Your independent source for Harvard news since 1898

Your independent source for Harvard news since 1898

Right Now | Self-Sequencing

A Personal Genome Machine?

March-April 2007

In a laboratory behind the Science Center, researchers are working on a high-stakes project at the nexus of physics and biology. If all goes according to plan—admittedly a big if, when treading terrain this new—the project could yield within a few months’ time a way to map a complete human genome in less than 24 hours, for a price in the hundreds or thousands, rather than millions, of dollars. The research, led by Rumford professor of physics Jene Golovchenko and Higgins professor of biology emeritus Daniel Branton, has immediate implications for preventing and curing disease.

The project aims to make DNA sequencing faster and cheaper by using a technology totally different from earlier methods that required creating many copies of each DNA fragment and used massive amounts of chemicals for the analysis. Golovchenko and Branton are using nanopores: similar to the tiny holes in cell membranes that open and close to let sodium and potassium ions in and out in a process crucial to just about every function of the human body, including heartbeat, nerve impulses, and circulation.

Everything nano is trendy these days; the utility of extremely small particles is being studied in fields as diverse as medicine, defense, communications, and public health. But Branton, a cell biologist, has been working on this particular small-scale problem for more than a decade.

His research began with an observation that Staphylococcus aureus, the bacterial culprit in staph infections, creates a toxin that makes holes in cell walls just big enough for a strand of DNA to pass through. Branton reasoned that because the four bases that make up DNA (commonly called A, C, G, and T after the first letter of each chemical compound’s multisyllabic name) each have different dimensions, he could tell which one was passing through the hole at a given moment by observing to what degree the pore was blocked, based on the number of ions that got through along with the DNA.

The process worked. And because the bases traveled through the pore at a rate approaching one million per second, “You really had something that was orders of magnitude faster than anyone had ever dreamed of,” says Golovchenko.

But the pores were unstable, so the method worked only for short segments of DNA, nowhere near long enough to sequence the six billion base pairs that represent a human being’s full set of 46 chromosomes (23 from each parent). That’s where Golovchenko came in. Drawing on his background in physics, he suggested using a nanopore made of silicon nitride (a glass-like compound known for its hardness that is used in electronic products to protect transistors from air and moisture) rather than Branton’s original choice of a lipid membrane (similar to the fatty envelope around cells).

The sturdier material enables the researchers to use a different type of current—electric rather than ionic—that is stronger and therefore easier to measure. They place wires on each side of the pore and cause electrons to jump between them and through the DNA, in a process called tunneling. The DNA doesn’t conduct the current, per se, but “if you get the wires close enough, it jumps,” Branton says, leaving a characteristic electrical signature that distinguishes the four bases.

The pores in the silicon membrane are just 1.5 nanometers wide. That’s 1.5 x 10-9 meters, a billion times smaller than a meter, and about one ten-thousandth the breadth of a human hair. Nanotechnology is generally defined as involving objects measuring 100 nanometers or less, so the pores used by Branton and Golovchenko are small even by these standards.

The Harvard colleagues are competing with many other scientists in a challenge issued by the National Institutes of Health to produce a sequencing method that costs less than $10,000 per genome by 2009, and a method for $1,000 or less by 2014. Even after a reliable technology is developed, however, widespread availability will depend on creating the complementary infrastructure. Scientists are well on their way to creating a relatively cheap and easy genetic fingerprint; from there, someone must still devise a way to analyze it. It remains to be seen who will develop the mega-fast processing capabilities and giga-scale data storage capacity this nanotechnology requires.

DNA sequencing research is sure to touch off a complex ethical debate. Using DNA profiles to identify people, for example, would presumably be far more reliable than using a fingerprint, but also far more invasive of privacy. The research also raises the specter of genetic engineering of humans, although that’s something “no responsible scientist is contemplating,” Branton notes. He and Golovchenko consider themselves lucky to be working on a topic with such obvious benefits to society. “Some scientists have a very fuzzy idea of how their science might be useful in the future,” says Golovchenko, but “we have a very clear idea of how it could be used.”  

~Elizabeth Gudrais

Nanopore group website: www.mcb.harvard.edu/branton