“We’re trying to predict what will happen with drugs in people,” says Mike Tarbit, head of drug metabolism at Glaxo’s research and development centre at Ware in Hertfordshire. Some potential drugs look promising but may not be able to withstand the body’s natural defences long enough to do much good before they are broken down. Some of the compounds that result from their breakdown can also lead to unpredictable side effects.

The liver is the main organ involved in processing and getting rid of drugs. Drugs that are taken orally are carried in the bloodstream straight to the liver where, in some cases, as much as 90 per cent of the dose is immediately broken down, according to Barry Ross, the Glaxo group’s research director. Even if drugs are injected, they only have around three minutes to circulate before they reach the liver. But more of an injected dose tends to survive its encounter with the liver because the drug is more diluted by the time it gets there.

The main culprits are a family of enzymes produced in the liver called cytochrome P-450s. Their main function is to act as catalysts, promoting chemical reactions that add hydroxyl groups to vulnerable parts of foreign molecules, including drugs. Some of the enzymes also tackle the invading molecules by cutting them into fragments. The resulting molecules are usually more soluble in water than fat and so can be easily excreted, either via the bile duct or the kidneys.

People typically have a dozen different P-450s in their livers, and Tarbit’s team has so far successfully transferred the genes for manufacturing nine of them – three into yeast cells, Saccharomyces cerevisiae, and six into Escherichia coli bacteria.

The project was complicated by the need to limit the production of the enzymes. If the microorganisms produce too much, the enzymes clot to form useless lumps of protein. The scientists also had to include the genetic code for another enzyme, called reductase, which helps the P-450s to work. Cultures of the genetically engineered bacteria and yeast are fed on samples of potential new drugs. Their waste products are then tested to determine what molecules the enzymes have produced and how fast they work.

The information allows Glaxo’s chemists to tweak their drugs to make them less susceptible to enzyme attack. And because it gives pharmacologists an extra test for toxicity and other possible side effects, the researchers hope that it may also allow them to cut down on the amount of animal testing needed later on in the development of a drug.

In countries such as Mexico, Azerbaijan and the Philippines, repairing earthquake damage often means strengthening columns made of reinforced concrete – the principal building material in many developing countries. The two most common techniques for repairing damaged pillars are injecting epoxy glue into the cracks, and jacketing the whole column with a casing of reinforced concrete.

Amr Elnashai of Imperial College specialises in earthquake engineering. He says that squirting glue into the columns is not very satisfactory because the results are unpredictable. But more importantly, Elnashai’s research has shown that jacketing with reinforced concrete could leave buildings in worse shape for facing another earthquake.

Elnashai and his colleague Nicholas Ambraseys ran laboratory tests on concrete walls, subjecting them to shocks from a computer-controlled hydraulic jack. Data from 30 strain gauges and other monitors were then fed into a computer model of a five-storey building from Erzincan, Turkey, along with records of a real earthquake in the area. It showed that if the building’s ground floor had been repaired with a concrete jacket, it would have collapsed at the first-floor level in a subsequent quake.

The reason, says Elnashai, is stiffness. Jacketing is designed to increase the strength of a load-bearing structure so that it can support the weight of the storeys above. This may be fine under normal conditions, but during earthquakes structures are put under unusual horizontal stresses. Stiffness (how hard it is to bend an object) and ductility (how far the object will bend before it snaps) become critical. Engineers usually repair only part of a building, but when one part of a structure is altered, the response of the whole building is modified.

While putting a concrete jacket around a pillar can restore its strength, it invariably makes it stiffer. The minimum practical thickness of a jacket is 10 centimetres, to allow room inside for reinforcing steel bars. But even a 5-centimetre jacket will increase stiffness by 200 per cent.

Stiffer columns resonate to higher-frequency shock waves – the kind experienced close to the epicentre of a tremor. “It’s like tightening a violin string,” says Elnashai. If the seismic waves are in tune with the repaired pillar it will resonate and amplify the signal, causing much greater damage to other parts of the building. On the other hand, leaving a pillar in its damaged state would make it more vulnerable to low-frequency waves. But these are typically recorded far from the focal point of an earthquake, where the strength of the tremor has dissipated.

The researchers have devised several new techniques. By bolting plates of a precise size to the side of a damaged column, for example, builders could control exactly how much they improve its stiffness. A system of steel rods that take up stress as a column bends can make it stronger without affecting the other two characteristics. And U-shaped steel bands fitted around the column can improve ductility.

The techniques have been used by a Turkish engineering company, Seyas of Istanbul. British companies specialising in earthquake repairs say they are aware of Elnashai’s research, but are waiting for his final results before testing his ideas.

The crop pests are running riot largely because they had a bad start
to the year. Few hatched in May and June, leaving their main predators little
to eat. By summer, with few predators around, the pest populations soared
along with the temperatures.

That pattern is familiar to Wilf Powell, an entomologist studying aphids
at the Rothamsted Experimental Station in Harpenden, Hertfordshire. Farming
has a similar effect. The tiny green aphids desert fields after the crops
have been harvested, and so do their predators. But in the spring, the aphids
return and gain a foothold before their enemies, which never quite catch
up.

Powell is trying to give one group of predators a head start by creating
wintering grounds for them around the edges of fields. He has discovered
that the parasitic Braconid wasps are attracted by two pheromones, nepetalactone
and nepetal-actol, given off by aphids when they mate in autumn. He is luring
wasps with synthetic pheromones derived from catmint, and supplying them
with aphids in which they lay their eggs. Powell hopes the next generation
of wasps will be lying in wait for the aphid swarm when it returns in the
spring.

The orange wheat blossom midge, Sitodiplosis mosselani, is the worst
crop pest this year, but the midge’s life cycle is not as well understood
as the aphid’s. The last time the midges were extensively studied in Britain
was during a relatively minor outbreak in the 1970s. The numbers have been
climbing steadily since 1991 when the population of their main predator,
another group of parasitic wasps, the Cynepid wasps, crashed.

Because so little is known about them, farmers are forced to use a broad-spectrum
organophosphate pesticide to kill them. But the pesticide also kills beneficial
insects, including the pests’ enemies. John Oakley at the Agricultural
Development and Advice Service in Reading is studying the midges in the
hope of finding more specific chemicals. He is testing a group of insect
nerve poisons called pyrethroids to determine at what stage of the life
cycle they are most effective.

Up to now, NASA has trained astronauts to work in zero gravity by dangling
them on springs from cranes or dunking them into water-filled tanks. Both
have drawbacks. Water gives buoyancy similar to weightlessness, but resists
movement. The cranes reduce the sense of gravity, but are awkward to use.

With just one moving part, the virtual reality system uses a completely
different approach. Trainees recline in contoured chairs, like those featured
in the film Lawnmower Man. The seats mimic the relaxed pose that people
adopt naturally when they sleep in space, so that on the ground there is
no feedback from muscles tensed against the force of gravity. A virtual
reality headset means the astronauts ‘see’ the inside of a space capsule.

To make the simulation as realistic as possible, NASA did not want to
use hydraulic supports, because their pistons are rigid and do not flex
when the astronaut’s weight shifts. Hydraulics are also slow and inefficient,
wasting up to 97 per cent of the energy used to pressurise them. Pinprick
leaks and dust in the hydraulic lines also require costly maintenance.

The solution was the Pemram motion base, invented by Phillip Denne,
who runs his own company based in Wimborne, Dorset. Instead of hydraulics
the base has gas springs and a linear induction motor which has just one
moving part. Three cylinders, each 360 millimetres long, are filled with
air at twice atmospheric pressure to support the chair. Each tube has 18
electromagnetic coils, and a piston made of neodymium-iron-boron -a strongly
magnetic material used to fish broken drilling bits from the bottom of oil
wells – shaped into a disc 25 millimetres thick. Steel caps over the poles
of the magnet concentrate its magnetic field towards the sides of the cylinder,
to maximise interaction with any current flowing in the coils.

The coils are controlled by microchips, which turn them on to either
pull or push the piston along the cylinder, according to the direction of
the current. This creates accelerations of up to 0.5 g, more than enough
to give the astronaut the sensation of movement. Response times are much
faster than with hydraulics because there are no valves to open, and energy
efficiency could theoretically reach 80 per cent.

The device is so accurate and efficient it could replace hydraulic pistons
in a wide range of industrial and commercial applications. Denne is now
investigating other uses for his invention. The device could move lifts
and replace suspension systems in vehicles. The food industry is also interested
because a leaking hydraulic line can contaminate products.

Kangaroos first descended from the trees 50 million years ago when
Australia’s climate began to dry out. They lost their anatomical adaptations
for climbing trees, such as opposable thumbs, and gained new ones to suit
life on the plains, such as a fore-stomach to help them digest grass. One
genus, Dendrolagus, returned to the rainforest some 5 million years ago,
branching into the nine species of tree kangaroo known today.

Tim Flannery of the Australian Museum in Sydney, who led last month’s
expedition to Irian Jaya, the Indonesian part of the island of New Guinea,
thinks the bondegezou is a tenth species of tree kangaroo because it has
broad, flat feet suited to tree climbing, legs of equal length and a skull
that more closely resembles those of the Dendrolagus species than ground-dwelling
kangaroos. But some of its features suggest that the kangaroo is in the
process of moving down from the stunted trees that grow in its alpine habitat.
The lower leg bones in the specimens he examined are close together, limiting
the animal’s ability to grasp tree trunks. The bones are lighter than expected
for a tree-living species, which have to bear the stress of jumping or falling
out of trees. And it lacks the long tails of those animals that balance
on branches.

Flannery and colleagues from Indonesia’s Bogor Museum have so far seen
only dead specimens of the bondegezou, including one killed by hounds as
they approached with a hunting party from the Dani tribe. But reports from
the hunters, and the neighbouring Moni people, who revere the animal as
an ancestor and will not eat it, indicate that the bondegezou is a poor
climber. It descends trees tail first and spends most of its time on the
ground.

The researchers hope to mount a second expedition to learn more about
the animal. The bondegezou weighs around 15 kilograms and is 1.2 metres
tall. Flannery estimates that thousands live in the mountains, but he is
concerned that their friendliness makes them easy prey for hunters.