ÐÓ°ÉÔ­´´

Between the nervous tissue of the brain and the blood there is a barrier
that prevents substances in the blood from entering the brain. This ‘blood-brain
barrier’ is formed by the cells that line the blood vessels of the brain.
It has evolved because nerve cells need a stable chemical environment. Without
it, nerve cells would be subjected as we eat or exercise to fluctuations
in concentrations of glucose, amino acids and hormones among other compounds,
leading to uncontrolled nervous activity and even fits.

But although the barrier protects the brain from unwanted chemical invasions,
it has one disadvantage: it also effectively excludes many drugs that could
treat diseases of the brain. Much research is now focused on trying to understand
the way the barrier develops and functions, and to find ways of conveying
drugs across it to treat certain diseases.

The basis of the blood-brain barrier is the endothelium, the layer of
cells that lines the countless small blood vessels, or capillaries, that
permeate the brain, supplying it with nutrients and carrying away waste
products. Between the brain’s endothelial cells are special junctions that
prevent substances in the blood from diffusing through the walls of the
vessels to reach the nervous tissue. So the endothelial cells form a virtually
impermeable layer between the blood and the brain.

But the brain requires nutrients including oxygen, glucose and amino
acids. How does it get them? If the barrier were complete, the brain would
starve to death. Because they have special structures to keep out unwanted
substances, the cells that line the brain capillaries also must have special
mechanisms to transport vital molecules into and out of the brain.

Much research on the blood-brain barrier has been focusing on these
transport systems. It is now known that endothelial cells have several types
of transporter, each carrying a specific molecule or type of molecule. How
well any molecule is excluded from the brain or transported across the barrier
depends on the structure and function of the capillaries and on the physics
and biochemistry of the molecule. Using recently developed preparations
of isolated brain capillaries and cultured brain endothelial cells, investigators
are now building up a picture of how molecules are transported across the
blood-brain barrier. This is leading to a better understanding of ways to
deliver therapeutic drugs to the brain.

The idea of a blood-brain barrier arose in the late 19th century when
the German bacteriologist Paul Ehrlich noticed that certain dyes injected
intravenously into small animals stained all tissues except the brain. Even
so, Ehrlich was sceptical of the existence of a selective barrier. Twenty
years later, one of his associates, Edwin Goldman, working with rabbits
and dogs, did the experiment in reverse: he injecting the dye, trypan blue,
directly into the cerebrospinal fluid, which fills the cavities or ventricles
of the brain. This time, the dye stained the entire brain – but it did not
enter the bloodstream to stain the other internal organs. Goldman concluded
that the central nervous system was separated from the blood by some kind
of barrier, probably in the brain’s capillaries.

With the introduction of electron microscopy in the 1950s, Goldman was
proved correct. Although it is still not entirely clear how the capillaries
of the brain function so differently from vessels in other organs, they
have at least three distinctive features. First, the ‘tight’ junctions between
the endothelial cells are extremely tight – the cell membranes are fused
along the length of their contact, rather like a zip fastener. In contrast,
the capillaries elsewhere in the body usually have gaps in the junctions
between their endothelial cells. Secondly, the brain’s capillary cells have
very few pinocytotic vesicles, the membrane-bound capsules within the cell
that may help to ferry small amounts of fluid and solutes across the cell
wall. These vesicles are common in endothelial cells outside the brain.
Thirdly, the brain side of the capillaries is encased in a type of brain
cell known as astrocytes. Because the astrocytes hold the capillaries in
their grasp, some of the early investigators thought they might be the blood-brain
barrier. Although this idea has been disproved, the astrocytes may well
play a part in inducing the brain endothelial cells to form their characteristic
tight junctions. Indeed, astrocytes can make endothelial cells from other
parts of the body act like a blood-brain barrier when artificially brought
into contact with them.

That the capillary endothelium is the site of the barrier was finally
established in the US in the 1960s by Thomas Reese and Morris Karnovsky
at the Harvard Medical School. They used an enzyme, horseradish peroxidase
(HRP), as a tracer in a modern version of the experiments with trypan blue.
HRP, which is similar in size to proteins normally present in the blood,
produces dark patches of electron-dense deposit within and between cells
when it is treated with specific chemicals. Reese and Karnovsky found that
HRP injected into the bloodstream penetrated the capillary walls of most
organs easily – either by passing through endothelial junctions or by being
taken up into endothelial cells in pinocytotic vesicles. In the brain, however,
HRP found its way barred by the tight junctions between the endothelial
cells, and little was carried into the endothelium in vesicles.

In 1969, Reese, by then at the Institute of Neurological and Communicative
Diseases and Stroke at Bethesda, Maryland, and Milton Brightman repeated
Goldmann’s experiment. When they injected HRP into one of the brain ventricles,
they saw that it was the tightly joined endothelial cells, and not the astrocytes,
that prevented the HRP leaving the brain. Thus the endothelium was established
as the blood-brain barrier.

Neuroscientists then began to study how substances cross the barrier
by labelling compounds with radioactive markers. In general, the physics
and chemistry of a molecule determine how easily it gets into the brain.
The most decisive factor is how soluable it is in the lipids (fats) and
make up cell membranes, which is roughly equivalent to how easily it dissolves
in oil. Lipid-soluble molecules readily cross the blood-brain barrier. They
include nicotine, ethanol and heroim, which is one reason why these drugs
are so effective. Charged molecules, which are not lipid-soluble, enter
the brain slowly, or not at all, unless they have specific transport systems.
These molecules range from compounds as large as proteins down to ions as
small as sodium – proving that the size of the molecule alone is not a key
factor.

Yet some of the brain’s main requirements, such as glucose, its main
source of energy, and amino acids that its cells cannot make for themselves,
are not lipid-soluble and so cannot simply diffuse across the barrier. Each
nutrient must be recognised and taken across the membrane by a transporter
that is specific for it.

The human brain uses more than 120 grams of glucose a day but cannot
store much more than about 2 grams. So a constant supply of glucose must
be transported across the barrier. And, as Christian Crone, of the University
of Copenhagen, first showed, the transporter system will accept only one
form of glucose. In a series of sophisticated experiments, he compared the
rates at which the two mirror-image forms of the sugar, D-glucose and L-glucose,
crossed the blood-brain barrier. The body tolerates both, but it can use
only D-glucose as a source of fuel. Crone found that only D-glucose could
be extracted from the blood as it flowed through the brain. He concluded
that the special endothelium of the capillaries serving the brain has a
transport system that is highly specific for this form of glucose. The situation
is quite different from that in muscles, where both D-glucose and L-glucose
are ‘extracted’ at similar rates, because they can both diffuse out between
the leaky tight junctions of the endothelium.

Border crossing

More detailed studies of the glucose transporter began when techniques
for isolating brain capillaries were developed by Gary Goldstein, then at
the University of California’s San Francisco Medical School. With isolated
capillaries, researchers can see the blood-brain barrier at work without
the confusing influences of all the other cell associations in the brain.
The functional characteristics of the glucose transport system, the busiest
transport system of the blood-brain barrier, are now clear: each endothelial
cell is richly endowed with transporters that enable it to take up large
amounts of glucose from the blood. Only a small amount of this sugar is
used in the endothelial cell itself. The rest is transferred into the brain.

The detailed structure of the transporter molecules is, however, still
a mystery. Each transporter is probably one or more protein molecules that
span the cell embrane to create a channel through which glucose passes,
but the precise nature of the proteins is still being worked out.

The transporter systems for amino acids are much more complicated, because
the 20 essential amino acids all have different molecular structures. They
can be divided according to their chemical properties into four categories:
the large neutral, the small neutral, the basic and the acidic. Each category
has its own transport system. Like the transporters for glucose, those for
the large neutral amino acids are present on both sides of the barrier,
and so these amino acids can pass across the endothelial cell both into
and out of the brain. Small neutral amino acids, on the other hand, which
can be synthesised by the cells of the brain, so no transport system is
needed to carry them into the brain; there is, however, a transporter to
ferry them out into the blood.

Recent studies using isolated brain capillaries indicate that peptides
and certain proteins circulating in the blood can find other ways across
the barrier. One route is receptor-mediated endocytosis, where a receptor
molecule help a substance into and across the endothelial cell. Receptors
systems have been found for several molecules including insulin, low-density
lipoprotein (which carries cholesterol around in the blood), insulin-like
growth factors, and transferrin, an iron-binding plasma protein. Studies
using radioactively labelled transferrin and insulin indicate that these
molecules do cross over into the brain tissue. It is not yet clear whether
the other substances are merely delivered into the endothelial cells, or
whether they actually cross the barrier.

A third path across is by hitching a lift on plasma proteins, such as
certain globulins and albumin. A number of free fatty acids and steroid
hormones, such as the male and female sex hormones, testosterone and oestradiol,
can readily cross the blood-brain barrier, not only because they dissolve
easily in lipids, but also because they are avidly bound by albumin, which
delivers them smartly to the brain endothelial cells. Once at the endothelium,
the hormones are delivered to ‘docking’ sites, such as receptors, where
they can dissociate from the albumin, cross the cell layer and enter the
brain.

But the blood-brain barrier does not always function effectively. In
many diseases, substances that are normally excluded may enter the brain.
Sometimes this can be beneficial. The antibiotic penicillin, for example,
is normally virtually excluded from the brain as it is insoluble in lipids.
In some forms of meningitis, however, the barrier is unusually permeable
and sufficient antibiotic can enter to be effective. But not all compounds
are as beneficial as penicillin and many could be poisonous.

Normal transport processes may also provide an entry for harmful substances
or may themselves become impaired. For instance, the aluminium deposits
that can appear in the brain in Alzheimer’s disease might get in by hitching
a lift on the transferrin receptor, which cannot distinguish between iron
and aluminium. And the plaques that are characteristic of Alzheimer’s often
have the blood protein immunoglobulin G associated with them. The immunoglobulin
may get into the brain as a result of a breakdown of the barrier. The glucose
transporter may also go awry in older people, resulting in abnormal sensitivity
to normal levels of glucose in the blood. The elederly also seem to be more
susceptible to the sedative effects of the neurotransmitter serotonin, by
a mechanism involving the neutral amino-acid transport sites at the barrier
– something that the pharmaceuticals industry has exploited as a treatment
for agitation in the elderly.

Breakdown of the barrier has been implicated in several other diseases,
including tumours, brain lesions and strokes, in which a major role is played
by cerebral oedema – the abnormal swelling of tissue when fluids and proteins
build up in the brain. The barrier becomes leakier because there are more
pinocytotic vesicles in the endothelial cells, or because the tight junctions
between the cells are forced apart, or because of a combination of both.

Damage to the barrier may play a role in the accumulation of fluid in
the brain tissue that occurs in lead poisoning. Studies have shown that
the metal first moves into the endothelial cells, and later into the astrocytes.
One possibility is that the brain is more vulnerable to attack from other
substances after lead in the blood has broken the barrier down.

One prospect is that the blood-brain barrier could be opened artificially
as a way of getting therapeutic agents into the brain. Researchers in the
US have found that raising the osmotic pressure in the blood for brief periods
can temporarily make the barrier more permeable, apparently by ‘unzipping’
the tight junctions between capillary endothelial cells. But such an assault
on the barrier leaves the brain temporarily unprotected from harmful substances
or fluctuating concentrations of salts and so on that are circulating in
the blood. Another problem is that the entire barrier throughout the brain
is opened, so drugs cannot be targeted to a particular area.

An alternative is to bypass the barrier by injecting the drug into the
fluid in the brain cavities. This method is commonly used for treating brain
tumours because it delivers the medicine directly to where it is needed.
Alzheimer’s disease has been treated by injecting a drug, bethanecol, into
the brain ventricles. And more recently, a constant infusion of the drug
has been made possible by implanting a pump into the patient’s ventricles.
But such invasive methods are both uncomfortable and somewhat risky; clearly,
opening the barrier osmotically or injecting into the ventricles are not
suitable for routine administration of drugs when the problem is chronic,
as in depression or degenerative diseases.

It is far more elegant to design drugs that interact with specific receptors
or transport systems on the endothelial cells, or to link drugs to molecules
– peptides, for example – that can cross the blood-brain barrier. Peptides
are broadly either transportable (that is, they have an affinity for a blood-brain
barrier transport system or receptor as do glucose, amino acids, insulin
or transferrin) or nontransportable, such as the morphine-like painkillers
that act on the central nervous system, the enkephalins and beta-endorphin.
A ‘chimeric’ peptide can be formed by coupling a nontransportable peptide
to a transportable carrier molecule. Once the chimera reaches the brain,
the link between them is cleaved, the carrier molecule can leave the brain
again and the therapeutic substance remains in the brain, where it is needed.
William Pardridge and his colleagues at the UCLA School of Medicine in California
have recently shown that in isolated brain capillaries from rats, such a
chimera formed by coupling beta-endorphin to a positively charged albumin
molecule is taken up into the endothelial cells; so far, however, no one
knows whether beta-endorphin can escape from the endothelial cells into
the brain.

A useful way to study both the origins of the barrier and how it functions
once fully formed, as well as the mechanisms and routes by which compounds
can cross it, is to grow an artificial blood-brain barrier in the laboratory.
Joan Abbott’s research group at King’s College in London is one of several
that have refined ways of growing endothelial cells from brain capillaries
in plastic dishes to provide cultures that display some of the essential
characteristics of the barrier: the cells contain few pinocytotic vesicles
and are bound together by tight junctions; these membranes are impermeable
enough to prevent larger molecules from diffusing through.

At a recent meeting on glial-neuronal interactions, at Cambridge, Werner
Risau of the Max Planck Institute for Psychiatry near Munich described how
endothelial cells form a much tighter barrier when they are grown on a permeable
membrane that has astrocytes on the other side. This provides more evidence
for the idea that astrocytes produce molecules essential for maintaining
the special nature of brain endothelial cells. Pharmaceuticals companies
could use cultures such as these to make a preliminary assessment of a drug’s
ability to cross the blood-brain barrier, avoiding animal tests.

Another approach has been used by a group at Athena Neurosciences, a
biotechnology company in San Francisco. At the Cambridge meeting, Lee Rubin
reported how he and his colleagues are finding ways to manipulate the permeability
of the barrier in cultures of endothelial cells. Instead of using astrocytes
themselves, they bathe the endothelial cell cultures with fluid containing
substances produced by astrocytes. But these alone are not enough to get
a really tight barrier. They also have to add chemicals that increase the
level of the signalling molecule cyclic AMP in the endothelial cells. As
the level of cyclic AMP goes up, the barrier gets tighter and the cells
behave more like endothelial cells from the brain.

Cyclic AMP levels are normally regulated by substances that attach to
receptor molecules on cell membranes. Rubin’s group does not yet know what
naturally regulates the cyclic AMP in endothelial cells but they are testing
a range of candidates. Using some of these on animals, they have been able
to open the barrier for brief periods. But such substances are likely to
have effects on other systems as well as the blood-brain barrier, so they
are at present too crude to use for administering drugs to the brain. Instead,
Rubin sees more potential for developing ways of closing the barrier when
it is damaged by trauma, stroke or disease.

The blood-brain barrier is no longer seen as a passive structure, but
as a dynamic interface between the blood and the brain. Yet out understanding
of its maintenance and transport mechanisms is still incomplete. There is
much to be learnt before we can manipulate the barrier well enough to turn
it into a reliable route for getting useful drugs to selected areas of the
brain. Meanwhile, much is being learnt about how the barrier maintains its
integrity, which may place tools for repairing leaky barriers in our hands.

Nerina Ramlakhan has just completed a PhD in the Department of Physiology
at King’s College London. Dr Jennifer Altman is a science writer and researcher
in the neurosciences.