Craighead reports on use of nanotechnology to count biological molecules

By Bill Steele

SAN FRANCISCO -- Up to now, most biologists have studied the molecules of life in test tubes, watching how large numbers of them behave. But now Cornell researchers are using nanotechnology to build microscopic silicon devices with features comparable in size to DNA, proteins and other biological molecules -- to count molecules, analyze them, separate them, perhaps even work with them one at a time. The work is called "nanofluidics."

"This will expand the methods for analyzing very small amounts of biochemicals and create new abilities unanticipated by the test-tube methods," said Harold Craighead, Cornell professor of applied and engineering physics and director of the Cornell Nanobiotechnology Center (NBTC).

Craighead described some of his laboratory's work in a talk, "Separation and Analysis of DNA in Nanofluidic Systems," at the annual meeting of the American Association for the Advancement of Science (AAAS) here, Feb. 15. The talk was part of a two-day seminar on nanotechnology.

Craighead's work began with a quest for an "artificial gel" that would replace the organic gels used to separate fragments of DNA for analysis. Traditionally this has been done by a process called gel electrophoresis. Enzymes are used to chop DNA strands into many short pieces of varying length. The sample is placed at one end of a column of organic gel and an electric field is applied to force the DNA to move through the gel. As they slowly snake their way through the tiny pores of the material, DNA fragments of different lengths move at different speeds and eventually collect in a series of bands as a ladder-like structure that can be photographed using fluorescent or radioactive tags.

The resulting image, Craighead explained, is just a list of the lengths of the fragments, from which scientists can read out genetic information. So he looked for other ways to sort DNA fragments by length that would allow scientists to work rapidly with small amounts of material.

The Craighead research group, which includes Jongyoon Han, Stephen W. Turner, Mathieu Foquet and Mario Cabodi, manufactured a variety of silicon-based nanostructures with pores comparable to the size of a large DNA molecule.

A microphoto of a capillary (horizontal) passing between waveguides (vertical) to enable detection and measurement of single DNA molecules. The image combines a bright field light source and laser excitation. The bright spot is due to scattered light in the probe volume. The capillary is one micrometer wide and 30 microns long. Courtesy Craighead Group/Cornell

The group has explored three approaches to DNA separation:

• Devices with openings that are less than the "radius of gyration" of the DNA fragments: When suspended in water, a DNA chain quickly coils into a roughly spherical blob. When pressed against a barrier with openings smaller than the radius of this form it must uncoil to pass through. Paradoxically, larger molecules do this more quickly, because they press a broader area against the obstacle, offering more places where a bit of the chain can be drawn in to start the uncoiling. When an electric field is used to drive a mix of DNA fragments along a channel with several such barriers, fragments of different lengths will move at different speeds, arriving at the far end in a series of bands not unlike those seen in gel electrophoresis.

• Sorting by physical length: A DNA sample is placed just outside a "forest" of tiny pillars arranged in a square grid, and an electric field applied to force the molecules to move into the grid. (Imagine pulling a coiled watch spring into a long, straight alley.) If the electric field is turned on just briefly and turned off before the molecule gets all the way in, the uncoiled portion will snap back out, just as the watch spring will pull back into its coil. But once the entire molecule is inside the grid, there is nothing to pull it back out. By varying the length of the electric field pulse, the researchers can control the length of the DNA strands that are collected in the grid. In addition to providing a way to measure strand length, Craighead said, this tool could be used to separate DNA for other work.

• Lateral diffusion by length: When moving through a grid of tiny pillars, DNA chains are constantly buffeted by moving water molecules that can knock them off-track, a process called "Brownian motion." If the pillars are flat vanes, all angled in the same direction, movement of all the chains will be skewed to one particular side. Shorter, lighter molecules will be pushed farther, so molecules can be sorted or measured based on the distance they are moved across the track when they emerge from the grid. Craighead calls these devices "Brownian ratchets."

These techniques all work with molecules en masse, but Craighead's group is also studying ways to work with single molecules, or at least to work with molecules one at a time. They have built microscopic tunnels just large enough for DNA molecules to run through in single file. Nanofabricated light pipes are placed on either side of the tunnel. Although very large, DNA molecules are still too small to be seen directly by visible light, but they can be tagged with other molecules that fluoresce when exposed to an ultraviolet laser and the fluorescence can be detected, with larger molecules giving off longer pulses of light.

In addition to counting the number of molecules of a given size in a sample, these devices could incorporate switches that could shunt molecules of different sizes into different channels, Craighead said.

The Craighead group collaborates with Watt W. Webb, Cornell professor of applied physics and S.B. Eckert Professor in Engineering, and his group, Jonas Korlach, Michael Levene and Dan Larson. They also work with professor Tom Duke of Cambridge University and professors Ted Cox and Bob Austin of Princeton University and their research group, Jeff Chou, Olgica Bakajin and S.S. Chan.