Stanford Scientists Use Noise To Sort Proteins

The device, which sorts molecules found in the cell membrane, is powered by an unconventional energy source, "thermal noise" -- the random variations in energy found in a population of molecules at a given temperature.

"The ability to separate membrane-associated molecules is extremely important," says Steven G. Boxer, professor of chemistry at Stanford, "because they play a vital role in the life of the cell and have proven exceptionally difficult to purify, characterize and study."

Boxer and postdoctoral fellow Alexander van Oudenaarden developed the new sorting system and report on its design and operation in the Aug. 13 issue of the journal Science.

Unlike conventional separation techniques, their system can separate membrane molecules in their native environment and do so continuously, rather than in batches. Because it relies on fabrication techniques developed in the microelectronics industry, the device should be inexpensive to manufacture in volume and relatively simply to modify for specific purposes, Boxer said.

The array is a realization of a theoretical device called a Brownian ratchet, popularized by Nobel physicist Richard Feynman, who used it to illustrate the inviolability of the second law of thermodynamics, the law that explains why it is always easier to make a mess than it is to clean it up.

The original Feynman version of a Brownian ratchet is a mechanical device that consists of two boxes joined by an axle. Four small vanes are attached to one end of the axle. Attached to the axle's other end is a ratchet, a saw-tooth wheel and a small pawl attached to a spring that permits the wheel to turn in only one direction.

The ratchet is powered by Brownian motion, the random motion exhibited by small particles floating through the air. The jerky motion is caused by the impact of air molecules that pound the particles from all sides. The ratchet allows the axle to turn in only one direction. So, when the ratchet's vanes are exposed to the same random bombardment, it appears to convert random motion into directional motion. As a result, the ratchet appears capable of performing work without adding energy. If the spring is light enough to allow the wheel to turn, however, it is also light enough so that thermal vibrations will cause the pawl to unlatch intermittently. As a result, molecular collisions are equally likely to push the wheel backward as forward.

Since Feynman's description, a number of scientists have explored Brownian ratchets. Most of the work has been theoretical. University of Chicago biochemist R. Dean Astumian, for example, has proposed a variant that employs asymmetric, time-varying electrical fields that work with Brownian motion to force electrically charged particles to move in a direction perpendicular to the field.

Van Oudenaarden and Boxer's device is the first demonstration of another basic type of Brownian ratchet, called a geometrical ratchet, that uses asymmetric barriers to harness thermal fluctuations to produce directional motion. Such devices were proposed in two theoretical papers published last year by Robert Austin of Princeton and Thomas Duke of Cambridge, and Denis Ertas of Exxon.

One of Boxer's former students, Jay Groves, now a Director's Postdoctoral Fellow at the Lawrence Berkeley National Laboratory, happened to hear Austin give a talk on the concept of geometrical ratchets and realized that a system he had helped develop while at Stanford could be adapted for this purpose. "I never would have heard of these papers if Jay hadn't gone to that talk," Boxer said.

Two years ago, Groves and Nick Ulman, then a research associate in applied physics, working with Boxer, developed a method for partitioning two-dimensional fluid membranes called the membrane chip. Borrowing microfabrication techniques from electrical engineering, they created a specially prepared surface that holds millions of cell-sized squares of an artificial membrane that closely mimics the surface of living cells. The membrane chip provided an entirely new approach to studying membrane-related cellular processes.

Boxer and van Oudenaarden realized that if they took a membrane chip and, instead of creating enclosed corrals, filled it with an array of microscopic barriers of the right shape, they should be able to make a special kind of two-dimensional ratchet, one that works on membrane-associated molecules.

So they designed and constructed such an array, filled it with artificial membrane, and tested it by adding charged, fluorescently labeled membrane molecules, called phospholipids, to one corner. Then they applied a small electrical field across the array and watched the fluorescent molecules diffuse. The flow pattern clearly showed that the device was converting Brownian movement into a net motion perpendicular to the direction of the electrical field.

The two also demonstrated that such a system can separate different kinds of membrane molecules by introducing two fluorescently labeled molecules, one with a single and the other with a double negative charge. Their measurements showed that the migration paths of the two types of molecules through the array were substantially different.

The test showed that the device can sort a large class of membrane molecules, including a number of important cell surface receptors, that move freely around in a membrane when it is supported by a solid surface. In addition, the researchers are working on new ways to attach membranes that they hope will allow them to study another important class of molecules, called integrated membrane proteins.

"This system combines interesting physics and engineering to produce a device that may prove useful for biology," says Boxer, "and we're not violating the second law!"

Funding for the research was provided by the National Science Foundation, the Netherlands Organization for Scientific Research and SmithKline Beecham.

[SIDEBAR] Bar room analogy for operation of membrane ratchet

Brownian motion is an example of what scientists call a random walk, a movement whose steps are determined at random. A colloquial term for such a path is a drunken sailor's walk. So an appropriate analogy to understand how the membrane ratchet works is a waterfront bar visited by a stream of heavily drinking sailors, Boxer says. The bar stretches all the way across the back wall, and the front entrance is on the left side. If the space between the door and the bar is empty, then the sailors will head straight for the bar. Although the sailors spread out a bit as they stagger, they end up predominantly on the left side of the bar. But if the floor is filled with tables that are angled from left to right, they will deflect the staggering sailors further down the bar to the right.

The furniture-filled bar also would sort the sailors depending on how thirsty they are. Extremely thirsty sailors would head to the bar the most quickly and so their staggering course will not be deflected as much, on average, as that of the sailors who are only mildly thirsty. So more of the thirsty individuals would wind up at the left end of the bar and more of the mildly thirsty would end up on the right end. The amount of deflection is also affected by a sailor's degree of drunkenness. The drunker the sailor, the more pronounced his stagger and the more his course is deflected to the right. The distribution of the sailors at the bar also can be changed by altering the angle of the tables.

Other relevant material:

Boxer group home page: http://www-chem.stanford.edu/group/boxer/