Images of enzyme suggest way to improve DNA sequencing
St. Louis, Aug. 16, 1999 -- Like a temperamental copy machine, the most commonly used enzyme in DNA sequencing has a few annoying quirks. It generates pages with blank spots, for example. But new X-ray images of the enzyme at work have suggested a way to fix these problems. The strategy works in the lab and is being tested by several companies in the United States.
"We expect this will make the enzyme an even better sequencing tool for the Human Genome Project," says Gabriel Waksman, Ph.D., an associate professor of biochemistry and molecular biophysics at Washington University School of Medicine in St. Louis, one of the world's major sequencing sites. Waksman and colleagues report their findings about the enzyme, Taq DNA polymerase, in the Aug. 17 issue of Proceedings of the National Academy of Sciences.
The Sanger method is the most common method for determining the order of genetic letters, called nucleotides, in DNA. It uses Taq to copy segments of DNA into chains of different lengths. Because Taq comes from a bacterium, Thermus aquaticus, that flourishes in hot springs, it can withstand the temperature changes necessary for copying.
Four genetic letters -- A, T, C and G -- make up the DNA code, and all of these nucleotides are included in the reaction mixture when Taq is put to work. But because a nucleotide-like compound called a dideoxynucleotide triphosphate -- ddC, ddG, ddA or ddT -- also is included, Taq stops copying when it incorporates that compound, like a train that halts on the track when it hits a rock. Therefore, the length of a new DNA chain indicates where that particular nucleotide occurred. After the copies are separated by size and reacted with fluorescent chemicals, DNA sequencing machines can determine which nucleotide terminates each chain. Because A is complementary to T and C to G, this reveals the order of the nucleotides in the original piece of DNA.
Waksman and colleagues obtained X-ray images of Taq at work after crystallizing it with DNA and each of the four dideoxynucleotide triphosphates in turn. They published the first image -- Taq DNA polymerase combined with ddC -- in The EMBO Journal last year. In the current study, they compared all four images and uncovered the anomaly that hampers Taq's performance in the lab. Then they used this structural information to fine-tune the enzyme.
Taq is shaped like a hand with fingers, a thumb and a palm. Dideoxynucleotides interact with the fingers. By comparing the images of the four enzyme-DNA-dideoxynucleotide complexes, the researchers discovered that Taq behaves differently when it interacts with ddG than when it interacts with ddA, ddC, or ddT. When it interacts with ddG, one amino acid building block in its fingers -- residue 660 -- binds to the dideoxynucleotide. But the same amino acid -- arginine -- doesn't participate when the enzyme interacts with any of the other three dideoxynucleotides.
One of the enzyme's shortcomings in the lab is that it incorporates ddG 10 times faster than it incorporates ddA, ddC or ddT. "So we thought that Taq might behave this way because of the additional interaction it makes with this particular dideoxynucleotide," Waksman says. Sure enough, when the researchers changed residue 660 from arginine to various other amino acids, the enzyme incorporated ddG at the lower rate. So altering the enzyme allowed sequencing to proceed at an even pace.
Unexpectedly, changing this amino acid also solved the problem of gaps in the sequence -- the blank spots on the copied page. When the enzyme has the usual arginine at position 660, the resulting sequence often has missing letters. But the enzyme without arginine at 660 made copies in which every dideoxynucleotide was detectable, so the sequence had no gaps. The modified enzyme has aspartate, leucine, serine, tyrosine or phenylalanine at this critical position.
"So it is clearly beneficial to have this mutation," Waksman says. "We expect that Taq DNA polymerases mutated at position 660 will help limit sequencing errors and reduce the necessity to sequence the same pieces of DNA multiple times, thereby decreasing labor and costs."
A grant from the National Institutes of Health supported this research.
Li Y, Mitaxov V, Waksman G. Structure-based design of novel Taq DNA polymerases with improved properties of dideoxynucleotide incorporation. Proceedings of the National Academy of Sciences, 96(17), 9491-9496, Aug. 17, 1999.
Copies of the paper are available from the PNAS news office, 202-334-2138, or by e-mail at: firstname.lastname@example.org