Contact

Duke Health News

919-660-1306

DURHAM, N.C. -- Biochemists have pinpointed how a flaw in
DNA that is central to mutations in cancer and aging fools the
cellular enzyme that copies DNA. Their finding explains how
oxidative DNA damage -- a process long believed to underlie
cancers and aging -- can create permanent genetic damage.

The Duke University Medical Center researchers' findings
were published online Aug. 22, 2004, by the journal Nature. The scientists were led by
Associate Professor of Biochemistry Lorena Beese, Ph.D., and
the paper's lead author was Gerald Hsu, a Duke M.D./Ph.D.
student. The other co-authors are Thomas Carell and Matthias
Ober of Ludwig Maximillians University in Germany. Their
research was supported mainly by the National Cancer
Institute.

DNA is a double stranded molecule shaped like a spiral
staircase. The two strands of the spiral are linked by
sequences of molecular subunits, or bases, called nucleotides.
The four nucleotides -- guanine, cytosine, adenine and thymine
-- naturally complement one another like puzzle pieces. In
normal DNA, a guanine matches with a cytosine, and an adenine
with a thymine. However, stray reactive oxidizing molecules in
the cell can alter guanine to become an "8-oxoguanine" that can
lead to a mismatch.

This mismatch occurs in the process of replicating DNA,
which begins when the two strands unzip. A protein enzyme
called DNA polymerase then works its way along one "template"
strand adding nucleotides to create a new double-stranded DNA.
In the replication process, the polymerase draws the DNA strand
through a small "active site" -- somewhat like a spaghetti
strand being drawn through a Cheerio.

Normally, this "high-fidelity" polymerase accurately adds
complementary nucleotides and detects any mistakes that have
been made. These mistakes or mismatches reveal themselves as
malformations that distort the active site -- like kinks in the
spaghetti strand that would clog the Cheerio. Such
malformations trigger a repair mechanism to correct the
mismatch.

The researchers' initial studies revealed that the
polymerase biochemically "prefers" to mismatch an 8-oxoguanine
with adenine rather than the correct cytosine. If not detected
and corrected, such a mismatch leads to errors in the cell's
machinery that can trigger the uncontrolled growth of cancer or
the death of cells in aging. However, researchers have long
known that the 8-oxoguanine-adenine mismatch seems to readily
avoid detection by the polymerase.

"There have been a number of studies of the kinetics and the
biochemistry of this mismatch reaction, but it was not
understood why this particular lesion evaded detection as well
as it does," said Beese. "It is one of a series of such
oxidative lesions, but it is considered the most mutagenic,
which is why we concentrated on understanding it."

In the experiments, Hsu worked with the particularly sturdy
polymerase enzyme from a thermostable strain of the bacterium,
Bacillus stearothermophilus, which thrives in geothermal hot
springs. He crystallized this enzyme along with a DNA strand
that contained an 8-oxoguanine. Because the polymerase retains
the ability to synthesize DNA in the crystal, Hsu then added
either the correct (cytosine) or incorrect (adenine)
nucleotides and observed the results.

Using X-ray crystallography, the researchers were able to
deduce with great precision the structure of the protein and
the DNA in the crystal. The series of crystals they analyzed
constituted snapshots of the polymerase's function as it
created both accurate and mutated strands from the
template.

The biochemists encountered a surprise when they analyzed
the polymerase crystals with either the correct or mismatched
nucleotides. "We saw that, ironically, when the polymerase
binds the correct cytosine opposite 8-oxoguanine, the structure
looked like DNA mispairs," said Beese. "This suggested that the
enzyme would stall and not readily proceed with
replication.

"But when we put in an incorrect adenine nucleotide, it
looked like a normal base pair in how it interacted with the
polymerase." The researchers' analyses revealed that the
mismatched combination of 8-oxoguanine and cytosine was
distorted, like a kink in a spaghetti strand that would jam the
active site. However, the mismatched 8-oxoguanine and adenine
showed no distortion so would proceed smoothly through the
polymerase to be incorporated into the new DNA.

"We were able to extend the replication process to show that
there were no distortions that would be detected by the
polymerase. This means that the DNA would continue to replicate
with this mispair, and that could potentially lead to stable
incorporation of a lethal mutation," said Beese.

In further analyses, Hsu confirmed that bacterial polymerase
would behave just as did the human polymerase in preferring to
incorporate the mismatch and failing to recognize it. Also,
they found, if the 8-oxoguanine-cytosine pair manages to pass
through the polymerase, the distortion disappears, meaning that
the chemically flawed guanine will persist in the DNA
strand.

In further studies, Beese and her colleagues are exploring
other types of DNA lesions and how they affect replication.
These lesions include those caused by major carcinogens. The
researchers also have developed a method to synchronize the DNA
replication process, so that they can make the equivalent of
X-ray crystallographic "movies" of the entire process, to
better understand it.