The first to be considered are communications satellites, which are useless
once their fuel runs out because they can no longer make the minor positional
adjustments鈥攃alled station keeping鈥攖hat keep them in their orbital
slots. A satellite that uses parts of its own structure as fuel would be lighter
at launch, as it would not have to carry so much onboard propellant. This would
mean that rocket-launching companies could carry larger payloads into orbit and
allow satellites to burn their own body parts to boost themselves into higher
orbits.

The research, which is funded by the US Air Force, is being managed by
Physical Sciences Incorporated, a technological consultancy based in Andover,
Massachusetts. Project leader Prakash Joshi confirms that studies are under way
to develop what he calls 鈥渄ual function structural propellant
materials鈥濃�攂ut the company declines to elaborate on the nature of the
substances under investigation.

However, New 杏吧原创 understands that the researchers are studying
novel kinds of thermoplastic composite materials which can offer a combination
of structural strength during launch and then be chemically converted or burnt
off in orbit.

At present, spacecraft like the space shuttle use materials such as aluminium
and graphite for their major body parts鈥攆or example, boosters and external
fuel tanks. But these are, in effect, 鈥渄ead weight鈥� that requires fuel to be
carried aloft. Although the shuttle鈥檚 solid rocket boosters parachute back to
Earth for retrieval and re-use, the huge external tank is simply jettisoned when
the craft reaches orbit.

Says Joshi: 鈥淭he idea is to look at new materials which you could use to
build a spacecraft and which could be consumed in orbit when their strength is
no longer needed. Most of the strength of the spacecraft is needed during launch
because of the G-forces, but once it gets up there you don鈥檛 need a massive
structure to hold the spacecraft together.鈥�

If suitable materials can be identified, Joshi鈥檚 team will develop a
prototype component to demonstrate that auto-cannibalism is viable. The next
step will be to decide how to incorporate the new materials into a
spacecraft.

鈥淚t may be that in order to take advantage of this material you will have to
design spacecraft a lot differently from the way you do now,鈥� Joshi says.
Possibilities might include burning off body parts in situ to create thrust in a
particular direction. Another method might be to use a central 鈥渞ecycling plant鈥�
with mechanical, chemical or solar systems to convert the structures into
fuel鈥攁lthough this would carry a weight penalty.

The gun looks like a pistol but works more like a shotgun, firing a number of tiny gold pellets. Genes are prepared using conventional cloning techniques, then precipitated onto beads of gold about one micrometre in diameter. These are fired into the target tissue at point-blank range by a pulse of compressed helium. By firing just a few pellets, using gas at a pressure of almost 35 atmospheres, the genes can be targeted in a tightly focused dose to the deepest regions of the tumour. Reducing the pulse pressure and increasing the number of pellets delivers a wider, shallower spread of the gene-carrying shot (see Diagram).

Wenn Sun, William Ershler and Ning-Sun Yang used the gun to attack tumours under the skin of mice. Once inside the tumour cells, the new genes stimulate them to produce molecules called cytokines. These act as 鈥渂eacons鈥�, alerting the body鈥檚 immune system to produce white blood cells that attack the tumour.

鈥淲e focused on the cytokine genes, including interleukin-2, tumour necrosis factor alpha and interferon gamma, and we saw major antitumour activity,鈥� says Sun. 鈥淲e鈥檝e even seen some animals become tumour-free. The big advantage of the gene gun is that it allows you to transfer a gene directly into a cell.鈥�

At present, gene therapists use viruses or globules of fat known as liposomes to carry genes into target cells. But these take a number of days to prepare, and their delivery can be hampered if they are intercepted by defences on the cell鈥檚 surface.

The researchers say that preparing genes for the gene gun is much faster, and the sheer force of the shot blasts the genes straight through the cell鈥檚 defences at near-supersonic speeds. The cell repairs the rupture within minutes.

The technique will need to be thoroughly tested for safety before any trials involving humans can begin. But because gold is relatively inert, the researchers are confident that it will prove to be safe. They hope to gain approval from the US Food and Drug Administration to begin clinical trials of the gun in a few months. The results of the tests on mice are published this month in the Proceedings of the National Academy of Sciences.

Karol Sikora, who works on gene therapy at the Hammersmith Hospital, London, says: 鈥淭he problem with gene therapy is how to get the foreign DNA to the right place. The gene gun is one of a number of approaches, each of which has advantages and disadvantages. At the moment no viruses target specifically to a tissue, so that鈥檚 the attraction of the gene gun.鈥� But Sikora sees gene guns only as an interim approach. 鈥淚f you could develop a virus that would target the right place it would be much less crude than this,鈥� he says.

The Wisconsin project is the latest in a long line of research into similar gene gun technologies. A device developed by Dennis McCabe of Agracetus in Middleton, Wisconsin, successfully inserted engineered genes into plants (Technology, 2 November 1991). And a team at Duke University, North Carolina, used tungsten pellets to carry the genes for human growth hormone and the luminous compound luciferase into mice (Technology, 15 September 1990).

Developed by Issam Mudawar and his colleagues at Purdue University, Indiana, the simplest prototype consists of a single microchannel 鈥� a thin tube, similar to a hypodermic needle. When the tube is held alongside a hot object it picks up heat. Water forced through the tube absorbs the heat and boils, creating thousands of tiny bubbles. The expanding bubbles are flushed out at the end of the tube, carrying heat away with them. The lost liquid is replaced via a pipe at the entrance to the tube and the cooling continues.

When a liquid is heated its temperature rises. But when boiling liquid is heated so that it vaporises, its capacity to absorb heat increases dramatically, even though the temperature may rise by just a few degrees. This is because heat, called the latent heat of vaporisation, is needed to turn the liquid into a vapour. The result is that a boiling liquid removes more heat than a straightforward liquid coolant.

The narrow diameter of the microchannel also increases the efficiency of the system because it gives a large surface area over which to transfer the heat, relative to the volume of liquid it contains.

Mudawar鈥檚 team recently used a single tube to dissipate the equivalent of 27 000 watts per square centimetre from a heated plate, the same as 270 100-watt light bulbs, focusing all their energy onto half a postage stamp. This smashed the previous record of 22 000 watts per square centimetre, which was set around 20 years ago by Soviet scientists.

In practical devices, the tubes are likely to be drilled in a row near to the surface of a solid slab. This makes them robust and easy to handle. Provided the coolant is inert, the tubes could even be replaced by grooves in the surface of the slab, allowing the coolant to come into direct contact with the surface to be cooled.

The microchannels could be open at one end or form part of a closed loop as in a fridge, (see Diagram), depending on the application and the coolant being used. 鈥淔or electronics you would use non-conducting liquids, and these can be recovered and recycled when they leave the end of the tube,鈥� says Mudawar.

He has already developed a micro channel device for aircraft electronics. At present, circuit boards are housed in boxes known as avionics modules, and the inefficiency of air cooling restricts the number of boards each module can hold. The microchannel cooling system can dissipate about 3000 watts 鈥� 10 times more heat than the latest modules produce 鈥� so more of the aircraft鈥檚 electronics can be crammed into a single, compact box.

Robert Morgan of the Naval Air Warfare Center in Indianapolis, says: 鈥淗eat dissipation is becoming the limiting factor in electronic miniaturisation. At present, our avionics boxes contain 12 to 15 boards. But if this system works, we could fit all the avionics onto one or two and this will save space and weight.鈥�

The Purdue team鈥檚 next goal is to produce the devices in a variety of shapes and sizes. Mudawar envisages small devices cooling supercomputers, CAT scanners, lasers and X-ray machines, while large arrays of heat exchangers could one day dissipate the heat from the walls of a fusion reactor.

Doctors have long suspected that oestrogen protects women from diseases of the circulatory system. After menopause, when the body鈥檚 supply of oestrogen dwindles, women become more susceptible to cardiovascular disease. Researchers have also found that women recover more quickly from strokes if they are given low doses of the hormone.

William Schnaper and colleagues at Northwestern University Medical School in Chicago have now found that the hormone performs a delicate balancing act in controlling the growth of two tissue types in blood vessels. The overall effect is to minimise narrowing of the vessels. This decreases the risk of clotting.

Other researchers have already shown that oestrogen can prevent the growth of muscle within the wall of a blood vessel. But Schnaper suspected that this was not the full story, and looked at how oestrogen affects the endothelial cells which lie inside the muscle layer. His team found that in this tissue, oestrogen has the opposite effect, causing it to grow.

The researchers experimented with endothelial tissue from the blood vessels of human placentas and mice. In both cases, oestrogen encouraged faster healing of small wounds in the endothelial layers. In the mice, it even stimulated the growth of new blood vessels (Circulation, vol 91, p 755).

If the endothelial layer is eroded, says Schnaper, muscle cells behind it proliferate and narrow the blood vessel. By encouraging endothelial cells to grow and migrate into damaged areas, while at the same time suppressing the growth of the muscle layer, oestrogen turns off the processes that cause blood vessel narrowing, Schnaper believes. He says his results also explain how oestrogen aids recovery from strokes. By encouraging the growth of new blood vessels, he suggests, the hormone diverts blood away from vessels that are blocked and damaged.

Nevertheless, oestrogen has many effects on the body, some beneficial and others not. Until the risks and benefits have been fully worked out, Schnaper says, it is too early to consider hormone replacement therapy for women who are showing the early signs of cardiovascular disease.

A recent expedition found these 鈥渂lack smokers鈥� at seven new sites in a
240-kilometre stretch of the mid-Atlantic ridge. Chris German of the Institute
of Oceanographic Sciences Deacon Laboratory in Wormley, Surrey, who led the
expedition, says it has doubled the number of known sites in only three weeks.
The new finds are spaced roughly every 30 kilometres along the ridge, at a
latitude of 36 to 38 degrees north. The seabed in this region is 3000 metres
below the surface.

German unveiled his findings at the annual science meeting of the British
Mid-Atlantic Ridge project (BRIDGE) in Oxford. He said that one of the new
black smoker sites, named Rainbow, was the biggest of its kind in the Atlantic
and one of the largest in the world. He confidently predicted that ten times
as many smoker sites could exist: 鈥淔rom what we know about the geology of
other parts of the ridge, there鈥檚 nothing to suggest that they are any
different from this new site,鈥� he told New 杏吧原创.

Previous expeditions had missed smokers because they had sampled the seabed
only every 24 kilometres, explained German. He found more sites by towing a
new instrument known as BRIDGET along the seafloor for over 240 kilometres.
BRIDGET鈥檚 chemical sensors continuously tested the water for the
characteristic sulphides and metals which the smokers spew out.

The rate at which toxic metals such as cadmium and lead pour from the
smokers is a key to assessing the impact of the much larger amount of metal
waste entering the oceans through human activity. The new finds mean that
estimates of this rate will have to be revised upwards. 鈥淎lthough these
natural fluxes are still quite small, perhaps less than 10 per cent of the
total, it鈥檚 time to think about the problem again,鈥� says German.

The discovery of the new sites also fits in with current observations on
how quickly the Earth is cooling. Hydrothermal vents are thought to be the
main route through which the Earth鈥檚 core loses heat, according to John
Edmond, Professor of Oceanography at the Massachusetts Institute of Technology
in Cambridge. Because scientists already know how much heat the core is
losing, Edmond estimates that there could be as many as 5000 vent sites in the
oceans.

The findings will also help to resolve the mystery of how animals spread
between vent sites. The seabed is sparsely inhabited at 3000 metres, and the
vent sites are oases for unusual species such as blind shrimps, echinoderms
and metal-digesting bacteria. When the only known Atlantic vent fields were
several hundred kilometres apart, yet supported similar species, scientists
wondered how the animals migrated from one to another. If sites are only 30 to
50 kilometres apart, they could simply drift between them.

Soot is part of a dangerous cocktail of airborne solids measuring less than
10 micrometres across and known as PM10. This cocktail is thought to be
responsible for up to 10 000 deaths a year in Britain (鈥淒ying from too much
dust鈥�, New 杏吧原创, 12 March 1994).

Unlike liquid pollutants, soot has no distinctive chemical signature, which
makes identifying its source and nature that much harder. Soot, mostly carbon,
does not dissolve, and cannot be analysed easily.

The fingerprinting technique was developed after Adel Sarofim and John
Vander Sande looked at soot through a transmission electron microscope (TEM)
and found that particles of soot from different sources had different
characteristics.

By magnifying the pictures from the TEM 2.5 million times, they produced
digitally enhanced images which showed the individual layers of graphite in
the soot, which are piled up to form the particles. They compared these with
data from the original images and found that different products such as soot
from diesel and from anthracite had different layer spacings. They reckon that
the structures will be unique 鈥� reflecting the varying chemical and
temperature environments that any particles will have gone through. 鈥淎t the
moment we aim to distinguish between soot from the major types of fuel such as
diesel, wood, fuel oil and gas,鈥� says Sarofim. 鈥淏ut in time we hope to say
whether a particle came from a particular type of engine or even a certain
蹿补肠迟辞谤测.鈥�

By fingerprinting hundreds of particles in a sample of air, MIT鈥檚
researchers hope to work out how much of a city鈥檚 soot comes from engines,
factories and fires. 鈥淲e need to know where it comes from before we can tackle
the problem,鈥� Sarofim says.

Soot particles are particularly dangerous because they can be very small
and often have a strong affinity for other chemicals. This means that they can
carry acids and other pollutants deep into the lungs where they can irritate
the sensitive membranes. Recent studies have even shown that soot alone can
cause tumours in the lungs of rats.

The technique can be tried on soot particles as small as 0.1 micrometres
across. The researchers hope to identify those particles which have a greater
number of regions that will trap smaller molecules.

Sarofim says this will help the environment in two ways. First, it will
enable engineers to tweak engines and other combustion processes so that they
burn more cleanly and produce a lower proportion of the smaller carrier
particles. Secondly, the technique should also identify which forms of carbon
are best at trapping airborne particles, and so make the best antipollution
filters.

Guy Beugnon and his colleagues at Paul Sabatier University in Toulouse,
France, set out to discover if ants could be trained to go to different
feeding sites at different times of the day. Earlier, other researchers
had found that some species of ants can be trained to feed at the same site
at different times.

Beugnon and his colleagues housed a colony of the tropical ant Ectatomma
ruidum in a small tank. This was connected to an arena linked through small
tubes, to three feeding dishes. Each dish offered honey for a period of
one hour at a different time o鈥檈 day.

Every day for three weeks Beugnon and his colleagues observed the foraging
ants. They found that by day 15 the ants were going to the right dish at
the right time. The researchers then marked nine foragers and watched their
reaction when no food was offered in any dish on day 22. Sure enough, the
ants headed for dishes at the time they expected them to be filled with
food (Animal Behaviour, vol 48, p 236).

Beugnon believes that the study shows that ants are capable of learning
that food is available only at a particular time in a particular place.
However, David Nash of the University of Bath says some ants may have been
following chemical trails laid down by others. For this reason, he says,
the study does not prove conclusively that individual ants can learn.

However, Nash says that if the study if confirmed it would challenge
the view that ants are merely behaving out of instinct and programming.
鈥楳ost people believed that ants have a limited repertoire of behaviours
which are switched on by social interaction rather than innate abilities,鈥�
he says. 鈥楤ut if this study is correct and individual ants can learn an
association between time and place then their nervous systems must be more
complex than we thought.鈥�

E. ruidum live in small colonies and so has fewer ants available to
forage for food such as nectar in flowers. Nash believes the ability to
learn what time of day a particular flower is producing nectar would be
an advantageous, because it would allow them to gather food more efficiently,
rather than waste energy in a random search.

鈥楾he level of ignorance about blue whales is extensive,鈥� says Russ Charif
of Cornell鈥檚 Bioacoustics Research Program. 鈥榃e don鈥檛 know how far they
swim, or how deep and often they dive, so almost everything we learn will
be new.鈥�

Last year, scientists working on the programme listened to blue whales
using the US Navy鈥檚 Integrated Undersea Sound Surveillance System (鈥楾he
Pentagon鈥檚 green flag of convenience鈥�, New 杏吧原创, 14 August 1993), but
they could track whales only when the creatures were calling within range
of one of the Navy鈥檚 listening devices.

By combining information from the tags with that obtained by the Navy,
a clearer picture of the whales鈥� behaviour will emerge. 鈥楾he Navy system
tracked one whale for 2400 kilometres over 43 days, but we don鈥檛 know if
this is typical,鈥� says Charif.

Cylindrical tags, 20 centimetres long and 5 centimetres in diameter,
were dev-eloped by Bruce Mate of Oregon State University. They will collect
details of how deep and for how long the whales dive, as well as information
about water pressure and temperature. The information is stored until the
whale surfaces, and then transmitted in less than a second to a communications
satellite, from which it is relayed back to the ground.

The sound of the whale calls is well known, and these will not be recorded
by the tags. But they will tell researchers how deep the whales are when
they call, and whether they are communicating along the sofar channel 鈥�
a region of the deep ocean where low-frequency sounds travel particularly
well.

The scientists are also keen to study whether low-frequency sounds from
other sources, such as ships and seismic disturbances, affect behaviour.
Mate hopes to use the tracking information to see if the whales are avoiding
such sources, but he admits that the problem with finding out how they react
to artificial sounds is that he and his colleagues do not yet know how the
whales behave under normal circumstances.

The research should give scientists a clearer idea of the likely impact
on the whales of the controversial plans to measure the temperature of the
oceans by creating loud explosions on the California coast.

Cancer spreads throughout the body when cells split off a primary tumour,
are swept away with the blood and form new tumours elsewhere. It is this
cleaving process, which is known as metastasis, that could be stopped by
the Velcro-like protein.

Known as integrin alpha4-beta1, the protein was discovered in 1987 by
Martin Hemler and his colleagues at Harvard University. Later, other scientists
noticed that activated alpha4-beta1 could help white blood cells stick to
each other, as well as to the surfaces of inflamed blood vessels.

鈥楢lpha4-beta1 is one of the body鈥檚 fundamental adhesive molecules,鈥�
says Irving Weissman of Stanford University鈥檚 School of Medicine. 鈥榃e have
evidence that it may attach to itself, rather than to other sites on the
outside of the cell, so it acts rather like Velcro.鈥�

Weissman and his colleague Fang Qian reasoned that tumour cells studded
with alpha4-beta1 would fasten to each other and be less likely to break
away from the tumour and spread (Cell, vol 77, p335). To test the idea,
they engineered mouse melanoma cells to produce the Velcro protein and injected
them under the skin of eight mice. A month later, they found that only
five tumours had appeared in the lungs of the mice. However, in a group
of eight mice used as controls, 45 new tumours had spread to the lungs.

Weissman and Qian now plan to study a variety of human tumours to see
how much of the protein they produce. They will then correlate their findings
with measurements of metastasis in the patients who develop the tumours.
Eventually, this may also allow doctors to predict how much a tumour will
spread, and adjust conventional treatment accordingly.

The protein can exist in a latent state on the surface of the cell,
and Weissman and Qian hope to find how it becomes activated. This could
lead to the development of a drug to activate the protein or to promote
its production. 鈥榃e still have a long way to go,鈥� says Qian. 鈥業t will be
several years before we can use this approach to prevent the spread of cancer
in humans.鈥�

Oxygen starvation, or hypoxia, results when blood vessels grow haphazardly
inside a developing cancer. Beverly Teicher of the Dana-Farber Cancer Institute
in Boston told the meeting that this lack of oxygen was hampering treatment
by protecting the tumour against conventional therapies such as radiation
and chemotherapy. Both treatments kill the malignant cells by generating
damaging free radicals from oxygen.

Teicher has used a new type of probe known as a pO₂ histograph
to map oxygen-starved regions in a tumour. Armed with this information,
she and her colleagues can increase the oxygen levels in the tumour to allow
conventional therapy to work, by infusing the patient with a synthetic blood
substitute. Trials on patients with superficial tumours such as melanomas
are promising.

Other cancer therapies have been designed to exploit the hypoxia, using
anticancer drugs designed to work in almost anaerobic conditions. British
and American researchers are using a bacterial 鈥榮ubmarine鈥� to deliver a
lethal cargo to the site of a tumour. Martin Brown and his colleagues at
Stanford University, California, use a genetically engineered bacterium
to produce an enzyme which activates an anticancer drug within the hypoxic
regions of the tumour.

The bacterium, Clostridium aceto-butylicum, is anaerobic and is normally
found in the soil. It can grow and reproduce only in low-oxygen environments;
tumours are the only oxygen-starved sites in the body, and scientists have
known for years that it can survive in tumours. It causes no human illness.

The cancer drug is CB 1954, which needs the presence of the enzyme nitroreductase
to transform it into a potent killer of cells. It is this enzyme that the
engineered form of Clostridium produces. The engineered bacterium was created
for Stanford by Nigel Minton of the Centre for Applied Microbiology and
Research at Porton Down near Salisbury.

Initial trials on mice have shown that the bacterium does indeed produce
nitroreductase only in tumours, and that it grows nowhere else in the mouse鈥檚
body. The next step is to inject the mice with the cancer drug in its benign
state, and then add the genetically altered Clostridium to see if the two-stage
delivery system really works.