The human tongue can distinguish between a dazzling array of subtle flavours
using a combination of just four elements of taste: sweet, sour, salt and
bitter. Each element is detected by one of the four types of taste bud located
on the tongue. Now researchers at the University of Texas in Austin have
designed an electronic tongue that works along the same lines (Journal of
the American Chemical Society, vol 120, p 6429).

Using chemical sensors, they have demonstrated that the electronic tongue can
“taste” different solutions. The team attached four well-known chemical sensors
to minute beads made of polyethylene glycol and polystyrene. The beads were
placed in micro-machined wells on a silicon wafer.

The sensors respond to chemical stimuli by changing colour, so the wafer was
placed between a light source and an image sensor. The responses were monitored
as red, green and blue light channels from each well. The first device they
built was designed to detect calcium and cerium ions, acidity and the simple
sugars. Each sensor responded differently to the various conditions: for
example, one turned yellow in response to high acidity, red if cerium ions were
present, and purple under basic conditions.

The sensors responded to different combinations of the four artificial taste
elements with unique combinations of red, green and blue, enabling the device to
analyse for several different chemical components simultaneously.

Eric Anslyn, a chemist at the university, hopes ultimately to make a device
“even more sensitive than the human tongue”. For now, his device has only nine
wells on a half-centimetre wafer, but he says current technology allows more
than 100 wells in the same area. The researchers plan to replace the numerous
tests carried out on a blood sample with a single test.

The electronic tongue could monitor the quality of simple solutions, says
John Warburton, a director of Neotronics Scientific, an Essex-based company that
manufactures electronic noses. Such instruments detect volatile compounds that
evaporate from a solution to give an odour. “A good application would be
something like mineral water, where you are trying to sell a product that should
have no taste or a very simple taste,” says Warburton.

He points out that some of our sensation of taste comes from our sense of
smell. The electronic nose and tongue could work together, he suggests, to
analyse compounds in solution as well as those that evaporate. “For a certain
type of material, there’s probably a great deal of synergy between the two,” he
says.

For a long time scientists have thought that the jellyfish chemical might have medical potential, because calcium ions play a vital role in many fundamental biological processes, including those involved in illnesses such as hypertension and heart disease. Coelenterazine would be ideal for detecting and monitoring them.

Until now, the chemical could only be made by extracting it from the jellyfish itself-a highly impractical task-or from a long, expensive and inefficient chemical synthesis. But a team of chemists led by Keith Jones at King’s College London has discovered a quick, efficient way of manufacturing coelenterazine that could make it readily available for research.

The main problem with synthesising the chemical was finding a way to attach a benzene ring to a pyrazine ring to form the molecule’s core. żěè¶ĚĘÓƵs had previously tried to build up this core in several steps, but Jones has developed a new process in which the benzene and pyrazine rings are made separately and joined together in a single step using the precious metal palladium as a catalyst (Chemical Communications, 1997, p 323).

This reaction, called a Suzuki coupling, allows coelenterazine to be prepared in only eight, highly efficient steps. “The entire synthesis takes just three weeks in the lab,” says Jones. The Suzuki coupling is one of several palladium-catalysed reactions that are becoming increasingly important in organic chemistry. As in the new synthesis of coelenterazine, they are used to replace several steps of a conventional process with a single reaction, significantly increasing its efficiency.

As well as tracking calcium ions, the chemical can be used to monitor cellular respiration, as it emits light when it interacts with superoxide ions, one of the products of respiration. The researchers have obtained a patent for their method of synthesising coelenterazine, and hope to use it to prepare related molecules that could have still more uses. For instance, says Jones, they could make chemicals to detect specific metals by attaching artificial metal-binding sites to coelenterazine.

“Being able to make coelenterazine in quantity will support the remarkable advances being made in the molecular biology of the cell,” says Frank McCapra, president of the Society for Bioluminescence and Chemiluminescence.

Many of the reactions that lead to drugs and agricultural chemicals involve
molecules that have left-handed and right-handed forms, only one of which leads
to the product with the desired biological activity. The mirror-image form may
even be toxic, as was the case with the drug thalidomide. One form cured morning
sickness, but the mirror image led to birth defects. Chemists have long
struggled to devise ways of selecting only the desired form.

One effective way of selecting one of the mirror-image forms is to use
microbial enzymes to catalyse the reaction, as they can make one form of a
compound react much faster than its mirror image. One example is an enzyme
produced by a bacterium of the genus Acinetobacter, which catalyses a
reaction known as the Baeyer-Villiger oxidation. This reaction, which converts
ketones to esters, is used widely for making fragrances, flavours, drugs and
other products.

But the bacterium is choosy about the conditions in which it thrives. So
chemists Jon Stewart from the University of Florida at Gainesville and Margaret
Kayser from the University of New Brunswick, St John, Canada transferred the
gene for the enzyme into baker’s yeast, which is much better suited to the
rigours of industrial conditions.

In one set of tests with the new strain of yeast, the researchers wanted to
obtain one form of a cyclic ester without its mirror-image form. The trouble is
that the two forms cannot be separated by physical means. But Stewart and Kayser
report in the Journal of Organic Chemistry (vol 61, p 7653) that the
yeast reacted almost exclusively with one form of the ketone to produce the
required ester, leaving the mirror-image form of the ketone behind (see
figure, top).
The ketone and the ester were not hard to separate. In many of the
reactions the researchers tested, the selectivity between mirror-image forms of
chemicals exceeded 98 per cent.

In a second series of experiments, the yeast was used to convert a
“non-handed” cyclic ketone to just the preferred form of the possible
mirror-image esters (see
figure, bottom).FIG-mg20573101.jpg

There is a wide variety of naturally occurring enzymes that can catalyse
useful reactions, say Stewart and Kayser. They are already experimenting with
other genetically engineered yeasts. “We see this as a way of making enzymes
available to the everyday chemist,” says Stewart.

David Bocian of the University of California at Riverside and Jonathan
Lindsey of Carnegie Mellon University in Pittsburgh are developing molecular
devices that transmit energy imparted by photons of light, rather than a flow of
electrons. At the level of single molecules, the absorption and emission of
light is much easier to detect than an electric current, Bocian explains. “Using
light to produce a transmission signal provides a much better interface between
the device and the outside world.”

In 1994, the chemists produced a photonic wire consisting of a a chain of
molecules called porphyrins. These large, ring-shaped molecules carry four
nitrogen atoms that can bind to other atoms which are then held at the centre of
the ring.

The porphyrin at one end of the wire absorbs light, entering an excited state
until it passes its excess energy on to the next porphyrin in the chain. The
energy is passed down the chain until it reaches the last porphyrin, which
re-emits it as light.

To add a switch, Bocian and Lindsey attached another porphyrin, bearing a
central magnesium atom, branching from the side of a short wire made from just
three porphyrins. When the chemists oxidised the magnesium atom, the energy was
shunted towards the magnesium porphyrin, where it was presumably dissipated as
heat, and no light emerged. If the magnesium was then chemically reduced, the
light output returned (Journal of the American Chemical Society, vol
118, p 3996).

One of the device’s most attractive features is that the switch does not sit
on the wire itself. It should be possible to add switches at many different
positions, and the chemists have already found the switch works on both the
middle or the final porphyrin in their three-porphyrin wire.

These simple units could be combined in different ways to mimic the function
of many electronic devices—eventually including computer chips. “It’s the
chemical equivalent of a set of Lego blocks,” says Lindsey. The chemists are
also trying to make a version of the switch that could be triggered by light,
rather than by a chemical oxidant.

Cellulose is nature’s pre-stressed concrete, comprising nearly 50 per cent of
the thick cell walls which help give plants their rigidity. The key to
cellulose’s strength is the way its component glucose molecules are joined.
Depending on the form of these links, glucose can make more than fifty different
polymers. Some, such as starch, readily curl or tangle up, but cellulose always
forms as a rigid, linear chain. In cellulose, the six-sided glucose rings are
tethered so that each joins to the upper face of one neighbouring ring and the
lower face of another. As a result of these ties, called (1,4)-&bgr;-glycosidic
links, the complete polymer looks like a gently sloping staircase.

The trouble for chemists is that left to their own devices glucose molecules
do not join together in this way. A glucose ring is bent into a chair-like shape
and this means that the molecules do not readily line up in the correct way to
form (1,4)-&bgr;-glycosidic links. Plants overcome this difficulty using an enzyme
called cellulase. And while cellulase can be made to work in the test tube,
chemists have been searching for a way to make cellulose that can be modified to
synthesise a range of similar molecules.

Fumiaki Nakatsubo and his colleagues at Kyoto University realised that the
key to manufacturing cellulose was to make glucose rings flat. To do this, they
added a modified ester that in the presence of another reagent called
N,N’-carbonyl-di-imidazole reacted with the hydroxyl groups carried by
three of the carbon atoms in each glucose ring, pulling the ring flat. When an
acid catalyst was added, the rings formed (1,4)-&bgr;-glycosidic links with each
other, springing back into shape and releasing the ester (Journal of the
American Chemical Society, vol 118, p 1677).

The resulting polymer was identical in structure to one of the crystalline
forms of natural cellulose. And by changing the temperature, the chemists found
they could produce chains of varying length. So far, the largest polymers they
can reliably produce contain an average of 20 glucose units, similar to some
natural celluloses.

Aside from its use by plants as a building material, cellulose has other
interesting properties. If various chemical groups are added to the basic
polymer, cellulose can prevent blood clotting, or become toxic to certain
tumours induced by viruses. Using variants of the Kyoto team’s technique,
chemists will now be able to prepare designer polymers closely related to
cellulose, and probe their possible uses. “This is an extremely elegant piece of
synthetic chemistry which could become commercially significant in producing a
new range of cellulose derivatives,” says Fraser Stoddart of the University of
Birmingham.

Many chemicals used as drugs or in agriculture exist in mirror image left and right-handed forms called enantiomers. Often, one enantiomer is more active than the other; and in some cases, such as the infamous drug thalidomide, one form can even be dangerous. So industrial chemists must devise methods, often using organic solvents, to exclude the unwanted form.

The Liverpool researchers, led by Phil Page and Don Bethell, have now come up with a catalyst that can extract one enantiomer from a mixture of the two forms of the alcohol butan-2-ol. The catalyst requires no solvent, as it can work when the alcohol is a gas, at around 115 °C.

The new catalyst is a modified zeolite – a naturally occurring crystal of aluminium silicate which has a porous, honeycomb structure. The Liverpool team doped their zeolite so that its cavities contained one enantiomer of a small organic molecule called a dithiane oxide.

Chemists understand the internal structures of zeolites very well, and so the Liverpool group could use computer simulations to work out exactly how the dithiane oxide was bound inside the honeycomb cavities. These simulations showed that when the butan-2-ol vapour passed through the catalyst, only one of the two enantiomers could fit into each cavity. Subsequent experiments showed that the enantiomer was converted into butene, while the other passed through the zeolite unchanged. Butene is a gas at room temperature, so by condensing the mixture as it leaves the catalyst, the researchers could remove the desired butan-2-ol enantiomer. The chemists found that they could obtain the other enantiomer if they doped the zeolite with a dithiane oxide to which a phenyl group had been added (Chemical Communications, 1995, 2409).

Using the same methods, the Liverpool researchers say it should be possible to devise doped zeolites to catalyse a wide range of industrially important reactions. “At the moment we can selectively destroy single enantiomers,” says Bethell. “Our goal is to selectively create them.”

For nearly a decade, scientists have known that specially designed antibodies could be made to catalyse chemical reactions. Antibodies, like enzymes, are large protein molecules with pockets on their surfaces that are exactly the right shape for other molecules to fit into. In the case of antibodies, the pockets fit a foreign molecule, or antigen, so that the antibody can bind to it. In the case of enzymes, the pockets bind chemical reactants, bringing them together so that they can interact more easily.

To make a catalytic antibody, scientists induce a mouse to produce antibodies to an antigen that has the shape and electronic charge of an intermediate in a chemical reaction. These antibodies will then behave just like enzymes, taking molecules into the pocket in the antibody surface and forcing them into just the right shape to undergo a chemical reaction.

Antibodies that catalyse simple one-step chemical reactions have been known since the late 1980s. But useful reactions often involve several steps and several different intermediates, making it much more difficult to devise a suitable catalyst.

But now a team led by Jean-Louis Reymond and Richard Lerner of the Scripps Research Institute in La Jolla, California has prepared an antibody that catalyses three of four steps in an important reaction known as intramolecular aldol condensation. The reaction is crucial to the synthesis of many pesticides and drugs, including oral contraceptives and steroids.

Reymond and Lerner prepared antibodies using a template antigen consisting of a ring of carbon atoms linked by a positively charged nitrogen atom. The antibody that resulted had a negatively charged carboxylic acid in the binding site opposite the positively charged nitrogen atom of the antigen. For two of the three catalysed steps this negative charge helps to remove a hydrogen ion from the molecule. Then the shape of the antibody’s pocket helps to form the bond that closes the ring on the molecule. The fourth step occurs spontaneously (Journal of the American Chemical Society, vol 117, p 9383).

The study represents “a major advance in a very exciting area”, says Keith Jones, an organic chemist at King’s College, London. “It opens up a wealth of new applications,” he says. (see Diagram)

The technique is an extension of an approach called combinatorial synthesis, in which chemists take a number of component molecules and join them together in all possible combinations. The resulting mix-and-match compounds can easily be screened for biological activity. The difficulty has been finding a way to label the molecules so that researchers can work out the structure of those that turn out to be promising as drugs.

To solve this problem, the chemists modified an approach called Furka-split synthesis. This method joins molecules together in a series of steps, while keeping one end of the compound being synthesised anchored to a polystyrene bead.

The researchers made their candidate drug molecules from three chemical sections, which can be thought of as “heads”, “bodies” and “tails”. They took a batch of polystyrene beads, divided it into seven portions and treated each portion with a different head unit. The result was that individual beads became coated with a large number of head units of the same kind.

Then the chemists mixed the seven portions together, redivided them at random into 31 samples and added different bodies to each. This gave them 217 head-body combinations. The 31 groups of beads were then mixed up once more, and again redivided into 31 samples so that all the headbody combinations were represented in every sample. Finally, different tails were added to each of these samples, making a grand total of 6727 compounds.

The Pharmacopeia team’s innovation is to tag the beads so that each one carries a record of the components from which the attached compound is built. Each time the chemists add a head, body or tail to a sample of beads, they also attach simple organic compounds called ketones containing carbon atom chains of varying lengths. As a result, each bead collects a combination of ketone labels whose sizes reveal the components of the molecule to which it is attached (Journal of the American Chemical Society, vol 117, p 5588).

Screening compounds made by this method is simplicity itself, the researchers say. Individual beads are simply removed, added to water and illuminated with ultraviolet light. The head units are bound to the beads by benzene rings bearing a single nitro (NO₂) group. Energy from the UV light is absorbed by the benzene ring and breaks the bond with the candidate drug molecule, releasing it to form a solution that can be tested for its biological activity.

The chemical composition of promising compounds can then be identified, the researchers say, by taking the relevant bead and exposing it to a strong oxidising agent. This releases the ketone tags, allowing them to be analysed by gas chromatography, which separates chemically similar molecules according to their size.

Five years ago, researchers working at the US Antarctic research station in McMurdo Sound found that the amphipod Hyperiella dilatata, which grows to a millimetre or so in length, regularly “kidnaps” small shell-less snails called Clione antarctica. These snails seem to be immune to attack from the plankton-feeding fish that are the amphipods’ main predators. So, to protect themselves, the amphipods often carry a snail on their backs, holding them firmly in place with two of their pairs of legs. The snails can grow to several times the size of H. dilatata, so the amphipods select only the smallest. Fish avoid amphipods carrying snails and will spit them out if they capture one by accident.

Ever since this remarkable behaviour was first observed, scientists have been working to find the chemical that makes the snails so unpalatable. Researchers led by Bill Baker of the Florida Institute of Technology extracted the chemicals produced by C. antarctica using two different solvents, which each extracted a different mixture of substances. They then added these chemical cocktails to pellets of fish food and compared the feeding patterns of fish given the doped food with those given untreated, control pellets. One of these mixtures stopped the fish feeding, indicating that it contained the defensive chemical.

The researchers purified each of the different chemicals in this mixture using a technique called high performance liquid chromatography, which separates chemicals on the basis of the differing strengths with which they bind to a silica gel. When these chemicals were tested individually only one made the fish stop feeding. This molecule, which the researchers call pteronenone, had a structure similar to that of chemicals which absorb the Sun’s damaging ultraviolet rays and are carried by other planktonic creatures (Journal of Organic Chemistry, vol 60, p 780). So it seems that the molecule which makes C. antarctica unpalatable to fish and causes the snails to get kidnapped may also stop them getting sunburnt.

A. fumigatus is harmless to healthy people, but for those whose immune systems are impaired by AIDS or immunosuppressive drugs, it can set up an infection in the lungs which is usually fatal. Yet this same fungus holds the promise of developing a new class of drugs which can control the way the body absorbs cholesterol, and so minimise the risk of heart attacks caused by a narrowing of the arteries.

Researchers led by Satoshi Omura of Kitasato University in Tokyo were looking for new drugs to block an enzyme called acyl-CoA:cholesterol acyltransferase, or ACAT. żěè¶ĚĘÓƵs believe this enzyme is involved in several processes which cause cholesterol to build up inside blood vessels, a condition called atherosclerosis. ACAT is thought to increase the absorption of dietary cholesterol from the gut and its accumulation throughout the body. So if ACAT, could be blocked, the amount of cholesterol entering the bloodstream would be reduced, preventing atherosclerosis.

The Japanese group suspected that fungi might produce molecules that could inhibit ACAT as fungi produce large numbers of biologically active chemicals. They found a promising lead in a soil sample from the Shinjuku area of Tokyo which contained a particular strain of A. fumigatus. From this fungus, the Japanese team isolated a group of previously unknown molecules, which they named the pyripyropenes. These compounds proved to be the most active naturally occurring inhibitors of ACAT ever discovered. Not only were they effective in test-tube experiments, but a single dose could reduce high blood cholesterol levels in hamsters by nearly 50 per cent.

The key step in turning this discovery into a new lead for coronary disease prevention, however, was to unravel the three-dimensional structure of the molecules. To achieve this, Omura was joined by an international team led by Amos Smith of the University of Pennsylvania, an expert in magnetic resonance spectroscopy, which can determine a molecule’s structure by analysing the behaviour of its constituent atoms in a powerful magnetic field. Using a variant of this technique, the researchers worked out the relative positions of each of the atoms within the molecule.

Biological molecules, however can exist in mirror-image forms called optical isomers. So the team then used another magnetic resonance technique which showed them which isomers were produced by the fungus (Journal of the American Chemical Society, vol 116, p 12097).

Now the precise structure of the pyripyropenes is known, chemists should be able to design similar molecules that may be even more effective at blocking ACAT and form the basis of drugs to treat atherosclerosis.