ÐÓ°ÉÔ­´´

Animals used in experiments face two sorts of hazards. First, they
are frequently kept in barren and unstimulating cages in environments that
disturb them. They may suffer because of the excessive noise of machinery
that emits high frequencies inaudible to humans, or because the smell of
other animals leads to sexual frustration or fear. Secondly, the experiments
themselves may cause the animals pain and stress and may expose them to
an unfamiliar room or cage, or an unfamiliar handler. Unaccustomed noise,
light intensity or confinement, or separation from cage mates may also affect
the animal.

To improve the lot of laboratory animals, scientists need to find objective
ways of quantifying stress. Researchers could then determine whether the
way animals are housed or the procedures they undergo are causing an unacceptable
level of stress or pain.

Stress occurs when an individual animal encounters something that disturbs
its normal physiological and mental equilibrium. The animal responds by
modifying its behaviour and physiology in an attempt to deal with these
adverse conditions.

An animal’s behaviour can sometimes be a clear sign of stress. A frightened
or anxious animal may adopt a characteristic vigilant posture, attempt
to flee or hide, or urinate or defecate more frequently than normal. It
may stop feeding, exploring or interacting with its fellows. Some animals,
particularly rodents, may ‘freeze’, or become motionless. When an animal
has no control over its situation, or no chance of escape, it may behave
in inappropriate ways – rats, for instance, may groom, fight or chew inedible
substances during electric shock. Animals confined in barren surroundings
in zoos, farms or laboratories often pace or circle endlessly in their
cages. So behavioural tests might help to quantify the degree of anxiety
an animal is suffering.

Robert Barclay and his colleagues at the University of Dundee recently
devised a test that monitors the fall-off in a rat’s normal exploratory
behaviour in order to measure how much fear or discomfort various procedures
cause. Barclay used an electronic monitor to measure the rat’s normal activity
in a box that the animal is accustomed to, and then again after an experimental
procedure such as an injection or restraint. He found that rats explored
less after the administration of saline by mouth or by injection, and after
restraint by an inexperienced handler, but not after restraint by an experienced
person.

But behaviour alone may not be a foolproof sign of stress. Stressed
animals often appear to damp down behavioural responses, perhaps to avoid
attracting the attention of predators. So it is useful to be able to measure
physiological responses as well.

Almost any form of emotional arousal elicits the release of glucocorticoid
hormones by the adrenal cortex, the outer part of the adrenal gland. Increased
levels of these hormones – corticosterone in rodents or cortisol in primates,
dogs and cats – follows courtship or anticipation of a meal, as well as
fear or pain. These hormones have been studied as indicators of the animal’s
response to a wide variety of unpleasant stimuli. For example, Michael Hennessy
and his colleagues at Stanford University in California showed that when
rats were exposed to new cages there was a rise in the level of corticosterone
in their blood. Levels increased even when rats were moved to a new cage
that was identical to their home cage. They were higher still if the cage
contained no sawdust and were highest of all in rats put into a cage with
an empty food container.

One problem with using glucocorticoids as indicators of stress is that
they are so easily raised that any handling of an animal tends to produce
a significant change. To measure glucocorticoids it used to be necessary
to take a blood sample, which itself is likely to cause stress. This could
only be avoided by collecting blood samples so rapidly that the adrenal
cortex did not have time to respond. Recently, it has become possible to
measure glucocorticoid in saliva, and the University of Cambridge’s animal
welfare group now measures the amount of cortisol in saliva samples from
pigs, calves, dogs and cats as part of its welfare investigations.

The catecholamines, the hormones adrenaline and noradrenaline, are responsible
for the fight-flight response in individuals who are frightened or angry.
These hormones are produced by the sympathetic nervous system and the adrenal
medulla, the inner part of the adrenal gland. They can be measured directly
in blood or urine, and their effects on heart rate and blood pressure can
be quantified too. However, these catecholamines are so sensitive to stress
that it is impossible to take blood samples without causing them to rise.
The only way to sample them in blood is to use an intravenous catheter and
take samples remotely, without disturbing the animal.

The alternative is to sample them in urine, although levels there vary
throughout the day, depending on how much urine is produced. There are also
drawbacks to trying to collect urine from animals: confining animals in
plastic ‘metabolism’ cages, constructed so that urine drains into a tray
below, is stressful in itself. But it may be feasible to avoid this by training
nonhuman primates and dogs to urinate in a particular place. Even mice will
urinate in a bottle if it is provided.

Catecholamines are also important in the brain, where they act as neurotransmitters
in certain nerve pathways, carrying nerve impulses from one nerve ending
to another. Stress increases activity in certain areas of the brain, leading
to an increased use of neurotransmitters, especially noradrenaline.

The breakdown products of noradrenaline in both brain and blood can
also be detected in blood or urine, and increasing levels indicate that
the animal is or has been stressed. An exciting new development is that
one of these products, vanillylmandelic acid, has now been measured in human
saliva. If it can be detected in the saliva of other species, this substance
might be a useful indicator of stress, which could be measured noninvasively.

A rapid heart rate can also indicate fear or anxiety. Researchers can
now monitor heart rate directly, without disturbing the animal, by placing
electrodes next to the skin and monitoring the signal by radio. The electrodes
and radiotransmitter can be fitted under a tight jacket or under an elasticated
strap fitted around the chest of animals as small as rats. Such research
has shown that heart rate increases in an animal responding actively to
fear, but falls below baseline when the animal ‘freezes’.

Blood pressure is another telltale sign of stress. Any emotional stress
that activates the sympathetic nervous system and prompts the fight-flight
response can cause blood pressure to rise. Researchers can measure blood
pressure, as in humans, by using a cuff on the limb of a trained nonhuman
primate or on the tail of a dog or rodent. But care must be taken to avoid
frightening the animal and so unintentionally raising the blood pressure.

Animals stressed over a long period may even develop permanently raised
blood pressure. Chronic activation of the sympathetic nervous system results
in the sustained constriction of blood vessels. Eventually these vessels
become thickened and lose their elasticity, leading to permanent high blood
pressure or hypertension. This can happen in animals exposed to long-term
experimental stress, or in the dominant animal of certain primate and rodent
colonies, presumably as a result of the stress of maintaining their position
in the social order.

Stressed mammals also produce higher than normal concentrations of
‘opioid’ peptides – substances with effects similar to compounds derived
from the juice of the opium poppy. The opioid peptides are produced by many
organs of the body, especially the brain, the anterior pituitary gland
and the adrenal medulla. When released into the brain, some of these opioid
peptides act as painkillers – accounting for the fact that severely injured
soliders may fight on, unaware of their wounds. A similar phenomenon has
been demonstrated in animals under stress. During stress, another set of
opioid peptides, particularly beta-endorphin, may be released into the circulation.
Their function here is not clear, but their levels may rise during stressful
situations.

There is now a great deal of evidence to suggest that stress in humans
and other animals can lead to a suppression of the immune system. Stress
leads to a reduction in the numbers and the efficiency of certain types
of lymphocyte, the white blood cells that are important in immune function.
In laboratory animals, a wide variety of stress factors have been found
to reduce the immune response. Vernon Riley at the University of Washington
in Seattle showed that even subtle improvements in housing such as the reduction
of noise, draughts and odours from other cages reduced the susceptibility
of mice to mammary tumours.

Susceptibility to tumours, either implanted or naturally occurring,
is one way of measuring the functioning of the immune system. Another simple,
but more useful, technique is the white cell count. While the number of
lymphocytes is reduced by stress, other white blood cells are unaffected.
So measuring the relative numbers of white cells in a blood sample, a so-called
differential count, can indicate whether an animal has been stressed.

More sophisticated methods of assessing immune function involve laboratory
techniques such as the ‘lymphocyte mitogenesis’ test. This test exploits
the finding that lymphocytes in culture begin to divide rapidly when certain
plant lectins are added. Lymphocytes from an animal with a suppressed immune
system multiply at a lower rate than those from a normal animal. Another
widely used test is based on the fact that a proportion of circulating lymphocytes
have so-called natural killer activity and can destroy host cells that have
been invaded by a virus or tumour. The test measures the ability of a sample
of lymphocytes to destroy tumour cells in laboratory cultures.

Both these tests have been widely used in studies of humans. People
who have been bereaved or suffered marital breakup may show reduced lymphocyte
mitogenesis and natural killer cell activity, as do students taking examinations.
The tests have been used in stressed animals too. For example, Stephen Keller
at Mount Sinai School of Medicine in New York found that the degree of
immunosuppression in rats, measured by mitogenesis, was linked to the intensity
of electric shocks that had been applied to their feet.

Stress over the long term can lead to gross changes in the anatomy of
organs which are clearly visible in postmortems. Rabbits and mice kept in
overcrowded cages, for example, have been found to have arteries with thickened
walls and heart muscle with microscopic lesions. Similar lesions have been
seen in pigs transported and slaughtered under stressful conditions, and
in mice after electric shock.

Stomach ulcers can occur in rats and mice that have been stressed by
being restrained for an hour or two, after being deprived of food for 24
hours. In fact, rats treated in this way are used as experimental models
to test drugs for possible ulcer-causing side effects.

An enlarged adrenal gland indicates that it has been chronically over-stimulated.
Stress-induced immune suppression often leads to smaller spleens and
thymus glands. Stress-induced cardio-vascular changes may lead to an increase
in the weight of the heart. By comparing organ weights to body weight at
post-mortem, researchers can confirm that an animal has been under stress.

A great deal of information on physiological responses to stress has
been gathered in research carried out mainly in the US. In this research,
animals were deliberately subjected to stressful experiences such as electric
shock, immobilisation in narrow plastic tubes or immersion in cold water.
Undoubtedly considerable suffering has been caused in this way. It seems
appropriate that laboratory animals should now benefit from our improved
understanding of the behavioural, physiological, immunological and pathological
effects of stress. Numerous systems of an organism are highly sensitive
to the effects of stress, and the physical response matches the intensity
of the stress factor. We are now in a position to use measurements of the
stress response to assess, and so improve, the welfare of animals in laboratories.

Caroline Manser is in the Department of Clinical Veterinary Medicine
at the University of Cambridge. This article is based on a detailed report
she prepared for the RSPCA which is to be published shortly.