Tying up loose ends of the human genome: A team boosts gene-probe efficiency

A straight line has two ends. Tie it in a circle, and it has none. That difference has helped Stanford genome researchers take a big step toward being able to examine all of an individual's genes in detail, swiftly and cheaply.

In a paper published July 8 in the Proceedings of the National Academy of Sciences, School of Medicine researchers describe a way to simultaneously amplify hundreds (and potentially hundreds of thousands) of separate snippets of an individual's genome in a single test tube.

There is no universal human genome: rather, there are about 6 billion human genomes (we each have our own, except for identical twins) distinguished by myriad minuscule variations in DNA sequences. These tiny differences from one person's genome to the next influence everything from eye color to susceptibility to disease. So it would be a real advance if every patient's entire set of genes could be scrutinized for known or previously unobserved sequence variations. Finding such a variation in a crucial position on an individual's genome could throw light on his or her disease susceptibility.

To do that, though, would require making lots of copies of that person's genetic instructions—quickly, accurately and cost-effectively. That's what this method may someday do, according to the Stanford team headed by Ronald Davis, PhD, director of the Stanford Genome Technology Center and professor of biochemistry and of genetics.

"Other people have tried doing this, but it has been very inefficient," said Sujatha Krishnakumar, PhD, a research associate at the genome center and the first author of the study. "We have come up with a method that is efficient."

The human genome, residing in the nuclei of virtually every cell in a person's body, is a sequence of about 3 billion chemical units of DNA strung together like beads on a necklace—a necklace that, if stretched end to end, would be an astonishing six feet long. Yet genes, which actually dictate the compositions (and therefore the capabilities) of the all-important proteins that carry out most of a cell's day-to-day activities, occupy only a tiny fraction of that sequence. A mutation in a gene could lead to a change in a protein's ability to do its job and, perhaps, to altered disease susceptibility in a person who has inherited that mutation.

To complicate things, most of a person's 30,000-odd genes aren't intact "stand-alone" sequences but, rather, are split by swaths of intervening DNA into 10 or 15 protein-coding fragments called exons. Those 400,000 or so exons account for less than 3 percent of the vast human genome, but because it is their sequences that determine those of all our proteins, they are the part that's most worth looking at right now, said Krishnakumar. (The other 97 percent of the genome has important functions, too, but these are only now being mapped.)

The team's advance builds on a technique called polymerase chain reaction, the biochemical equivalent of a copy machine.

PCR begins with the synthesis, in large quantities, of a pair of small DNA strips called "primers" that are designed to perch like bookmarks just before and after a stretch of genomic DNA—they tell the biochemical copy machines what "page" to go to. The newly copied DNA strand, stapled by the copying machinery to the two primers, floats off its "template." New primers hop onto the spots vacated by the old ones, and the process is repeated, up to millions of times.

Copying 500 exons requires designing 500 unique primer pairs. While this is possible, actually using PCR to amplify 500 exons—let alone all 400,000—at once is a nonstarter, said Michael Mindrinos, PhD, associate director of the genome center and study co-author. You'd have to unleash not just two but 1,000 short DNA strips into the same mixture, Mindrinos said, and they would get in one another's way and wreak havoc.

So the Stanford investigators modified the procedure by tethering each primer pair to a long, linking molecule also made of DNA, the way one might tie a pair of tennis shoes together by one of their shoelaces. The linker molecule is generic, so it can be used to connect any primer pair. Importantly, it is also lengthy enough to span entire exons. Until now, mass-producing such long DNA molecules via conventional chemistry has been far too expensive to be practical. A critical step forward by the Stanford investigators was figuring out how do this cheaply, using sophisticated biochemical methods.

Now, when those 500 tethered pairs of primer sequences (analogous to 500 pairs of shoes) are placed into solution with genomic material, each pair finds its way to the exon it was designed to bracket. Just as in standard PCR, the copying machinery comes along to generate copy after copy, in each case immediately tying the new copy to the two bracketing primers. But because the primers' far ends are hooked to the long "shoelace" connector molecule, the resulting product is a closed ring of DNA.

Next, enzymes are unleashed that chew up linear DNA debris, leaving unscathed the circular (that is, closed-up) DNA. Other enzymes pop those open and chop out the desired exon copies in abundance, ready for further analysis.

Careful tweaking allowed the researchers to copy hundreds of diverse exons at reasonably uniform rates—a critical step, Mindrinos said. If final quantities were too disparate, it would be like a gumball machine holding 100,000 green gumballs for each red one. You'd go broke before you fished out a red gumball.

The Stanford team succeeded in getting abundant copies for more than 90 percent of nearly 500 exons they targeted with primers in a first round of the reaction. Then, in a second round, they coaxed more-recalcitrant exons into copying properly by customizing reaction conditions to suit those exons' biochemical quirks, boosting the initial 90 percent success rate to 95 percent. The goal now is to amplify 5,000 exons, and eventually all 400,000, simultaneously.

Other authors on the study were Julie Wilhelmy, research assistant at the Stanford Genome Technology Center; and Jianbiao Zheng and Malek Faham, both of Affymetrix. The project was funded by grants from the National Institutes of Health.

Bruce Goldman is a freelance writer.