Biochemists have performed detailed structural studies that reveal for the first time how an enzyme key to DNA replication stalls when an error occurs, to allow it to be corrected. Without such instantaneous braking, such mistakes in DNA replication would wreak havoc on DNA replication, killing the cell.
To their surprise, the scientists observed how the enzyme, DNA polymerase, retains a "short-term memory" of mismatches, in some cases halting itself past the point of the mismatch, so that the repair machinery can go to work. They also found that the mismatch structures differed dramatically from those deduced from previous indirect biochemical studies.
In an article in the March 19, 2004, issue of the journal Cell, Duke University Medical Center biochemists Sean Johnson and Lorena Beese, Ph.D., described how they had conducted detailed structural analyses of DNA polymerase as it encountered each of the 12 possible kinds of mismatches possible in DNA replication.
In such replication, the polymerase sequentially attaches DNA units called bases along a single-stranded template DNA. The result is like constructing one rail of a spiral staircase, using the other rail as a guide; and the polymerase "translocates" the template strand through its active site like a thread through the eye of a needle.
In this replication process, the polymerase normally guides the template strand and assembles the complementary, growing "primer" strand by pairing each base with the correct counterpart -- always pairing adenine with thymine and cytosine with guanine.
When mismatches occur, the polymerase must instantly halt itself, triggering the mismatch repair machinery to launch into action, before replication can continue. This stalling is thought to occur because the polymerase-DNA structure is distorted by the mismatched bases, causing it to shut down.
The problem, said Beese, who is an associate professor of biochemistry, is that the critical molecular details of how such distortion acted to brake the polymerase have remained unknown.
"For 40 years, there have been biochemical studies trying to understand how polymerase achieves such a high fidelity of replication," said Beese. "It was known that the polymerase stalled, but it wasn't known why. However, these studies represent the first direct observation of the structural details of mismatches and how they interact with the polymerase. And they show why and how stalling occurs."
In their studies, Beese and graduate student Johnson used the analytical technique of X-ray crystallography. In this widely used technique, intense X-ray beams are directed through a crystal of a protein to be analyzed, and the pattern of diffractions analyzed to deduce the structure of the protein.
The first steps in their studies were to first crystallize the polymerase with a segment of DNA containing each type of mismatched pair of nucleotides. Importantly, said Beese, the loosely associated crystals were so constructed that the polymerase could actually carry out several replication steps within the crystal.
"We have been able to replicate and translocate up to six base pairs in the crystal -- to my knowledge the biggest such motion ever seen in a crystal," said Beese. Using this approach, the Duke biochemists engineered the polymerase to be error-prone, so that they could produce crystals with mispaired bases inserted in the active site.
What's more, they were able to move the mismatch away from the active site and still detect the distortion of the polymerase structure that would indicate the polymerase was "sensing" a mismatch. Thus, the enzyme could "remember" a mismatch after it had occurred.
"What was surprising about this finding is that prior to the study a mismatch was thought to induce only very small local distortions right around the mismatch," said Beese. "But what we saw is that the polymerase amplifies this distortion back to the active site." However, cautioned, Beese, the full details of the stalling mechanism under all possible conditions remain to be worked out. So, there could be other details of the stalling mechanism that could affect understanding of this "memory," she said.
Significantly, said Beese, she and Johnson discovered that both the growing primer strand and the template are involved in the stalling process.
"Although each mismatch is different, we saw that it isn't just on the primer side that the structure is disrupted by a mismatch, but also on the template side, and sometimes both. And we also saw a mechanism we hadn't expected at all, which is that some mismatches just get stuck and don't translocate."
Although Beese emphasized that their studies are quite basic, such findings could help explain how the polymerase-triggered repair system is affected by DNA damage from carcinogenic chemicals.
The next steps in their research, said Beese, will be to instantaneously capture the polymerase in the act of processing a mismatch. The researchers plan to use flashes of ultraviolet light to unleash "caged" chemicals that trigger replication -- and at the same time use flashes of X-rays to illuminate the crystal. This approach may allow the researchers to make a movie of the polymerase during the synthesis and mismatch detection process.