NINE children who were infected with HIV at birth have apparently shaken
off the infection. Researchers in Europe who have been following the fortunes
of children born to women infected with HIV have identified a tiny proportion
of infected babies who seem either to have rid themselves of the virus, or who
can live with it without becoming ill.

About 15 per cent of babies born to HIV-positive mothers become infected
while they are still in the womb or during delivery. But over the past few
years there have been sporadic reports of infected babies who have
subsequently stopped producing antibodies against HIV and remained well.
Marie-Louise Newell of the Institute of Child Health in London and colleagues
across Europe, who set up the European Collaborative Study to track down such
children, report that they have identified nine such cases.

鈥淭here are many people in the AIDS field who don鈥檛 believe this phenomenon
exists, and who insist that previous cases can be explained by errors made in
the laboratory,鈥� says Newell. By joining forces with other European centres,
Newell and her colleagues have been able to find and systematically follow up
a number of infected children who seem to have cleared the virus. 鈥淎lthough it
is rare, with only 2.7 per cent of babies clearing the virus, it appears to be
real,鈥� says Newell.

Because HIV antibodies can cross the placenta, it is common to find
antibodies in the blood of newborn babies whose mothers are infected. These
maternal antibodies disappear within a month. The babies Newell and her team
are following are those who carry on making their own antibodies, and so are
assumed to be infected by the virus, yet eventually test negative for the
antibodies. 鈥淭his suggests that they are somehow dealing with the infection,鈥�
says Newell. 鈥淚f we could find out how they do it, then we might find some
valuable clues to help in the design of vaccines,鈥� she says.

All nine of the babies who lost their antibodies were also tested for the
presence of the virus. They all tested positive at least once. 鈥淲e used either
the polymerase chain reaction (PCR) to detect the presence of viral DNA, or
cultured the virus,鈥� says Newell. Seven of the babies went on to test negative
for the virus, which suggests they have cleared the virus from their bodies,
the researchers say. The other two babies continue intermittently to give a
positive test for the virus but remain well. The team believes that this
indicates some sort of tolerance to the virus. A full report of the study
appears in the current issue of The Lancet.

In a highly publicised case last year, doctors in California claimed they
had identified a baby that had shaken off the virus (This Week, 8 April 1995).
However, they came under fire from some AIDS experts, including Myra McClure
from St Mary鈥檚 Hospital, London, and Richard Tedder at University College
London Medical School. They suggested that the most likely explanation was
contamination of the original blood sample or errors in labelling the samples.
The case was reviewed by international experts at a workshop in Oxford last
October.

鈥淲e reached a consensus at the meeting that much stricter diagnostic
criteria must be adhered to before cases like these can be considered genuine,
and not the result of error,鈥� says McClure. 鈥淭here should be at least two
different positive tests 鈥� one confirming the other 鈥� on two independent
occasions,鈥� she says. 鈥淭o date we have found no cases of viral clearance in
the UK using these criteria.鈥�

Tedder says he regards the new data with extreme caution, because the study
did not stick to the recommended criteria. He also points out that the PCR and
viral culture techniques used by Newell鈥檚 team are 鈥渙ld and prone to giving
false positive results鈥�.

鈥淚 believe that HIV clearance is biologically plausible, but the scientific
credibility of the data offered in support of the phenomenon must be
questioned,鈥� says Tedder. 鈥淚f you are trying to demonstrate something that
goes against the dogma, then you need to shoot three arrows in the same
direction, not just one.鈥�

Newell believes these criticisms are short-sighted. 鈥淚 accept that in an
ideal world these measures would be possible, but they are not very practical
in the real world,鈥� she says. 鈥淲e鈥檝e used computer simulation to show that
these nine cases cannot all be attributed to lab error,鈥� she says. 鈥淧eople are
reluctant to believe things that don鈥檛 fit neatly into life, but it鈥檚 about
time they accepted that there is something genuine going on here, and start
investigating how it鈥檚 happening.鈥�

Andrew McMichael at the Institute of Molecular Medicine in Oxford accepts
the possibility that the babies might have fought off a brief infection. He
and his colleagues have been looking at other groups of people, including
prostitutes in the Gambia, who have been exposed to the virus but have not
been infected. Antibodies alone are incapable of getting rid of HIV, so the
researchers focused on the branch of the immune system which responds to
foreign invaders by producing killer T cells. They found high levels of these
cells in the prostitutes. McMichael suggests that the babies might be fighting
off the virus in a similar way and that in future the killer T cell response
should be tested in such babies.

鈥淚f more cases keep cropping up there could be something in it and we
should be making a huge fuss about it,鈥� says McMichael. 鈥淏ut nothing will be
proved until the criteria are fulfilled,鈥� he says.

Dentists often use broad-spectrum antibiotics to treat gum disease and tooth decay. But long-term treatment is the perfect way to encourage antibiotic resistance. 鈥淚t鈥檚 asking for trouble,鈥� says Michael Wilson, a microbiologist at London鈥檚 Eastman Dental Institute.

Wilson and his colleagues have been testing a number of light-sensitive compounds which do not effect bacteria on their own, but kill the bacteria immediately when activated for just a few seconds by a low-powered laser. They decided to try the technique on bacteria after seeing it kill certain types of cancer cells.

鈥淭hese lasers are red, so they鈥檙e best absorbed by blue compounds,鈥� says Wilson. Under laser light the compounds produce singlet oxygen (an excited state of molecular oxygen) and toxic free radicals, which kill bacteria by punching holes in the cell membranes. Because bacteria are more susceptible to the energy of lasers than mammalian cells, it is possible to kill the bacteria and leave human tissue intact.

鈥淲e鈥檝e shown that this method works on two totally different sorts of organisms,鈥� says Wilson, who tried the technique on aerobic bacteria, which require the presence of oxygen, and anaerobic bacteria, which do not.

Although he has not yet tested the technique in people, Wilson believes that dental problems will be a suitable target. 鈥淲e could easily apply the compounds to the gap between tooth and gum in gum disease, or squirt it into the hole of a decayed tooth,鈥� he says. 鈥淭he laser could be pointed directly at the compound once it鈥檚 in place.鈥�

Wilson has also demonstrated that S. aureus can be killed in this way. 鈥淭his is the bacterium that commonly infects wounds and burns and can go on to cause fatal septicaemia,鈥� says Wilson. It is also becoming resistant to antibiotics.

So far however, only localised infections can be treated because it would be unsafe to inject a light-sensitive compound into the bloodstream or give it by mouth, and then irradiate the whole body. 鈥淒oing that would probably kill off all the other useful bacteria in the body,鈥� says Wilson.

But US researchers led by Martin Yarmush at Shreiners Burns Institute in Boston think they have found a solution to this problem. By attaching antibodies which target specific bacteria to the light-sensitive compounds only selected bacteria are killed, Yarmush says. He hopes to start human trials within two years.

At a meeting of the British Neuropathological Society, British researchers described the case of a 23-year-old boxer who died in the ring from a sudden and massive brain haemorrhage. They found that his brain had some of the structural abnormalities seen in people suffering from Alzheimer鈥檚 disease and in many old ex-boxers. They believe this is the first report of this type of brain damage in a young person.

The young boxer鈥檚 death provided a rare opportunity to study the earliest changes in the brain that might be brought on by repeated blows to the head. 鈥淢ost boxers who die in the ring are examined by forensic specialists who identify the immediate cause of death 鈥� usually a large bleed into the brain,鈥� says Jennian Geddes, the neuropathologist at the Royal London Hospital who presented the case to the meeting. They do not tend to look in detail at the rest of the brain, she says.

鈥淲e found signs of long-standing brain damage and because there was no evidence of any other disease when he was alive, we have to assume it must have been caused by repeated blows to the head,鈥� she says.

Elderly ex-boxers often suffer from slurred speech, disturbed balance and poor memory 鈥� a condition called the punch-drunk syndrome. At postmortem, their brains look very similar to those of people who have had Alzheimer鈥檚. 鈥淭ypically, their brains reveal a pattern of amyloid plaques and neurofibrillary tangles that looks virtually identical to what we see in Alzheimer鈥檚,鈥� says Geddes. Tangles are abnormal clumps of a protein known as tau, which accumulate inside cells. Plaques are lumps of amyloid protein outside cells.

The young boxer in this case had been fighting since he was 11, and as a professional for four years. Apart from being 鈥渟omewhat forgetful鈥�, there was nothing in his behaviour to suggest that he might have suffered any damage to his brain.

So Geddes was surprised when she discovered there were tangles in his brain. More puzzling still, the tangles were not in the places where she would expect to see them in older boxers 鈥� and nor were there any plaques. 鈥淭he tangles were wrapped around blood vessels at the base and sides of the brain, which is exactly where the brain would sustain the force of any blows to the head,鈥� explains Geddes. 鈥淭his picture is not typical of what we find in Alzheimer鈥檚 disease or retired boxers鈥� brains.鈥� Geddes thinks she may have found the earliest stage in a process that eventually leads to the extensive damage seen in punch-drunk boxers.

She also believes that in boxers a different pathological process is at work from that in Alzheimer鈥檚 patients. Repeated blows to the skull could damage nearby blood vessels. 鈥淪omething may be leaking from the damaged vessels which causes tangles to form,鈥� she says. Another possibility is that as the blood vessel is torn away from the surrounding brain tissue by the force of the blow, it causes a chemical reaction that triggers the development of tangles.

And in this case the tangles must have formed quickly rather than over the many decades it is believed to take for Alzheimer鈥檚 disease to develop. A detailed report of Geddes鈥檚 findings will appear in the February issue of Neuropathology and Applied Neurobiology.

Geddes believes her finding reinforces the argument against allowing children to box. 鈥淲e simply don鈥檛 know how quickly it takes for tangles to form,鈥� says Geddes. 鈥淎nd young boxers may show no outward signs of brain damage.鈥� This level of damage would not show on a brain scan, the usual method of monitoring professional boxers.

Last month, Geddes re-examined the brain of another young boxer who died eight years ago. This time, using new techniques for staining the tissue, she found exactly the same abnormalities as she had in the more recent case. This reinforces her belief that the damage is 鈥済enuinely attributable to boxing鈥�.

The British Boxing Board of Control has repeatedly argued that there is no scientific evidence that professional boxers who fight by the BBBC rules suffer any cumulative brain damage. 鈥淚 know there is research looking into the possibility of brain damage in young boxers, but I鈥檓 also aware that there is some scepticism among doctors about how it鈥檚 being carried out,鈥� says Simon Block, assistant general secretary of the BBBC. The board believes that restrictions on the number and length of rounds in a bout have made punch-drunk syndrome 鈥渁 thing of the past鈥�.

鈥淭hat鈥檚 nonsense,鈥� says Peter Harvey, a neurologist at the Royal Free Hospital in London and a fervent campaigner against boxing. 鈥淥ne can only say that brain damage has occurred once it鈥檚 there,鈥� he says. Harvey argues that it is impossible to say that boxing is safe below a certain threshold.

While it may be possible to stop brain damage getting any worse once it becomes apparent, these two cases show that the BBBC鈥檚 existing rules cannot prevent it altogether. 鈥淭his will blow the top off boxing,鈥� he says.

Sherre Florence and Jon Kaas at Vanderbilt University in Nashville, Tennessee, believe they have discovered the origin of these 鈥減hantom limb鈥� phenomena. After amputation, they say, new nerve connections which transmit the sensation of touch form in a tiny part of the cuneate nucleus, a structure within the brainstem, at the top of the spinal cord. The researchers now hope to find ways of preventing these unwanted connections.

Neuroscientists already know that the concentration of substances called 鈥済rowth associated proteins鈥� tends to rise in the nerves of injured limbs. They have also noticed an increase of growth proteins where these nerves connect into the brain. This suggests that new brain connections are being created.

Florence studied adult monkeys which had lost a hand. She linked a tracer molecule to a harmless chemical called horseradish peroxidase (HRP), which she injected into the monkeys鈥� skin just above their stumps. The HRP and its associated tracer were taken up by nerve cells and carried up to the brainstem. The tracer, which appears dark blue under the microscope, allowed Florence to identify where the HRP ended up.

As expected, some of the blue stain ended up in the region of the cuneate nucleus that usually connects with nerves from the lower arm. 鈥淏ut we were surprised to find these monkeys also had a small focus of blue in the part of the cuneate nucleus where nerves from the hand would normally have ended up,鈥� says Florence. Since most of the nerves from the severed hand would have died shortly after amputation, Florence believes the HRP had picked out new connections from nerves in the arm (The Journal of Neuroscience, vol 15, p 8083). 鈥淭hese new connections could explain why people still feel sensation from an amputated hand,鈥� suggests Florence.

Now that these connections have been found, says Florence, it should eventually be possible to find ways to prevent such unwanted brain reorganisation. 鈥淭he first step is to look for molecules such as growth associated proteins which allow such new growth to take place,鈥� she says. It may also be possible to promote new connections that could actually improve nerve recovery after injury.

Louis Tartaglia and his colleagues at Millennium Pharmaceuticals in Cambridge, Massachusetts, are beginning to provide the answers. They have identified a receptor on the surface of cells that binds to leptin, which in turn should help other scientists understand how leptin regulates body weight and aid the search for drugs to treat obesity.

Leptin is produced by a gene called ob, and is thought to suppress appetite and increase metabolic rate (This Week, 5 August 1995). To identify the leptin receptor, Tartaglia鈥檚 team attached molecules of leptin to another protein called alkaline phosphatase, which catalyses a reaction that produces a purple product. The researchers added the modified leptin to a variety of mouse tissues. After unbound leptin was washed away, the deepest purple stain appeared in a section of the brain called the choroid plexus. This tissue lines the brain鈥檚 fluid-filled cavities, or ventricles and secretes the cerebrospinal fluid that bathes the entire central nervous system.

Next, Tartaglia and his colleagues took choroid plexus cells and isolated their mRNA, the molecules which are the intermediate step between genes and the proteins they code for. Using an enzyme called reverse transcriptase, they reversed this process of MRNA production to identify the corresponding genes. 鈥淲e then had a library of millions of samples of DNA,鈥� says Tartaglia. By injecting the samples in batches into cultured cells that originally lacked the leptin receptor, the researchers were then able to use their modified leptin to identify cells which had started to produce the receptor.

From these cells, the researchers were able to pinpoint the receptor gene, which resides on the mouse chromosome 4. It codes for a receptor protein that straddles a cell鈥檚 outer membrane (Cell, vol 83, p 1263). Tartaglia suspects that there may be other leptin receptors still waiting to be discovered. 鈥淏ut if we can demonstrate that leptin is useless at regulating body weight without our receptor, then we will know we have cloned the essential one.鈥�

Once leptin binds to the portion of the receptor that protrudes from a cell鈥檚 surface, other proteins presumably transmit a chemical message to the cell鈥檚 interior. 鈥淥besity may be caused by a mutation in the receptor gene, or it may be due to defective processing inside the cell which occurs after the protein binds to the receptor,鈥� says Tartaglia.

One big puzzle remains, however. Obese strains of mice typically produce little leptin, which is why scientists injected them with it in the first place. But obese humans seem to have high levels of leptin. Tartaglia believes that the underlying cause of severe human obesity may be similar to that behind adult-onset diabetes, where patients become unable to respond to insulin, even though it is present in large quantities. 鈥淸Obese patients] appear to be resistant to leptin,鈥� he says. 鈥淚f the resistance is only partial, we may be able to overcome the problem by designing drugs which can overstimulate the receptor.鈥�

Screening for drugs that target the receptor can begin straight away. But if the problem lies in the subsequent signalling pathway, researchers will first have to identify the signalling proteins inside the cell, and then target drugs against them. 鈥淚dentifying the receptor is an important further step in understanding how the ob gene regulates body weight,鈥� says Colin Dayan, an endocrinologist at the University of Bristol. 鈥淏ut it doesn鈥檛 immediately link all the elements together.鈥�

Val Kitchen and her colleagues at St Mary鈥檚 Hospital in London are testing a series of potential vaginal virucides, compounds which they hope will prevent the transmission of HIV.

Eventually, virucides could offer women greater control over their sexual health. They may be marketed in the form of gels, creams or foams. Another advantage is that virucides could prevent HIV infection while still allowing conception. This could be particularly welcome in countries such as Uganda where women are culturally under pressure to have children, even at the expense of their health.

In the 1980s, scientists identified a number of compounds which in laboratory tests appear to be able to stop HIV invading susceptible cells. One of these is dextrin sulphate (D₂S), which seems to stop HIV entering and infecting target cells by coating them. Other compounds, including nonoxynol-9 (N-9), which is the active ingredient in most spermicides, appear to kill the virus by destroying its fatty outer coat. But healthy cells also have fatty coats which make them vulnerable to the chemical, so any benefits from killing the virus must be weighed carefully against the damage done to other cells.

From the start researchers realised the potential of these chemicals, but early work on vaginal virucides quickly ran into problems. In 1992, Joan Kreiss and her team from the University of Washington, Seattle, discovered that high doses of N-9 led to genital ulceration. Any damage to the lining of the vagina could increase the chances of HIV entering the white blood cells called CD4 cells which the virus invades.

Kitchen hopes to avoid this problem by using lower doses of N-9. The ideal virucide could be a combination of two substances: one that kills the virus and another that blocks susceptible cells and prevents N-9 attacking them. 鈥淭his would be one way of reducing the dose,鈥� suggests Kitchen.

The team at St Mary鈥檚 has completed a safety trial of D₂S. The researchers asked 36 women who were not sexually active to apply either D₂S gel, or a dummy gel, on five successive nights. The women reported no significant side effects, and biopsies revealed no damage to the lining of the vagina. The team will now test a series of other compounds in the same way.

Kitchen then plans to carry out larger trials with the most promising virucides. These will probably involve about 200 women in steady relationships who are unlikely to be exposed to HIV 鈥淲e can鈥檛 assume that D₂S protects against HIV, so we need to first try it out in a low-risk population,鈥� she says. The aim of this study will be to see how acceptable virucides are to both sexes in terms of comfort and convenience. But the ultimate test will be to study virucides in countries where the risk of HIV is high.

鈥淭hree years ago there was terrific resistance to the idea of virucides, particularly from industry,鈥� says Alan Stone, head of the Medical Research Council鈥檚 AIDS secretariat. 鈥淩umour had it that they were wary of low profitability and indemnification 鈥� with the possibility of being sued if their product failed or caused ill effects.鈥� He is now talking to European pharmaceuticals companies to find out whether these objections still hold.

鈥淲e want to widen people鈥檚 choices,鈥� says Stone. 鈥淚 don鈥檛 see virucides taking the place of condoms, nor will this line of work undermine vaccine research,鈥� he says. 鈥淏ut virucides should be effective enough to make an impact on HIV transmission, and are likely to be available sooner.鈥�

The placentas of pregnant African women, particularly those who are carrying their first child, are often clogged with malaria parasites. These parasites can cause anaemia, which in turn can lead to premature labour and undersized babies.

The researchers realised that the parasites which infect pregnant women are for some reason not recognised by the immune system. 鈥淚t鈥檚 either a different parasite altogether, or it鈥檚 the same parasite binding to a totally different receptor in the placenta,鈥� says Fried.

Malaria-infected red blood cells stick to the walls of blood vessels by binding to receptors called CD36 and ICAM-1. Fried and her colleagues screened a range of placental proteins to find the corresponding receptor in the placenta, and found that infected cells bound fast to a protein called chondroitin A. When they added these protein-bound cells to a sample of disease-free placenta, none could anchor to the tissue. 鈥淭his showed us that chondroitin A is the main protein receptor which malaria parasites bind to in the placenta,鈥� says Fried. Untreated infected cells were also unable to adhere to placental tissue that had been exposed to an enzyme that digests chondroitin A. Fried described these experiments last week in Washington DC, at the annual meeting of the American Society for Cell Biology.

The researchers still do not know if the parasites that clog the placenta are different from those found elsewhere, but they hope to find the answer by comparing the genetic characteristics of parasites taken from different sites. Fried also wants to identify the parasite protein that binds to chondroitin A. 鈥淲e might be able to find a good candidate for a vaccine, or design a drug which could interfere with the binding process,鈥� she says.

This may not be easy, however. 鈥淭o be able to interfere with the binding of the parasite to the host, you have to assume that both the parasite binding site and the host tissue receptor are constant, which may not be the case,鈥� warns Tony Holder of the National Institute for Medical Research in London. 鈥淏ut these findings certainly help explain why pregnant women in endemic areas are more likely to become infected with malaria.鈥�

Because HIV mutates very rapidly, strains that resist a drug鈥檚 toxic effects can evolve very quickly. But William Rice and his colleagues at the National Cancer Institute near Washington DC believe that the key to preventing the development of resistance may lie in structures within proteins called zinc fingers. These consist of four linked amino acids bound to an atom of zinc. Zinc fingers help maintain the three-dimensional structure of proteins. The protein which protects the inner core of HIV contains two zinc fingers. Unlike most other components of HIV, they do not readily mutate. 鈥淭hey have to stay the same for the virus to be able to multiply,鈥� says Rice.

女生小视频s at the cancer institute have identified several drugs from a family called the disulphide-substituted benzamides (DIBAs) which can remove zinc from the fingers, disabling the virus. In last week鈥檚 issue of Science (vol 270, p 1194), Rice鈥檚 team confirmed that HIV does not easily evolve resistance to these drugs. The researchers grew HIV in white blood cell cultures for over a year and continuously exposed them to DIBAs at doses that were too low to disable the virus. This is usually a reliable way to produce mutant viruses that are drug-resistant, but none appeared.

This means that DIBAs may be a useful addition to the drug cocktails that may one day hold HIV in check indefinitely. 鈥淚f zinc finger agents turn out to be as potent as these early experiments show, they might be very useful if we combined them [with other drugs],鈥� says Clive Loveday, a virologist at Middlesex Hospital in London.

In the same issue of Science (vol 270, p 1197), Che-Chung Tsai and his colleagues at the University of Washington in Seattle describe the first hints that it may be possible to eliminate HIV if drugs are given soon after the initial exposure to the virus.

The researchers worked on SIV infection in long-tailed macaques, which is thought to be very similar to HIV infection in humans. They gave a four-week course of a drug called PMPA to the monkeys, beginning either immediately before they were exposed to SIV or up to 24 hours after. After 13 months, they could find no trace of SIV in either group of treated animals.

Like the leading AIDS drug AZT, PMPA interferes with an enzyme called reverse transcriptase, which both HIV and SIV need to reproduce. AZT, however, cannot protect macaques from SIV infection. Tsai does not know why PMPA is more effective, but one difference is that PMPA is active in the form in which it is administered, while AZT must be phosphorylated inside the body before it becomes effective. Tsai adds that PMPA seems to stay longer inside cells. 鈥淚t鈥檚 something we鈥檝e observed, although we don鈥檛 know why it happens,鈥� he says.

The snag is that most HIV-positive people do not know exactly when they became infected, which means that giving a drug immediately after exposure is not often practical. But the new results suggest that PMPA or similar compounds could be useful in treating healthcare workers stabbed accidentally with an infected needle.

If the blood supply is cut off to cells, they develop holes in their outer membrane and die. When a person has a heart attack this is exactly what happens; the blood supply to the muscle of the heart is disrupted, creating holes in the membranes of the cells. Whether people survive an attack depends on how much of their heart muscle has been damaged in this way.

Ban-An Khaw and his colleagues wanted to see if they could limit the damage sustained by heart muscle by trying to patch up the holes in time to prevent the cells dying (Nature Medicine, vol 1, p 1195). 鈥淐ell death occurs because the holes in the membrane allow critical enzymes and salts to wash out of the cells,鈥� says Khaw. 鈥淚t鈥檚 like a car engine which runs out of petrol 鈥� the tank needs to be filled up again if it is going to carry on working.鈥�

The patching technique relies on little bundles of fat called liposomes, which in this case have antibodies attached. The antibodies Khaw has used so far are designed to bind to myosin, a protein which is plentiful inside heart muscle cells and which is responsible for the contraction of the muscle. The liposomes will be pulled over holes in the cell鈥檚 membrane because of the attraction between the antibodies on the liposome and the myosin inside the cell. Khaw believes the force of the attraction will pull the liposome into the hole, where it will eventually fuse with the cell membrane, which also consists of fat.

To test this theory, Khaw took cultured heart muscle cells that had been starved of oxygen and treated half of them with liposomes. The other half were left untreated. He reasoned that if the liposomes sealed the holes, the number of cells which survived would be a measure of success.

Without liposome treatment, only 5 per cent of the oxygen-starved cells survived. But with treatment, the survival rate increased to almost 90 per cent. Khaw also used silver grains trapped in the liposome to get an impression of how the cells were sealed. 鈥淚f the silver stays on the outside of the cells, we know they have been plugged,鈥� he says. But if the silver ends up inside the cells, the liposomes must have fused with the cell membranes.

For heart attack victims, Khaw estimates that plugging the cellular leaks and restoring the blood supply by administering standard clot-busting drugs could salvage up to 90 per cent of the heart muscle at risk of dying. Without plugging the holes, only around half the muscle is usually saved. Animal trials are due to start this week. If they are successful, Khaw says that human trials will follow.

Other researchers are already looking at liposomes as a delivery system for drugs and genes. Khaw also wants to explore this angle. 鈥淲e could use genes delivered by liposomes to convert fibroblast cells into muscle cells, so replacing scar tissue by real muscle,鈥� he says. (see Diagram)