杏吧原创

Sheffield

IT MIGHT not sound like the most delicate treatment in the world. In fact,
having your body operated on with hairpins and scissors sounds downright
dangerous. So it will probably come as a surprise to learn that chemists are
wielding these hairdressing tools in the fight against genetic diseases and
cancer. And it鈥檚 not hair that researchers are messing with, but the blueprint
of life itself鈥擠NA.

Our DNA contains more than 80 000 genes that control all the workings of our
bodies. Yet amid this complexity, even the tiniest of errors can be devastating:
fatal diseases such as sickle-cell anaemia and cystic fibrosis are caused by
simple, inherited genetic defects. Similarly, damage to DNA can trigger the kind
of uncontrolled cell growth that leads to cancer. Medical researchers would
dearly love to understand where and how this damage occurs, and to find ways of
switching off defective genes. With the help of their scissors and hairpins,
chemists may soon provide them with the tools.

Stripping DNA

Some of the things that give us cancer, such as carcinogenic chemicals and
ultraviolet light from the Sun, do their damage to DNA by stripping it of
electrons, thereby oxidising it. At the Beckman Institute of the California
Institute of Technology in Pasadena, Jacqueline Barton and her colleagues have
designed small molecules that mimic this destructive process. The molecules snip
DNA like miniature scissors and, in so doing, have revealed natural weak spots.
Also at Caltech, another team led by Peter Dervan has built a set of chemicals
shaped like hairpins, that can switch off specific genes. What both teams have
in common is that their molecules lock in specific ways to DNA鈥檚
superstructure.

DNA looks, at least in part, like a twisted ladder. Two long chains of units
known as nucleotides are connected by rungs made from pairs of chemical groups
called bases. The four types of base in DNA鈥攁denine (A), thymine (T),
guanine (G) and cytosine (C)鈥攃an only pair up in one way. Adenine pairs
with thymine and guanine with cytosine.

Molecules can dock with DNA in several ways. For example, just as a knife can
slide between the pages of a closed book, flat molecules can slip between two
rungs of DNA. This is known as intercalation. Alternatively, molecules can slot
into the grooves between the ladder鈥檚 鈥渦prights鈥�
(see Diagram).

Barton鈥檚 molecular scissors work by intercalating DNA. Each molecule is
composed of several large, flat hydrocarbon rings linked by a central metal ion.
When light shines on this ion, it tries to pull electrons from surrounding
molecules in a process known as photo-oxidation. If this happens when the
intercalator is locked to DNA, the metal ion grabs electrons from the DNA,
damaging nearby bases. When exposed to high-energy light, the molecule will even
cleave DNA in two as if it were a pair of light-activated scissors.

For the past two years, Barton鈥檚 team has been adding the scissors to
solutions containing strands of DNA with a known sequence of bases, and shining
low energy light on them. Although the scissors intercalate between all the
rungs of the ladder, the team has found that damage occurs only at places where
two guanine bases sit next to each other on one upright of the ladder. Clearly,
鈥淕G鈥� sites are oxidation weak spots. This finding has been confirmed by
researchers at the University of Utah headed by Cynthia Burrows. 鈥淕enerally, G
sites are more easily damaged, but we also find that GG and GGG sites are
particularly sensitive,鈥� says her colleague Steve Ross.

But the story doesn鈥檛 end there. Barton鈥檚 team also discovered that GG sites
are susceptible even when the scissors are docked some distance down the ladder.
Damage can occur when GG sites are 20 nanometres from the scissors, says Barton,
which on the molecular scale is a very long way away. Sites with three guanines
in a row are even more vulnerable.

The team is now working to understand how the electrons move along DNA. 鈥淲e
know that although the process is not particularly sensitive to distance, it is
extremely sensitive to how the base pairs are stacked,鈥� says Barton. For
example, when the team made a bend in DNA between a GG site and the
scissors鈥攂y inserting extra bases into one upright of the ladder鈥攖he
bend protected the GG site from damage.

Damage at a distance

Learning how oxidation causes damage at a distance like this is important
because it goes to the heart of the cancer process. Researchers have found that
guanine bases damaged by oxidation cause genetic mutations that may be important
in carcinogenesis. If researchers can locate vulnerable sites and understand how
harm is done, they can then look for ways to protect DNA from damage.

Barton鈥檚 team now has a number of different DNA scissors which intercalate
different sequences of base pairs and can be used for a variety of jobs. In
future, some could even be used to treat DNA damage. One of the most common DNA
lesions caused by exposure to ultraviolet light is for two thymine bases to bind
to each other. These thymine 鈥渄imers鈥� disrupt the information encoded
within DNA and are associated with skin cancer. In the past year, Barton and her
colleagues have shown that the scissors repair this damage by selectively
breaking the T-T bond.

The team has also used its chemical creations to explore the fine detail of
how DNA is exploited within living cells. The sequence of base pairs in a gene
contains all the information needed to make that gene鈥檚 protein. But between
gene and protein lies a long series of biochemical steps. First, a protein
called a transcription factor binds to the DNA. Then, a vast complex of other
proteins pull apart the two strands and makes an RNA copy of one of the strands
(鈥淧laying to win鈥�, New 杏吧原创, 12 April 1997, p 38). After this
process, known as transcription, the RNA migrates out of the cell nucleus, where
it is used as a template to make the protein.

One key piece of information that researchers want to know is where the
transcription factor for a gene binds to DNA. Block this and you stop that
gene鈥檚 protein being produced. Barton can use her scissors to identify these
sites.

Her first step is to mix the gene鈥檚 transcription factor with its DNA in
order to bind them together. Next, she adds her scissors which intercalate all
along the DNA ladder and chop it to pieces. But the scissors cannot cut DNA
where it is bound to the protein. By comparing the fragments generated by this
procedure with those produced when the same 鈥渦nbound鈥� strand of DNA is cut up,
Barton can identify where the transcription factor attaches to DNA. It is then a
short step to read off the sequence of base pairs. Once that is known,
researchers can think about designing molecules to block them.

It is this challenge that Dervan has taken up. But it鈥檚 no easy task because
the transcription factors that bind to DNA vary greatly. 鈥淭here is no simple
motif for recognition,鈥� he says. What Dervan needed was a group of chemical
units that could be assembled in any order to bind to any sequence of DNA base
pairs.

Twisting and winding

The first breakthrough came back in 1987 when Dervan and researchers at the
French National Natural History Museum in Paris led by Claude
H茅l猫ne proved that they could use a single strand of DNA to bind a
specific sequence of the double helix. The extra strand winds round the normal
twisted ladder structure rather like cotton winds round the thread of a screw.
This 鈥渢riplex鈥� structure physically blocks access to the DNA.

Dervan and H茅l猫ne then tried to build customised, single
strands of DNA to block any sequence of base pairs. 鈥淲e want to regulate the
expression of a single gene at a time,鈥� says H茅l猫ne. By 1992, they
were making single strands capable of latching onto a target sequence of around
20 base pairs on a DNA strand of 340 000 base pairs. 鈥淲e had proved that there
was a way forward,鈥� says Dervan.

However, there are still problems with this approach. One difficulty is that
T-A and C-G base pairs do not form stable bonds with the third strand of DNA so
only certain sequences of base pairs can be bound. Another obstacle is getting
the extra strand of DNA into the cell nucleus, where it is needed. The strands
are made up of comparatively large molecules and it is not easy for them to move
through the membranes of cells.

H茅l猫ne believes these difficulties can be overcome. His group
has inserted strands up to 15 bases long into cells. Also, any sizeable chunk of
DNA will contain stretches that do not contain T-A and C-G base pairs, so he
could target triplex formation to these stretches. 鈥淪o, potentially any gene can
be recognised,鈥� he says. His group is also making single DNA strands that
include artificial bases designed to bind with all the base pairs.

鈥淗茅l猫ne and his colleagues are really moving the triplex
approach forward,鈥� says Dervan. 鈥淏ut we thought鈥攚hy put all your eggs in
one basket?鈥� So Dervan set out on a different path. He took inspiration from
antibiotics, such as distamycin, which lodge in DNA鈥檚 minor groove. The
attraction of these crescent-shaped molecules was that they home in on specific
sequences of DNA. They are made from two or three small, ring-shaped groups,
called pyrroles, which bind only to sites where several T-A or A-T base pairs
sit next to each other.

Dervan set about changing the structure of the crescents by adding slightly
different ring-shaped groups called imidazoles. 鈥淲e came across a major
surprise,鈥� he says. The new molecules slotted into the DNA groove in pairs, and
bound to G-C and C-G base pairs as well. In 1994, Dervan joined pairs of
crescents to make a shape resembling an elongated letter U鈥攈is first
hairpin. He found that these molecules could bind tightly to DNA.

The researchers have now worked out the rules for how the hairpins dock with
DNA. Each hairpin has two 鈥渁rms鈥� made from pyrrole and imidazole rings. When
different pairs of rings are sited opposite each other on the arms, they bind to
different DNA base pairs. An imidazole ring on one arm opposite a pyrrole ring
on the other binds to a G-C rung. If the position of the rings are reversed,
they bind to a C-G rung. Two pyrrole rings opposite each other bind to both T-A
and A-T rungs
(see Diagram).
Not being able to distinguish A-T from T-A
presented a thorny problem. But last week in Nature, Dervan鈥檚 group
announced the solution. A third type of ring, hydroxypyrrole, placed opposite a
pyrrole ring binds to T-A, while the opposite arrangement binds to A-T. 鈥淭his is
the holy grail for me,鈥� says Dervan. 鈥淲e now have a code that distinguishes each
of the four base pairs.鈥�

So now Dervan can design hairpins to bind any short strand of DNA. And thanks
to the small size of the molecules, high concentrations of hairpins diffuse into
cells. His team has shown that hairpins can move into cells taken from toads,
penetrate the nucleus and lock onto a target gene more strongly than the cell鈥檚
own transcription proteins. This silences the target gene, but lets others in
the cell be transcribed as normal.FIG-mg21204802.JPG

H茅l猫ne is impressed by Dervan鈥檚 results but points to
limitations. The hairpins can only be used to block short sequences of six or
seven base pairs, and this may not be long enough to block the transcription
factor binding sites of all genes. Dervan does not see this as an insurmountable
problem. He is working to create molecules that will lock onto longer
sequences.

Despite the rapid advances made in the past few years, Dervan is playing down
the immediate potential of his technique. A lot of basic biology must be done in
cultured cells before moving into living creatures, he says. But the long-term
potential is huge. Hairpins could be a powerful new tool for research. 鈥淲e
could,鈥� he says, 鈥減roduce lab animals with specific genes knocked out
temporarily.鈥� At the moment, animals and plants with 鈥渒nock out鈥� genes must be
produced through complex breeding and genetic engineering.

In addition, the molecules might make drug treatments for many diseases. They
could be used to block transcription of 鈥渙ncogenes鈥�, which cause cells to turn
cancerous. Chemists could also design hairpins to switch off genes that cause
inherited diseases.

In fact, in future, chemists could be tailor-making hairpins to dock into the
sequence of base pair rungs in any of our genes. 鈥淵ou are a product of about 80
000 genes,鈥� says Dervan. 鈥淛ust imagine the power of a set of chemical keys that
can unlock and control that information.鈥� Thanks to the chemists鈥� rapidly
growing molecular tool kit, we might soon have our hands on those keys.