Å®ÉúСÊÓƵ

Chemists in the US have used an ‘antibody’ to catalyse a reaction for
which no enzyme exists. In the body, antibodies are complex proteins that
defend the body from infection by recognising and binding very specifically
to foreign molecules. This specificity has attracted the attention of chemists,
who have wanted to use them as ‘artificial enzymes’ to speed up chemical
reactions.

Peter Schultz and Andrew Braisted of the University of California at
Berkeley have made an antibody that can catalyse a so-called Diels-Alder
reaction (Journal of the American Chemical Society, vol 112, p 1430). This
is the classic route by which organic chemists make carbon rings of the
cyclohexene type. The antibody makes the reaction more than a million times
faster.

Last year, Don Hilvert of the Research Institute of the Scripps Clinic
in La Jolla, California, was able to use an antibody to speed up a similar
chemical reaction by a greater amount than this. But Schultz’s approach
is more widely applicable.

‘This is the first documented case where anyone has been able to do
something with antibodies that enzymes haven’t been shown to do,’ says Schultz.
‘We have done something in nine months that Mother Nature hasn’t managed
in a hundred million years.’

Enzymes work by lowering the energy needed to reach the transition state
– the unstable intermediate stage – through which a chemical reaction passes.
Chemists use a trick to mimic this. They make an antibody that will bind
a stable ‘analogue’ molecule (in this case bicyclooctene) that closely resembles
this transition state. The antibody will then catalyse the chemical reaction.

Schultz thinks that such a catalyst would be useful for making intermediates
for drugs, and in organic syntheses, which chemists would then be able to
carry out at lower temperatures. More importantly, they may be able to control
the selectivity. It is this, rather than speed, that may be the real advantage
to chemists of the new catalysts.

Now Schultz and his colleagues have a related antibody that forms a
complex with an iron porphyrin. This gives them ‘a foot in the door of oxidation
chemistry’, he says. In nature, it is such molecules – known as co-factors
– that often do the chemical work for enzymes.

Schultz and his team used the antibody-porphyrin complex to catalyse
several typical reactions involving the reduction of hydrogen peroxide.
In nature, these reactions are catalysed by the enzyme peroxidase. In common
with similar enzyme-catalysed reactions, the antibody would not work with
its cofactor, and vice versa.

‘The antibody is slow compared with the enzyme – but this enzyme is
outrageously fast,’ says Schultz. ‘The rates we are getting now with the
antibody compare well with chemical models of the reaction. The real question
is whether we can get selectivity.’

An antibody that binds both cofactor and substrate would allow the chemists
to transfer selected oxygen atoms. They are now making antibodies to an
N-steroidal porphyrin in the hope of catalysing the selective hydroxylation
of steroids. This would give chemists selectivity in many organic reactions.
Stereospecific hydroxylation is an important step in the synthesis of organic
molecules, fine chemicals and pharmaceuticals.

Equally important is the work of a team led by Richard Lerner of the
Scripps Clinic and Stephen Benkovic of the State University of Pennsylvania.
They made the first step towards a new class of catalytic antibodies – metalloantibodies
– when they succeeded in inserting zinc atoms into an antibody (Science,
vol 249, p 659; Proceedings of the National Academy of Sciences, vol 87,
p 6654).

‘A large number of chemical reactions cannot be done by amino acids
alone,’ says Lerner, ‘and nature solves the problem by throwing a metal
at it – a third of all enzymes are metalloenzymes.’ In nature these metals
often act as electron ‘sinks’. Many useful reactions, including hydrolysis,
are catalysed by enzymes that contain zinc.

The key to Lerner’s technique is that all antibodies have a common feature
in their binding sites – the relationship between two loops of amino acids
called the L1 and L3 chains. By changing the exact sequence of amino acids
in these loops, the researchers were able to insert the metal between the
two. Now the researchers will try to design a metalloantibody with catalytic
behaviour.

According to Lerner, metalloantibodies will expand the field of catalytic
antibodies to do all the things that metals now do in biology, such as catalyse
the hydrolysis of amide and ester bonds and redox reactions. The advantage
of metalloantibodies is that they allow precise control of the position
of the metal. This means that chemists will be able to program an antibody
to recognise a particular substrate.

Lerner has no qualms about trying to improve on nature: ‘There are only
two to four thousand enzymes – enough to keep you alive, but not enough
to do the kind of things people want to do,’ he says, ‘such as the hydroxylation
of steroids, or hydrolysis of AIDS and TPA (tissue plasminogen activator,
a clot-dissolving protein).’

The last two reactions involve what many researchers and pharmaceutical
companies see as the most desirable aim for designers of catalytic antibodies:
cleavage of specific peptide bonds. No one has done this yet. But neither
has nature – no enzyme has evolved that can break the bonds that hold together
coat proteins of the AIDS virus, for example.