Valium belongs to the class of tranquillising drugs known as benzo diazepines.
These drugs enhance the activity of a chemical messenger, or neurotransmitter,
that reduces the level of excitement in the brain. The neurotransmitter,
which is called GABA, acts as a signal between neurons that recognise it.
Benzodiazepines bind to GABA receptors on the neurons, enhancing the signal.

Researchers have known since the 1980s that humans, rats and several
other species have natural benzodiazepines in their brains. In rats, benzodiazepine
levels rise in the hippocampal areas of the brain – structures associated
with perception of anxiety – when the animals are stressed. But researchers
have also suspected for several years that these molecules might be involved
in memory: first, because the American scientist Jim McGaugh of the University
of California at Irvine showed in the mid-1980s that the GABA system helps
to control memory; and secondly because people taking benzodiazepine drugs
sometimes have memory lapses.

Izquierdo and his team used antibodies to detect levels of benzodiazepines
in different parts of rats’ brains during two types of training experience.
In the first, the rats were allowed to learn about a new environment – a
box. The researchers found that benzodiazepine levels rose in areas of the
brain associated with alertness and learning – the cortex, amygdala and
medial septum – but not in the hippocampus. This did not happen in other
rats that were not learning.

In a second experiment, the rats received a small electric shock through
their feet as they began to explore the box. This meant that they were learning
about the same environment but under stressful conditions. This time the
levels of benzodiazepines were two or three times higher in the cortex,
amygdala and medial septum as in the first experiment – as well as higher
in the hippocampus, as you might expect because the animals were stressed.

Izquierdo then did some experiments to see what would happen if the
effect of the benzodiazepines that were released during learning was blocked.
He used flumazenil, a drug that stops the binding of benzodiazepines to
their receptor (and therefore tends to cause anxiety). When Izquierdo injected
flumazenil into rats just before they were trained to avoid unpleasant stimuli,
they were unable to remember what they were taught.

Izquierdo concludes that ‘a degree of arousal, anxiety or stress may
be necessary for learning’. In other words, new information that does not
excite arousal is deemed unimportant by the brain and filtered from the
memory. At the same time, arousal in its exaggerated form – anxiety – will
have the same effect on memory.

Peter Rogers, from the Agricultural and Food Research Council’s Institute
of Food Research in Reading, and John Blundell from the University of Leeds,
gave aspartame in capsules to 16 undergraduates one hour before a sandwich
lunch of 1200 kilocalories. Each capsule contained 200 milligrams of aspartame
– equivalent to the amount in a can of diet cola. A control group took capsules
containing no aspartame.

Before the meal, both groups felt equally hungry, according to standard
psychological tests. However, the group that had taken aspartame ate 15
per cent fewer kilocalories than the controls, as measured by the amount
of food left on their plates.

The researchers believe that, while aspartame does not reduce hunger,
it exaggerates the satiating effects of food. The experiment has been repeated
six times, Rogers said. The researchers believe aspartame could eventually
be used as a therapy against obesity, although they warn that much more
research would be needed.

There have been numerous studies on aspartame but no one has found this
effect before. One explanation, suggests Rogers, is that most of the early
studies were in rats; yet the sweetener has no effect on these animals’
food intake, he says.

Another possible explanation for the failure to note the effect before,
says Rogers, is that the taste of sweeteners may actually stimulate appetite.
The team’s own research on saccharin and aspartame has found that these
sweeteners do increase appetite when people can taste them, although other
researchers have found no such effect (Physiology and Behaviour, vol 39,
p 247). The reason that Rogers and Blundell gave the aspartame in capsules
was to avoid any possible effect of sweetness on appetite.

The team cannot provide any evidence for a mechanism by which the sweetener
makes people feel full, but they say this is the first demonstration that
aspartame has a biological effect other than tasting sweet. A hormone called
cholecystokinin, released from the gut, is involved in the feeling of satiety
after eating. The researchers speculate that receptors on cells that release
the hormone in the gut wall may be sensitive to aspartame, and that the
sweetener may trigger the release of additional cholecystokinin.

Erik Millstone, a researcher at the Science Policy Research Unit at
the University of Sussex, says it would be misleading to put too much weight
on one study. ‘Using single doses of aspartame in capsule form does not
faithfully reproduce normal patterns of human consumption,’ he said.

Maureen Mackey, associate director of nutritional science at the NutraSweet
Company, said she could not comment until the results had been published
in a journal. But she was sceptical about the claim that aspartame could
be used to treat obesity.

Lisa Matsuda, a molecular biologist, working in the laboratory of Tom
Bonner at the National Institutes of Health (NIH) in Washington DC discovered
the receptor. The details of her work are published in today’s issue of
Nature (vol 346, p 561).

It is unlikely that nature has endowed humans with such a specific receptor
simply to tempt us to smoke a plant indigenous to central Asia, so there
must be a more fundamental biological reason why the ‘cannabinoid’ receptor
exists. The most obvious reason is that the brain may actually make a cannabis-like
chemical which, by binding and activating the receptor, acts as a messenger
between cells. This would mean that the receptor could, in certain circumstances,
cause some of the effects experienced by smokers of marijuana.

Cannabis is not only used as a ‘recreational drug’; cannabinoids – the
active constituents – have a medical use. For example, the cannabinoids
suppress the vomiting reflex, and synthetic analogues, such as nabilone,
are used clinically.

But until recently scientists were not clear how these drugs worked.
Cannabinoids, unlike most drugs, are soluble in lipids rather than in water.
This led scientists to believe that the substances simply dissolved in the
lipid part of cell membranes, disrupting the organisation of membrane proteins
which are responsible for communications between cells.

However, in the mid-1980s, Allyn Howlett of St Louis University Medical
School in the US found evidence that the action of cannabinoids might be
more specific – that is, a receptor might be involved. Howlett’s work suggested
that the receptor might be one of a family of receptors for neurotransmitters
– substances that act as messengers between brain cells – which act by reducing
the activity of the key enzyme in cells, adenyly cyclase. This family of
receptors belongs, in turn, to a superfamily known as G protein-linked receptors.

Ironically, the cloning of the receptor, reported in today’s issue of
Nature, did not follow directly from Howlett’s work. Instead, it was a quite
serendipitous discovery by Matsuda at NIH.

Researchers in Bonner’s laboratory have successfully cloned a series
of neurotransmitter receptors using a so-called ‘fishing’ process. Receptors
in the superfamily have similar structures, so it is possible to develop
a probe for one receptor – with a known structure – and use it to ‘recognise’
a shared sequence on a similar receptor in the same family. The techniques
of molecular biology allow biologists to separate the new receptor in such
a way that cells grown in culture can be persuaded to synthesise it and
insert it into their own membranes. Researchers can then study and identify
it.

Things were not so simple, however, with the cannabinoid receptor. When
Matsuda expressed the receptor in cultured cells, she could not find any
drug that would bind to it and activate the cells. In such cases, clues
to the identify of a receptor may come from studying its distribution. But
even this technique yielded no positive leads for Matsuda.

Finally, Matsuda hit on the idea that the receptor might recognise cannabinoids.
Experiments proved that she had indeed found a cannabinoid receptor.

Bonner believes it very likely that a natural cannabinoid exists in
the brain. He has preliminary evidence that the gene for the cannabinoid
receptor is also present in the fruit fly Drosophila, which suggests that
ancestral genes must have been present very early in evolution. ‘If this
gene has been conserved throughout evolution,’ he says, ‘then this suggests
that its product (the receptor) must serve an important function.’

Matsuda admits to being ‘intrigued’ but prefers to remain cautious about
the significance of her findings. ‘The function of this system may be one
of dampening inputs into the brain to allow certain memory processes to
occur, but much more work is needed.’

Drugs companies will be hoping to exploit the finding but, to date,
they have been wary of trying to market products associated with ‘drugs
of abuse’.

The mechanism is simple. The cell that receives the signal – from a
hormone, for example – responds by synthesising nitric oxide, which then
diffuses between that cell and the true target cell, where it acts.

Scientists first identified nitric oxide as a messenger, or biological
mediator, only two years ago. They discovered it in two different systems.
In 1987, Salvador Moncada and his colleagues at the Wellcome Institute in
Beckenham, Kent, found that a much-studied substance in the body, called
endothelium-derived relaxing factor (EDRF) was nitric oxide (Nature, vol
327, p 524). EDRF had been the focus of much interest because it seems to
influence blood pressure and the clotting of blood. It is released from
the endothelium, the sheet of cells which line the smooth muscle that forms
blood vessels.

Many biological signals, such as neurotransmitters, control the tone
of blood vessels but not all act directly on the vessel walls. Moncado and
his colleagues found, for example, that some trigger the endothelial cells
to release nitric oxide, which can then diffuse across the small gap between
endothelium and blood vessel wall and cause the muscle to relax, dilating
the vessel. They also found that nitric oxide inhibits the clumping of platelets,
a primary event in the development of a clot.

Also in 1987, John Gibbs and his colleagues at the University of Utah
reported that macrophages, a kind of immune cell, release nitric oxide when
they are activated by bacteria (Science, vol 235, p 473). The molecule also
diffuses short distances to tumour cells and is capable of killing them.
Gibbs also discovered that cells produced nitric oxide from the amino acid
arginine, and this has been confirmed by other researchers. Most believe
that the rise in levels of intracellular calcium which is a common response
of cells to activating signal, switches on the enzyme, nitric oxide synthase,
which makes the molecule.

These major discoveries have prompted biologists to look for nitric
oxide in other cells. John Garthwaite of the University of Liverpool has
shown that the brain makes nitric oxide when at least two types of receptor
for glutamate are excited.

Glutamate is an amino acid neurotransmitter that excites nerve cells.
Researchers believe that one of these receptor types, the NMDA receptor,
is involved in the processes of learning and memory. The receptor is also
involved in causing the irreversible brain damage that occurs after a stroke.
Garthwaite found that the nitric oxide diffuses out of the receptor-bearing
cell to different types of cell close by. It somehow modifies their activity.

S. H. Ferreira in Brazil has shown that nitric oxide regulates peripheral
nerves involved in suppressing pain. Cells known as neutrophils, the immune
system’s first line of defence against acute infection, also generate nitric
oxide. The mediator here may have a dual role in killing bacteria and in
regulating the blood flow at the site of infection.

How can such a small and simple molecule have such a diversity of effects?
And why should some cells see nitric oxide and respond appropriately, while
others see nitric oxide and die? One answer could be that more than a threshold
concentration of nitric oxide kills. Cells that make nitric oxide may protect
themselves from a build-up of nitric oxide by also producing substances,
so far unidentified, which promote its degradation.

Now researchers have found that it is possible to manipulate pharmacologically
the production of nitric oxide in humans (1989, The Lancet, ii, p 997).
Joe Collier and Patrick Vallance from St George’s Hospital Medical School
collaborated with Moncada to show that an inhibitor of nitric oxide synthase,
L-NMMA, reduces the blood flow through the forearm of healthy volunteers.
Moreover, the drug glyceryl trinitrate (GTN), commonly used in treating
angina, had the opposite effect. Doctors know the body converts GTN to nitric
oxide.

Thus we already have a drug which exploits nitric oxide. Because of
the molecule’s diverse action, compounds that interfere with its production
may have wide-ranging therapeutic implications.