Scientists across the country have been poking and prodding at nanopores, and high-tech companies are starting to clue in to the technology's potentially revolutionary value. Still, the diminutive pore is a work in progress that has yet to prove itself. And if this high-speed sequencer does make its way into the wider community, a new stream of questions will need to be answered. For Americans leery of national identification cards, the advent of individual DNA sequencing might open the door to unwanted peeking into the most private details of a person's identity.

The nanopore sequencing idea first surfaced more than a decade ago on a scientist's road trip from Oregon to California. Reminiscing in his Santa Cruz office, that scientist-biochemist David Deamer pulls out a notebook from a triple-tiered file cabinet. He flips open the cardboard cover and peruses the first two pages, covered with cursive notes and sketches in red ink. His main drawing shows a series of shapes parading in a single file through what looks like a narrow tunnel. Each shape is a pentagon with squiggles sprouting from its corners like wildly-waving arms and legs. The pentagons come in four types. Each one is a building block — called a base — that makes up the DNA of genes. Genes, in turn, are the blueprint for the body's appearance, attributes, and often, its predisposition to disease.

Deamer's scribblings serendipitously came out of his past research into the origins of life. In order for life to begin, certain chemical compounds had to get together and start interacting from scratch. In one project, Deamer looked specifically at enzymes, which he collected in a hollow ball of a cell membrane called a vesicle. During his experiments, he noticed that a tiny hole in the vesicle could let in molecules like ATP — the cell's energy source — while also keeping in the enzymes needed to jump-start life. "I said, well, if I can get ATP in, why can't I get a molecule of DNA in?" Deamer recalls. "And that's really what bubbled up in my mind."

Then, during Deamer's road trip, his musings led him to a brilliant idea: What if he ran an electric current through the pore while a DNA molecule went in? As each base of the DNA passed through, it would momentarily block the current. And because each base is slightly different, it would interfere with the current in a unique way that would single it out, just like a name tag. Deamer could then pick out bases in their order of appearance-and at a rate hardly imaginable even with today's high-speed computers. "If we could have this working right now in our laboratory, one instrument would equal the entire world's ability to sequence DNA," he says. "That's how fast it really is."

In 1991, Deamer teamed up with Branton, who was his post-doctoral adviser when the two worked at the University of California at Berkeley, to test his road-trip daydream in the lab. Searching for a suitable pore, the pair looked at a well-studied protein called alpha-hemolysin, which comes from a toxic bacterium. The bacterium, which causes staph infections, uses the protein to punch holes in a cell's membranes in order to sink the cell like a ship bombarded with cannonballs. The cell-killing protein forms a passageway just large enough for only one chain of the double-stranded DNA molecule — which resembles a twisted ladder — to pass through.

By this time, Deamer had contacted biophysicist John Kasianowicz at the National Institute of Science and Technology in Washington, D.C., who was studying the same alpha-hemolysin pore in his own research. In 1993, the two scientists conducted a test in Kasianowicz's lab. They placed the protein in an artificial membrane in a solution of potassium chloride. Then they set up an electrical current through the fluid from one side of the membrane to the other through the nanopore. With the current up and running, they threw genetic molecules into the mix.

The first molecule to squeeze through the tiny tunnel was RNA, a single-stranded relative of DNA that helps cells make proteins. At its smallest point, the pore is less than 2 nanometers wide — 2 billionths of a meter. Just as they'd hoped, the RNA molecules made the current plummet as they slid through.

"When Kasianowicz and I saw that, we realized that we really did have a detector," Deamer recalls. He points to a picture of the current — a straight horizontal line, with long, thin spikes poking down from it. Each downward spike shows the current was blocked for a fraction of a second by a passing RNA molecule.

Deamer's group had shown that the pore could tell when genetic material was traveling through. But what they really wanted was for it to detect the sequence of RNA or DNA bases as they zoomed single file through the hole. The researcher who began tackling that challenge in 1995 was biochemist Mark Akeson, who joined Deamer's lab from the National Institutes of Health. Akeson began studying how to use the nanopore to distinguish between different bases on long strands of molecules. He made his own strands, consisting of a group of the same building blocks all in a row. For example, one molecule was a long series of adenines-one kind of base-while another was all cytosines. Then he pushed each chain through the pore. As each one went through, it blocked the current in a different way.

Akeson was the first to demonstrate that the method could actually detect sequences, albeit not on a base by base level, says Deamer, leaning back in his office chair and interlacing his fingers. "We realized then-Mark realized-that we might be able to fulfill our dream of sequencing in a very rough way," he says.

HAVING SET THE STAGE with a prologue explaining his team's work, Deamer displays the star of the show. The prototype nanopore lives in a cramped corner of his lab, within the walls of a cubicle. From the outside, a small, thin aluminum box sitting on the cubicle's desk looks like it might hold dominoes or a small video game. Actually, the box is the stage where all the action happens.

Graduate student Veronica DeGuzman squints through a microscope, focusing on a glass slide positioned atop the box where she's trying to place the pore. Manufacturing the pore itself for an experiment is no easy trick. The slide must be coated with the thinnest layer of fatty molecules called lipids, creating a cell membrane structure. DeGuzman dips a small paintbrush into a finger-sized test tube containing the oily lipids. She scrapes out a tiny, sticky droplet with the brush. Then, she picks up another brush, this one topped with a single sable hair. She holds up the two brushes to the light, trying to tease away one slippery lipid from the droplet to transfer to the slide.

After assembling the basic lipid membrane, DeGuzman adds the alpha-hemolysin protein to create the pore. Then it's a waiting game. "It's just like fishing," says Akeson. "You have to wait and see if you've caught something." Sometimes this process takes an hour, sometimes a day, but eventually the protein punches a hole through the lipid layer. Because the pore narrows as it goes through the membrane, it prevents double-stranded DNA from squeezing by. But single-stranded DNA and RNA can speed right through. "It would look like somebody slurping up a spaghetti noodle," says Akeson, pursing his lips into an O-shaped circle.

But so far, the pore slurps too quickly to read DNA base by base. A single base zips through the pore in microseconds. This doesn't give the computer program that reads the electric current enough time to differentiate between the normal and blocked current. To slow things down, Deamer's lab tried working with a convoluted form of DNA. By using single-stranded hairpin DNA — a string of bases that folds up into a U-shaped structure-the two found that the pore could savor a single base at a time. As it tries to fit through the narrowing pore, the hairpin DNA gets stuck. Then it begins to unfold, stretching one end into the pore. By measuring how the current is obstructed, the researchers can distinguish between the last sets of bases on different hairpins. This has given the researchers a hint of single-strand sequencing, though only for the bases sitting at the very end of the hairpin.

So far, no one has yet reached the goal of reading a full strand of DNA base by base. Even the researchers confess they aren't completely sure it can be done. "It still remains, from the point of view of sequencing, a high-risk project with a high-risk payoff. There's no guarantee it's going to work," says Harvard's Branton.

Even so, nanotechnology companies are taking the gamble. In May 2001, Agilent Technologies in Mountain View, California, began working with Branton, hoping to develop the first commercially available nanopore. But Agilent's pore is different: it's carved out of a synthetic material called silicon nitride. Branton and Harvard physicist Jene Golovchenko blasted a stream of ions at a sheet of the synthetic material to drill a hole for DNA to pass through.

The silicon nitride nanopore is much tougher than its biological counterpart. It should be able to handle endless streams of DNA for widespread sequencing use. "Unlike alpha-hemolysin, these pores could be made and stored in a can, so to speak," Branton says. The researchers have shown that strands of 30,000 DNA bases can pass through their synthetic pore-chains thirty times longer than those sequenced by conventional machines. But so far, their pore has only been able to identify bases in groups of ten.

Despite the potential advantages of a synthetic pore, naturally-occurring pores may be able to read DNA more precisely, Akeson says. "Nature has made this protein to fit exactly with DNA," he says. The successful pore of the future, the one which can read DNA one base at a time, could have both biological and synthetic components, he adds.

Both biological and synthetic nanopores could beat out conventional sequencing when it comes to efficiency. All current sequencing is based on the Sanger method, developed in the 1970s. With this technique, enzymes chop strands of DNA into segments. The researchers separate the fragments by length, then piece them together to determine the sequence of their bases. Many fragments are needed to solve the puzzle, so the scientists must start out with many, many copies of the DNA.

Today, although many improvements to the Sanger method have been made over the decades, sequencing machines still require large volumes of DNA. But with the nanopore method, just one copy of the DNA is all it would take: A single strand could be decoded base by base without any amplifying, chopping, or reorganizing. "The next sea change [in sequencing] is going to have to be with a single molecule," Akeson says.

Akeson is hoping that grad student DeGuzman's current project is the work that will really establish single-strand nanopore sequencing as a viable technique. DeGuzman is investigating a different way of slowing down DNA's transit through the tunnel, so that the pore can read it. She's trying to use a DNA-chewing enzyme to control DNA's entrance into the nanopore. This enzyme grabs onto the double-stranded molecule and chops one of its chains into individual bases. The other strand escapes unscathed and feeds into the hole at the same slow rate at which the enzyme munches along.

DeGUZMAN AND OTHERS in Deamer's lab are working with collaborators across the country who are studying the nanopore from many angles. Many of the people working on the project predict that a commercially available pore could be running within ten years. But not everyone is so confident. For large-scale challenges such as the exploration of the human genome, a working nanopore sequencer would certainly be a godsend. "It's conceivable that everything you could do with a nanopore would be better," says Jeffrey Schloss, a cell biologist who heads the Technology Development Coordination program at the National Human Genome Research Institute. "The big question is, will it ever work?"

But even if the technology doesn't turn out to be the single-base sequencer of Deamer's brainstorm, Schloss says, it may have other useful applications. For instance, Andre Marziali, a physicist at the University of British Columbia in Vancouver, wants to use the nanopore to count pieces of DNA after they're already been separated according to length by conventional sequencing. Right now, scientists use a laser the size of a boxy computer to count the chopped-up DNA. Many sequencers also use expensive dyes to detect the DNA fragments. Nanopores could replace both the lasers and these dyes-bringing the whole process down to a smaller, cheaper scale-by directly tallying individual DNA molecules as they whiz through the pore. Marziali believes this application of the nanopore can be put into practice sooner than the elusive nanopore sequencer.

If a faster sequencing method succeeds, however, it is certain to raise new worries about the privacy of genetic information. A person's DNA sequence could become like a social security number, following a person from place to place-but with far more intimate personal details. But chemist Rashid Bashir, who is developing a synthetic nanopore at Purdue University, thinks the potential gains in health care outweigh these concerns. "Science and technology are really like tools," Bashir says. "Technology can be used for good reasons and bad reasons, like anything else."

Indeed, this technology may go places Deamer never imagined on that long car ride years ago. He's confident that scientists will come up with new ways of improving the basic nanopore-sequencing idea that no one's even dreamed of yet. "Someone's going to make it work," Deamer says. "We just hope that it's us."