Natural antibody molecules have two binding sites, which both bind to a
single type of antigen—a chemical marker on the surface of a cell. Now the
researchers have made antibodies that are bispecific—they bind to two
different antigens. They reasoned that if an antibody binds to antigens on two
different types of cell—T cells and tumour cells—then the molecule
would act as a bridge to hold the cells together.

But simply holding immune T cells and tumour cells together is not enough to
induce the T cells to kill the tumour cells. The T cells have to be activated,
and this only happens when receptor molecules on the T cell engage their
counterparts on the tumour cell. The researchers mimicked this by stimulating
the receptors with one of the antibody’s binding sites.

Pfreundschuh and his team tested a combination of two bispecific antibodies
to treat Hodgkin’s tumour of the lymph glands. One binding site of each antibody
attaches to CD30, an antigen on the surface of Hodgkin’s tumour cells. The other
binds to either CD3 or CD28, receptors which together activate the T cell.

The team used mice with a genetic defect in their immune system, which means
they can be injected with human cells without rejecting them. The mice injected
with Hodgkin’s tumour cells alone all died in around 40 days. In contrast, when
the researchers injected the antibodies and human T cells into mice with
widespread tumours they were all completely cured.

“This is a very good model because it uses human cells and so it is as close
to the human situation as we can get,” says Pfreundschuh.

They showed that T cells taken from patients with Hodgkin’s tumour worked as
well as T cells from healthy people. So Pfreundschuh speculates that the tumour
cells in cancer patients would be killed.

However, the mice were only cured if the treatment was given within seven
days of the injection of tumour cells. If it was left any later, then the tumour
always outstripped the treatment in the end, although the treated mice still
survived longer than the untreated ones. “The timing is critical,” says
Pfreundschuh. “It is a reflection of tumour load. It is difficult to relate this
to the clinical situation, but I speculate that this treatment will be
˛őłÜł¦ł¦±đ˛ő˛ő´ÚłÜ±ô.”

“A cure in a mouse does not necessarily translate into a cure in humans,”
says Peter Beverley, scientific head of The Edward Jenner Institute for Vaccine
Research near Oxford. “My worry is that the antibodies won’t get to the tumour
cells, so I am rather sceptical as to whether every last cell will be killed.
However, if this treatment is given after standard treatment has substantially
reduced the tumour load then I would be cautiously optimistic”.

Pfreundschuh and his colleagues have published their findings in
Blood (vol 87, p 2930).

Francis Chisari, who led the team at the Scripps Research Institute in La Jolla, California, thinks that the immune system may use the same stealthy mechanism to tackle other viruses. “Everyone has focused on the killing function of T cells. Now we must look at their other functions.” He hopes that the discovery will lead to new ways to treat viral infections.

Researchers have known for years that killer T cells play a key role in eliminating hepatitis B infection, and have assumed that they did so by killing virus-infected liver cells. But Chisari and his team became curious when they calculated that it would take the body’s entire population of killer T cells – around 100 billion cells – to kill all the infected liver cells.

To find out more about what the T cells do, the researchers devised a transgenic mouse in which all the liver cells make hepatitis B virus. When they injected killer T cells into the transgenic mice, the virus was cleared even though very few liver cells were killed.

Eventually, the team discovered how the killer T cells do this. They produce two messenger proteins, or cytokines, called gamma interferon and tumour necrosis factor alpha (TNF-α). The cytokines activate special inflammatory cells in the liver called Kupffer cells, which in turn produce more gamma interferon and TNF-α and other inflammatory cytokines. Finally, the cytokines tell the infected liver cells to stop producing the virus. The team publishes its results in Immunity (vol 4, p 25).

By injecting the mice with antibodies that block the action of gamma interferon and TNF-α, the researchers showed that these cytokines are responsible for setting off this chain of events. When the researchers blocked either cytokine singly, the infected cells were partially suppressed but still produced some virus. When they blocked both, the cells returned to producing large amounts of virus.

The obvious question is whether the cytokines could be used to treat people with chronic hepatitis B infection. Past trials have been disappointing. Howard Thomas of Imperial College School of Medicine in London found that neither gamma interferon nor TNF-α has much effect on the levels of hepatitis B virus in the blood of infected people. Also, the cytokines can produce toxic reactions, which limits the dose that people can be given.

But Thomas does not dismiss the possibility of future treatments. The previous trials might have failed because the level of the cytokines was too low, he says, and two cytokines given together might be more effective than either separately.

Chisari wants to sidestep the problem of toxic reactions by targeting the cytokines specifically to liver cells. “High doses are needed,” he says, “because cytokine receptors on other cells mop them up and only a very small amount ever reaches the liver.”

As well as devising ways to deliver the cytokines directly to the liver, his team is trying to find a short cut. He would like to find a way to activate the Kupffer cells directly so that they produce their cytokines and tell the infected liver cells to stop producing virus.

The ultimate goal, says Chisari, is to identify the genes in the liver cells that stop the production of viruses, and then find a way to switch them on. “This may give rise to a whole new class of antivirals,” he says. But he is cautious. “There are pretty stringent requirements for this to occur.”

The mutation is in a gene known only as X. It showed up in 17 out of 19 apparently healthy carriers, but was absent in 9 patients with hepatitis (The Journal of Infectious Diseases, vol 172, p 1191). The X gene codes for a protein that activates other genes which are needed for viral reproduction. Fukuda and his colleagues believe the mutation they have identified suppresses the activity of these genes, and so limits the production of new virus particles. In disease-free carriers, they point out, HBV reproduces much more slowly than in patients with chronic hepatitis.

But why should a mutation that damps down viral reproduction have become so common? Ian Weller, a specialist in sexually transmitted diseases at University College London, speculates that HBV can benefit from curtailing its reproduction because this makes it less likely to attract unwelcome attention from the immune system. Immune T cells kill infected cells, causing liver damage.

The new discovery means that HBV can now be classified into different strains on the basis of the X gene. “Other X mutations may exist that increase viral reproduction,” says Weller. Such mutations could explain why HBV sometimes causes fulminating hepatitis, a fatal form of the disease that causes liver failure within a month of infection. Identifying virulent strains and the people who carry them would allow researchers to pinpoint populations that should be targeted in vaccination programmes.

Some 85 per cent of the population have rhesus D-positive red blood cells, giving them a blood group known as D-positive. The remaining 15 per cent are D-negative. Rhesus disease occurs when a D-negative mother and a D-positive father conceive a baby whose red blood cells carry the D-positive blood group of the father. Just before or during birth some of the baby’s red blood cells pass into the mother. Her immune system sees them as “foreign” and makes antibodies against them called anti-D.

This causes no immediate problems, but the next time she carries a D-positive baby, the mother’s anti-D antibodies can pass through the placenta into the developing baby and destroy its red blood cells. This causes the severe anaemia, jaundice, and enlarged liver and spleen typical of babies suffering from the disease. Around 700 000 babies are born in Britain each year and 70 000 of these are at risk of developing rhesus factor disease.

Fortunately, rhesus disease can be prevented successfully by injecting pregnant women who are D-negative with anti-D from other sources. These injected antibodies destroy any of the baby’s red cells that find their way into the mother’s bloodstream before her immune system can react and produce its own anti-D.

At present, the anti-D used to prevent the disease comes mainly from the plasma of D-negative mothers of affected babies. But this form of treatment is a victim of its own success. Thanks to the prevention programme, there are very few mothers with antibodies. So D-negative male volunteers are now being immunised with D-positive red blood cells to spur their immune systems into making anti-D.

This is expensive, costing the National Health Service ÂŁ1.5 million a year. It can also be risky for the volunteers. Because they develop antibodies to rhesus-D, it may be more difficult to find compatible blood if they need an emergency transfusion.

“We rely on a panel of highly motivated donors who come to the transfusion centre every two to four weeks to donate plasma,” says Belinda Kumpel, of the International Blood Group Reference Laboratory in Bristol, which is one of the organisations conducting the research. “But there is still a shortage.”

Now researchers have taken the cells that produce anti-D from immunised volunteers and are growing them successfully in culture at Bio Products Laboratories, Elstree. This enables them to produce an unlimited supply of monoclonal anti-D antibodies.

To test the effectiveness of these antibodies in preventing rhesus factor disease, a team at the Blood Transfusion Centre in Bristol used D-negative male volunteers to mimic what happens in at-risk mothers. First, they injected them with monoclonal anti-D and followed it up with a dose of D-positive red blood cells. The cells were tagged with a radioactive label so that they could be tracked in the body. The researchers showed that the D-positive red blood cells were cleared very quickly from the blood and that none of the 27 immunised volunteers developed their own antibodies.

Treatments relying on monoclonal antibodies have run into difficulties in the past. Sometimes the immune system recognises monoclonal antibodies as foreign and produces an anti-antibody, which makes the treatment ineffective. None of the volunteers in the study produced this type of response after a single dose of anti-D.

“But the critical question is whether immunised individuals produce antibody against the injected anti-D after multiple doses,” says Willem Ouwehand, a lecturer in transfusion medicine at the University of Cambridge. “This could neutralise the effect of the monoclonal anti-D. And 27 volunteers is a low number to draw conclusions from. There is a great need to expand the numbers.”

Kumpel says that further tests are underway: “The number of pregnancies with a possible rhesus-D incompatibility will always be the same so we aim to prevent the disease in the safest possible way.”

Insulin-dependent diabetics all carry antibodies against insulin and another protein found in the pancreas called glutamic acid decarboxylase (GAD). But some people who have these antibodies do not have diabetes. And for years, researchers have struggled to find out what triggers the T cells that wreak havoc in the pancreas.

Andrew Lew and his colleagues at the Walter and Eliza Hall Institute of Medical Research in Melbourne think they have solved the problem. They believe that the T cells target proinsulin, the molecule from which insulin is made, rather than insulin itself. “Everyone else may have been looking at the wrong thing,” says Lew.

The key appears to be a tiny protein fragment, or peptide, that forms part of proinsulin. It may one day be possible, the researchers say, to manipulate the immune system so that it ignores the peptide and the cells that produce it.

Lew and his team were led to the finding by previous laboratory studies, which showed that some T cells from people with insulin-dependent diabetes respond to GAD by starting to proliferate. The Australian researchers were intrigued by the fact that a portion of GAD is similar to part of the proinsulin molecule. This peptide, just 13 amino acids long, occurs in part of the proinsulin molecule that is removed to form insulin.

To test whether they had found the elusive T cell target, the researchers added the peptide to cultured T cells. Cells taken from people who already had antibodies to GAD and insulin – and were therefore at high risk of developing diabetes – began to proliferate rapidly, just as they do in people who have the disease. When the cells came from people without GAD and insulin antibodies, they did not proliferate in response to the proinsulin sequence. In 40 per cent of cases, however, the T cells of people without GAD and proinsulin antibodies responded weakly to the GAD version of the peptide (Molecular Medicine, vol 1, p 625).

The researchers speculate that among people with T cells that react to GAD, a few will develop T cells that also respond to the proinsulin sequence, and then become diabetic. If so, it may be possible to desensitise the T cells of people at risk of developing insulin-dependent diabetes by exposing them to abnormally large quantities of the proinsulin sequence.

Animal studies have shown that T cells that respond to a particular peptide can be inactivated by taking other immune system cells from the body, coating them with the peptide, and then returning them to the bloodstream. “This has not been done in humans yet but it could be tried,” says Lew.

The problems with existing sirens make road junctions dangerous places when an ambulance is approaching at speed. In one year alone in the US, accidents involving ambulances killed 67 people and injured 537.

Researchers at Sound Alert asked 52 people with normal hearing to locate the source of a siren sound coming from one of eight hidden speakers during a simulated driving trial. These tests showed that today’s sirens, including the “hi-lo”, pulsar, wail and “yelp” varieties, are good at alerting people but are difficult to locate. “Even the best of the sirens in use at the moment has an error of localisation of over 45 degrees,” said Withington. “Road users always tell me that they know there is an ambulance coming but they don’t know where it is.”

For people to locate a sound without seeing its source, it must include a broad range of frequencies. The human ear can hear sounds at frequencies between 20 hertz and 20 kilohertz, but existing sirens only use the range from 500 hertz to 1.5 kilohertz. A mixture of all audible frequencies – known as white noise – would be the easiest to locate, but this would not make an effective warning signal.

Sound Alert has developed and patented new patterns of sound which can be pinpointed more easily, said Withington. The pattern its researchers have chosen for road trials is a series of “whaa” sounds of rising frequency followed by a blast of white noise. The end result sounds something like “whaa, whaa, whaa, shhhhhhh”.

Derek Smith of the West Yorkshire Metropolitan Ambulance Service said: “We endorse the project and we will be doing road tests on the new sirens.”

“We do the implant as soon as we can because after meningitis new bone growth may block the cochlea so that we can’t get the implant in,” said McCormick. “The youngest child we have treated is just 20 months old.”

Queen’s Medical Centre, which pioneered the technique in children, has carried out nearly a hundred implants. “We can do up to forty a year but it is very labour-intensive. One full-time staff member is needed for every 10 children implanted,” McCormick said. In Britain, about 300 children a year need a cochlear implant.

The device is a tiny, multichannel tube which is inserted into the spiral of the cochlea through a hole in the side of the skull. The tube, just half a millimetre in diameter, contains up to 20 electrodes which do the job of around 30 000 nerve endings that would normally transmit sound information from the cochlea to the brain. Once the implant is in place it has to be fine-tuned to the needs of the individual child. “Each one is as unique to the child as a fingerprint,” said McCormick.

The implants, which are made in Australia, are manufactured and soldered by hand under a microscope and cost £12 000 each – the entire operation costs around £31 000. However, this expense must be weighed against the cost of caring for a totally deaf person for life, said McCormick.

Henshaw measured the levels of alpha radiation emitted over several days from the tissue of babies and children born around the Severn Estuary. Newborn babies acquire their high levels of alpha radiation in the womb. “At the time of birth they already have a quarter of the level usually seen in a 10-year-old,” he told the BA last week.

Alpha radiation is produced by petrochemical pollution and is associated with a high incidence of leukaemia. Alpha particles, emitted by radioactive elements as they decay, damage chromosomes by breaking the DNA, causing permanent genetic damage that may lead to cancer.

Earlier studies by Ray Cartwright, an epidemiologist at the Leukaemia Research Fund Centre in Leeds, showed a two to threefold increase in childhood leukaemias in children living near river estuaries. Cartwright attributed this to pollution.

“In running the motor car we are very efficiently concentrating petrochemical pollution on roads, which is then washed into rivers,” said Henshaw. The action of tides concentrates pollution in estuaries. Local people breathe in the pollution when it becomes airborne in water vapour and spray. Henshaw said that petrochemical pollution in the Severn Estuary had been monitored for many years. It measured one part per million in 1850, but by 1970 it had reached more than 200 parts per million.

Alpha particles are produced by the decay of naturally occurring uranium and radon to lead-210, which is found in petrol and oils. In the body this isotope is concentrated in bone, where it decays into short-lived “daughters”, such as polonium-210, which emit alpha radiation.

Radon in particular is taken up by the fat cells of the bone marrow where it can reach concentrations sixteen times as high as in the blood. “Fat cells act as a sponge soaking up the radon,” said Henshaw. From here the “daughters” deliver alpha radiation to the developing blood cells in the bone marrow. These are the cells which give rise to leukaemia. Children are more sensitive to alpha irradiation than adults because they are growing and their cells are dividing more quickly.

According to researchers at the Institute of Naval Medicine in Gosport, Hampshire, cold shock is to blame. The body’s first reaction to immersion in icy waters is to hyperventilate, the BA heard, and the sudden intake of breath is followed by an influx of water – and drowning.

With a large tank of cold water, a ducking stool and a stream of volunteers, Michael Tipton and his colleagues investigated the body’s reaction to immersion in cold water. The naval team found that the hyperventilation response is so strong that it cannot be prevented even if volunteers try to hold their breath. The only way to survive cold shock is to keep the head out of the water until the response dies down. This usually takes two to three minutes, although swimmers who are used to cold water are able to suppress it faster.

Paradoxically, cold water may help people to survive if they live through the acute cold shock. “People who survive the cold shock response may then live for long periods in the water. The longest recorded case is a child who was revived after 40 minutes of immersion in ice-covered water,” said Mark Harries of Northwick Park Hospital, London.

Even people who are not breathing and whose hearts have stopped beating can be revived. Out of water, however, brain damage and death occur within minutes of the heart stopping. “This difference is due to body temperature. A cold brain can withstand a cardiac arrest much longer than a warm one,” said Harries.

The cells in cassava root contain highly poisonous cyanoglucoside. The latex contains enzymes which break down the poison, but the enzyme and the poison do not come into contact in the living plant. Processing cassava breaks down the cells, allowing the enzymes to reach the poisonous cyanoglucoside and convert it to hydrogen cyanide gas. The gas is given off, leaving the root safe to eat.

Any cyanoglucoside not removed before eating is converted into hydrogen cyanide in the body. This can cause chronic cyanide poisoning and paralysis, a condition known as “konzo” in rural Nigeria. In the past ten years up to 10 000 women and children in Zaire and Mozambique have developed chronic poisoning. In the worst-hit villages one in five has been permanently crippled.

There are “around 184 different methods” of processing cassava in Africa, says Hughes. The commonest way of removing the poison is by fermentation in which naturally occurring lactobacteria break down the cell walls and release the cyanoglucoside so that the enzymes can detoxify it. But this is not very reliable, and often some poison is left in the food.

Hughes told the BA that it might be possible to use the newly cloned genes to improve the fermentation process. She aims to make a starter culture for fermentation with lactobacteria that have been genetically engineered to manufacture the cyanoglucoside-destroying enzymes. “We will use the natural bacteria from the fermentation processing plants in Africa and distribute the starter culture free. It will then be self-perpetuating,” she says.

A longer-term strategy is to alter genes in the cassava plant itself so that the roots produce more enzyme. “The level in the leaves is quite high, but it is low in the root,” says Hughes. “It is now possible to improve detoxification during the fermentation by redesigning the plant to increase the amount of enzyme in the root.” From then on, providing the improved plants would be easy. “The plant is propagated vegetatively just by replanting sticks, so we only need to modify one plant and give it back to the farmers,” says Hughes.