ÐÓ°ÉÔ­´´

Every day of your life, each of your cells is bombarded by hostile forces
that damage the DNA bases forming the alphabet of your genetic code. The
enemy – which includes chemicals, drugs and radiation, as well as naturally
produced free radicals – causes an unknown but probably vast number of DNA
defects that can garble the genetic message.

Not to worry. Your body is equipped with teams of repair enzymes that
provide a roving service to fix these glitches. Should these enzymes fail
to do their job, your damaged cells will normally either commit suicide
by the process of apoptosis (‘Making friends with death-wish genes’, New
ÐÓ°ÉÔ­´´, 30 July) or be destroyed by your immune system. This may sound
draconian, but the alternative is far worse. If these safety mechanisms
fail, defective cells are left free to proliferate uncontrolled. The result
is cancer.

The part DNA repair plays in all this is a complex one. Biologists know
that at least some common forms of cancer may be directly linked to faulty
DNA repair. This became clear last December when two teams of scientists,
one in Boston the other in Baltimore, made the headlines with their discovery
of a genetic mutation that predisposes people to colon cancer and cancer
of the uterus. It turns out that the healthy gene codes for an enzyme involved
in correcting DNA defects that might otherwise corrupt genes – a ‘DNA repair
enzyme’. Now the hunt is on for other genes encoding DNA repair enzymes,
and for evidence that mutations to repair genes might be linked to cancer.

So far, researchers have identified over 20 mammalian DNA repair genes,
and they expect to find several more. A dozen repair genes have been cloned,
and scientists are excitedly studying them. How do the enzymes they code
for accomplish the massive task of correcting huge numbers of errors in
DNA? And what happens when they fail to do the job? By piecing together
a detailed picture of how DNA is repaired, and fitting this into the broader
scheme of how cells choose whether to divide or die, researchers are hoping
to increase their understanding of cancer. One important question facing
them concerns the age-old problem of nature-versus-nurture: how big a part
does innate genetic make-up play in causing cancer, compared with exposure
to outside influences?

SIMPLE TEST

Policy makers aiming to protect the public from carcinogenic agents
would love to know the answer. Unfortunately, there seems little hope of
finding a simple relationship in which inherited defects in the genes that
control DNA repair lead directly to cancer. Much more likely is the alternative
that individuals born with defective DNA repair enzymes may be particularly
sensitive to certain carcinogens. It is difficult to measure the genetic
component here because faulty DNA repair in itself does not normally cause
cancer; apoptosis and immune cells can usually compensate. So researchers
trying to make sense of DNA repair mutations must also consider mutations
in genes that control apoptosis and the immune response.

Despite this complexity, the discovery of the gene associated with familial
colon cancer means a simple diagnostic test is on the cards to tell people
whether they carry the mutant form of a repair enzyme. Armed with such
information, some people might decide to avoid certain jobs or habits that
could lead to damaged DNA. But a test would be a mixed blessing, as medical
and life insurance companies could use its results to discriminate against
people with defective genes. One thing is certain: if mutations in DNA repair
enzymes do turn out to be powerful pre-dictors of who is most at risk of
developing cancer, the interest won’t be confined to cancer researchers
and doctors.

FINDING FAULT

Less clear is whether understanding the mechanisms responsible for DNA
repair will produce a fast track to cancer therapies. Clinicians are, however,
starting to see that faulty DNA repair is a common factor in many cancers.
The link was first made as long ago as 1968, but until recently faulty
DNA repair was known to be involved only in a small group of extremely rare
heritable diseases where the connection with cancer is unclear. Patients
with the conditions known as xeroderma pigmentosum, Cockayne’s syndrome
and PIBIDS are all extremely sensitivity to the damaging effects of ultraviolet
rays, yet only those with xeroderma pigmentosum suffer multiple skin cancers.
Only now, with a more complete understanding of DNA repair, can we explain
why this is.

In the past decade, researchers have identified several mechanisms by
which DNA is repaired. At the Imperial Cancer Research Fund’s Clare Hall
Laboratories in South Mimms, Hertfordshire, Thomas Lindahl has been studying
a process which relies on direct reversal of the damage. One way in which
damage can occur is by a methyl group becoming attached to a DNA strand.
It can be removed without further disruption to the genetic material, Lindahl
has found, but this process is the exception. Repair mechanisms usually
require the cutting out of a single damaged base or a string of nucleotides
– the base-and-sugar units that link together to form DNA’s backbone – around
the damaged site. These include ‘nucleotide excision repair’, characterised
by the removal of around 30 nucleotides, and ‘mismatch repair’ where a much
larger chunk of genetic material is replaced.

Despite the variety of processes, several stages are common to most
DNA repair. These in turn are controlled by a wide variety of enzymes. In
the first step, enzymes must find the damaged DNA. Then the DNA helix must
be unwound. Next, the damaged area is cut out, and finally it is replaced
with new bases. To complicate matters further, repair follows two distinct
pathways, depending on whether the damage is in ‘silent’ genes – those that
are inactive – or ones that are replicating as part of the process of manufacturing
proteins or during cell division. The former, known as ‘overall genomic
repair’, is slow, but if it fails altogether mutations become permanent
when the cell’s DNA is replicated, which may then lead to runaway growth
and cancer. This is what happens in people with xeroderma pigmentosum. The
second pathway, called ‘active gene repair’ is much faster. Here the consequences
of the repair failing are less serious: the replication process will usually
be aborted, so reducing the risk of cancer. This is largely what happens
in patients with Cockayne’s syndrome and PIBIDS.

Repair of active and silent genes occurs through quite separate mechanisms
– though they do have some enzymes in common. In the inactive genome, recognising
damage is a formidable task since the bases that carry the genetic message
could be buried anywhere within the twisted DNA molecule. Making things
doubly difficult is the fact that DNA is often wrapped in sticky proteins,
while damage can lead to other bulky molecules becoming attached to it.
So how do enzymes distinguish between a protein that should be attached
and one that shouldn’t? And how do they decide which loops of DNA are normal
and which are the result of damage? So far, these mysteries remain unsolved.
But it is clear that each enzyme is capable of recognising particular kinds
of DNA errors.

In active gene repair DNA damage is flagged while a gene’s code is being
read and converted into the RNA messages that function as the recipes for
building proteins. The workhorse in this process is an enzyme complex known
as RNA polymerase, which reads genes by sliding along DNA, like a nun fingering
her rosary beads. When RNA polymerase encounters certain glitches, it gets
hung up and stops moving along the gene. At this point RNA polymerase is
pushed out of the way by an enzyme called transcription repair coupling
factor, which summons repair enzymes.

Aziz Sancar and his colleagues at the University of North Carolina at
Chapel Hill have identified transcription repair coupling factor in Escherichia
coli bacteria, and believe they are on the trail of its human counterpart.
The protein is thought to be the defective link in people with Cockayne’s
syndrome and PIBIDS. Unable to repair certain kinds of DNA damage, these
patients suffer stunted growth and neurological degeneration – probably
caused by large numbers of cells dying. But they do not have an increased
risk of cancer because damaged cells are cleared by apoptosis and by scavenger
cells in the immune system.

CUT AND PASTE

A second enzyme under investigation at Sancar’s laboratory has wider-ranging
application. Excision nuclease, or exci-nuclease, is responsible for nucleotide
excision repair. Richard Wood, a senior scientist at the Clare Hall Laboratories,
explains that excinuclease cuts DNA strands on either side of the damage
at characteristic distances of around 25 nucleotides on one side and 5 nucleotides
on the other. The enzyme plays a role in repair of both active and silent
genes and is the defective enzyme associated with xeroderma pigmento-sum.
Having probed the talents of excinuclease for a decade, Sancar realises
he has struck gold. ‘I found it repaired everything,’ he says. ‘It’s probably
the most important defence mechanism against cancer.’

Teams in Britain and the US are now busy working out the details of
how this multifaceted enzyme works. A complex picture is emerging. Human
excinuclease is made up of no less than 16 protein subunits. People with
xeroderma pigmento-sum show a variety of defects in the genes for these
subunits. Indeed, Sancar and others now suspect that faults in this enzyme
may be much more common than the incidence of xeroderma pigmentosum suggests.
They think that people who suffer from symptoms similar to those associated
with the disorder, but in a milder form, may harbour subtle defects in
their excinuclease genes.

This idea has gained favour since the discovery of the gene associated
with hereditary nonpolyposis colon cancer. Teams led by Bert Vogelstein
from Johns Hopkins University in Baltimore and Richard Kolodner of the Dana-Farber
Cancer Institute in Boston found that people with mutations in this gene
– which encodes a protein called mismatch repair enzyme – have a 70 per
cent chance of developing colon cancer at some time in their lives, and
women with the mutation also have at least a 50 per cent chance of developing
cancer of the uterus (This Week, 11 December 1993). Researchers now suspect
that one person in every hundred has a faulty mismatch repair enzyme and
that it predisposes them to all sorts of cancers, particularly in tissue,
like the colon, where apoptosis is not triggered as easily as elsewhere.

Mismatch repair enzyme is involved in active gene repair. It normally
steps in when base pairs slip out of alignment during replication and produce
a mismatch of bases. According to Kolodner, the enzyme’s job is the cellular
equivalent of a spellchecker. Mutations in the gene encoding it impair its
function but do not halt the replication process. So defective DNA can replicate,
and multiple mutations accumulate. This is the pattern found in cancer cells
taken from patients with hereditary colon cancer.

DEADLY PATTERNS

This form of cancer is not the only one in which tumours show distinctive
patterns of mutations, with some genes much more commonly damaged than others.
These patterns had puzzled clinicians for decades, but it now appears that
a better understanding of DNA repair may provide the answers. Mutagens
such as reactive chemicals and radiation seem to attack only specific target
sequences in the base code. In the past, researchers assumed that these
hot spots for genetic damage were the same ones that appear mutated in cancer
cells. But oddly, some commonly damaged sites in DNA are rarely found mutated
in tumour cells. Conversely, some sites that are rarely damaged in DNA are
commonly found to be mutated in tumour cells.

The missing part of the story is the rate at which damaged DNA is repaired.
Two new studies show that not all regions of DNA receive the same attention
from repair enzymes. And the pattern of mutations seen in cancer cells
seems to be as much a function of the variable speed of DNA repair as of
the agent inducing the damage. This is particularly crucial when the mutated
gene is involved in cell growth, and could trigger a tumour if the fault
is not repaired quickly. Gerd Pfeifer and Silvia Tornaletti at the Beckman
Research Institute at the City of Hope Hospital in Duarte, California, have
studied such a gene, and reported their findings in March this year. Known
as p53, the gene responds to DNA damage in one of two ways: by arresting
growth so that the damage can be repaired, or by choosing the suicide pathway.
Mutated p53 appears in half of all human cancers.

Using a new technique that distinguishes individual nucleotides, Pfeifer
and Tornaletti measured the repair rate at eight positions in the p53 gene
that are frequently seen mutated. In human fibroblast cells irradiated with
ultraviolet light, they found that seven of these dam-age sites underwent
very slow repair. So it seems that the DNA repair enzymes are unable to
do their job before the mutated gene starts replicating out of control.

In a related study, Gerald Holmquist and his colleagues, also at Beckman,
showed that the rate of DNA repair can vary 15-fold from nucleotide to
nucleotide within a human gene. The researchers irradiated human fibroblasts,
and then mapped repair rates along a well-studied gene called PGK1, which
produces an enzyme involved in energy production. The team showed that rates
of DNA repair can be altered by the presence of proteins that bind to DNA,
such as transcription factors. These are known to alter the shape of DNA,
and it seems likely that this makes some damaged sites more accessible to
repair enzymes while obscuring others. PGK1 seems to have two binding sites
where transcription factors reduce the rate of repair. But such findings
are still far from explaining the huge variation in the quality of repair
service.

Looking at the puzzle from another angle, some scientists think there
is a link between cancer, DNA repair and ageing. After all, the incidence
of cancer climbs dramatically between the ages of 50 and 70. ‘My view of
cancer is that it’s a premature ageing disease,’ says Lawrence Grossman,
a biochemist at Johns Hopkins School of Public Health in Baltimore.

Grossman’s group recently tested this theory by evaluating the DNA repair
capacities in people whose ages ranged from 20 to 60. Of these subjects,
88 had basal cell carcinoma and 135 were free of the disease. The researchers
collected white blood cells from these donors, transfected the cells with
damaged DNA and then 40 hours later measured how much damage had been repaired.
The group reported last year that the DNA repair capacity of cells from
the 135 healthy individuals varied strongly with age. Indeed, the results
suggest that every year between the ages of 30 and 60 we lose about 1 per
cent of our ability to fix faulty DNA. Even a 1 per cent change is significant:
it means that 10 damaged nucleotides within a gene may go unrepaired, says
Grossman. And it only takes one damaged nucleotide to produce a defective
protein.

The researchers also found something striking from the group with skin
cancer. ‘When we look at people 30 to 40 years of age with basal cell carcinoma,
we see they have the repair capacity of someone 40 years older,’ says Grossman.

OLD COPY

Which enzymes may be affected by age is not yet known. But some cancers
and diseases of ageing might result from repair enzymes faltering as their
own genes become damaged. One idea is that these genes end up as victims
of attack by free radicals – highly reactive and destructive chemicals generated
during inflammation and within all cells when oxygen is metabolised. Bruce
Demple, a biochemist at the Harvard School of Public Health in Boston, who
studies the repair enzymes that fix free-radical damage to DNA, is keen
on the idea. But as he admits, evidence that this is what actually happens
is proving hard to come by. ‘We haven’t found any human disease that results
from a deficiency in oxidative repair,’ he says. ‘And believe me, we’ve
looked. It’s very surprising.’

International efforts to map and sequence the human genome could help
reveal new links between the genetics of DNA repair, cancer and ageing.
Philip Hanawalt is a molecular biologist at Stanford University in California
who also sits on a committee in California that is reviewing the Environmental
Protection Agency’s list of carcinogens. ‘As we get more mechanistic information
about DNA repair, we should be able to come up with more rational lists
of carcinogens,’ he says. The link between mutation of the mismatch repair
gene and colon cancer is just the beginning.

Karen Schmidt is a writer based in Chapel HIll, North Carolina