ÐÓ°ÉÔ­´´

IF A BARNACLE could tell you the critical point in its career that determined
its success, it would most likely recall the moment when, as a free-swimming
larva, it finally attached itself to the bottom of a ship’s hull. From that
point on, it was free from the attention of predators and the distraction
of the tides. Like barnacles, bacteria also like to stick to solid surfaces.
Any object dipped into the sea will rapidly become coated with bacteria.
Once the bacteria are attached, the water cannot sweep them away to an unpredictable
fate.

Bacteria also like to stick to the human body. Soon after a baby is
born, bacteria adhere to the surfaces of its body, including the skin and
the membranes lining the nose and the gut. The bacteria remain in variable
numbers and differing proportions of species. In the gut, for example, many
of the bacteria are Escherichia coli. They help in digestion and keep disease-causing
bacteria at bay. This symbiotic relationship lasts until death.

Occasionally, the arrival of harmful bacteria upsets this symbiosis.
The invading organisms adhere to and colonise body surfaces, including parts
of the body that are normally sterile, such as the urinary tract and the
lower airways of the lung. Most researchers assume that any disease-causing
bacterium moving from the outside of the body to the inside must first make
contact with a surface. Over the past 10 years, biomedical scientists have
been examining this process.

Drugs that kill or inhibit microbes dominate current strategies aimed
at combating bacterial diseases. This sledgehammer approach has several
shortcomings. For example, drugs may be toxic: one very useful antibiotic
(chloramphenicol) will, on rare occasions, destroy the patient’s bone marrow.
Microbes can also become resistant to the action of certain drugs. An alternative
and more subtle strategy might be to interfere with the ability of pathogens
to gain a hold on the body in the first place.

The surfaces of human cells have a net negative electric charge. So
do those of bacteria. Researchers believe that water-resisting (hydrophobic)
molecules present on both surfaces partially overcome the resulting electrostatic
repulsion, in much the same way that the barnacle’s sucker excludes water
from the interface between the mollusc and the boat. This is not the whole
story. If it were, any bacterium could stick to any part of the human body,
which is not the case. Bacteria prefer to colonise or infect particular
tissues. Researchers call this phenomenon ’tissue tropism’. They can partly
explain it by factors including the local pH, the temperature and the host’s
secretion of antibacterial chemicals, antibodies and mucus; only a few bacteria
will have adapted themselves to overcome these lines of defence. Bacteria
can also stick to particular tissues by producing molecules, called adhesins,
that project through their own cell walls and interact with complementary
receptors on the surface of the host’s cells. Adhesins are usually made
of proteins. The receptors on the host’s cells, however, are complex molecules
of protein or lipid that contain carbohydrates, called glycoproteins or
glycolipids, respectively.

Ed Beachey, professor of medicine at the University of Tennessee College
of Medicine, in Memphis, has pioneered work on Streptococcus pyogenes, the
bacterium that commonly causes tonsillitis (it may also produce more serious
consequences including kidney and heart disease). Electron microscopy of
streptococci reveals many short stubby projections, called fibrillae, on
the surface of the bacteria. The fibrillae are made up of proteins and probably
help the bacteria to attach to surfaces. Beachey and his colleagues have
found that the glycolipid called lipoteichoic acid, which is present in
the cell walls of many streptococci, becomes associated with positively
charged regions of the proteins in the fibrillae. As a result, the lipid
end of the lipoteichoic acid molecule juts out of the bacterium. These highly
reactive sections of lipid may then react with lipids within the walls of
cells lining the throat, firmly anchoring the bacterium.

Once attached, bacteria can proliferate in a relatively stable environment.
Occasionally, they may break out of the throat and invade the bloodstream,
a move which confers enormous advantages in terms of nutrients and room
to multiply in. This spreads the infection around the body but, at the same
time, exposes the bacteria to the host’s white blood cells. It is an apparent
paradox that the same fibrillae that enable the streptococci to cling on
to the throat should also help the white blood cells known as phagocytes
to attach to the bacteria and destroy them. Beachey’s team showed that proteins
in the blood, such as fibrinogen, coat the fibrillae and may have a role
in hiding the adhesins from the white blood cells: a classic example of
a wolf in sheep’s clothing.

* * *

A hike across the groin

Researchers have now described similar interactions between adhesins
and cell receptors for a variety of disease-causing microbes, including
E. coli, which causes 80 per cent of urinary infections. Such infections
are particularly common in women, in whom the urethra, the tube that leads
from the bladder to the outside, is short. E. coli is normally carried harmlessly
in the gut, but can migrate across the skin of the groin to the urinary
tract, where it is certainly not harmless.

During the day, frequent urination discourages any bacteria within the
bladder from gaining a foothold. At night, when someone is asleep, urine
in the bladder is stagnant, allowing the balance to tip in favour of the
bacteria. This effect is especially apparent in people with neurological
disease of the bladder – after a spinal injury, for example – in whom the
flow of urine is obstructed. But some women with apparently normal bladder
function still have frequent urinary infections, a puzzling phenomenon until
relatively recently. The term ‘cystitis’ describes inflammation of the bladder
lining, leading to symptoms of frequent urination and pain on passing urine.
Although a bacterial infection is often to blame, it is not always possible
to isolate bacteria from people who suffer repeated attacks.

E. coli attaches to the surface of the bladder wall by adhesin proteins
that are usually located on the tips of hair-like structures called fimbriae.
These are much longer than fibrillae and project several micrometres from
the surface, in a fashion that resembles the bristles of a toilet brush.
Fimbriae are common features of those bacteria that colonise the gut and
the nose. Each bacterium may possess up to 200 fimbriae. So it is not surprising
that fimbriated forms of E. coli are good at sticking to the bladder’s surface
– when E. coli manages to infect the bladder, its fimbriated forms are usually
to blame.

Researchers have divided fimbriae into two major classes on the basis
of the receptors on the cells they stick to. AllE. coli infecting the urine
produce type 1 fimbriae. If you isolate and mix type 1 fimbriae withered
blood cells, the red cells clump together, or agglutinate, because the fimbriae
glue them together. If you first add mannose, a common carbohydrate in the
cell walls of the urinary tract, to the red cells you prevent the cells
from agglutinating, presumably because mannose competes with the red cells
for the bacteria’s adhesins. So researchers call type 1 fimbriae ‘mannose
sensitive’, and they think that these fimbriae help bacteria to stick to
those glycoproteins in the walls of bladder cells that contain mannose.

The lining of the bladder probably deals with these unwelcome advances
by secreting a gel-like substance, called mucoprotein, into the urine. The
mucoprotein includes Tamm-Horsfall protein, or THP, a glycoprotein that
contains mannose. This sticks to type 1 fimbriae and researchers believe
that it escorts E. coli out of the bladder during urination. The tempting
conclusion is that women who are prone to urinary infections either have
more receptors for bacteria in their bladder or secrete smaller amounts
of Tamm-Horsfall protein. Evidence favours the first possibility: isolated
bladder cells from women prone to infection bind more E. coli than do cells
from healthy women.

The composition of the carbohydrates that help to make up the glycoproteins
and glycolipids on the bladder’s surface is extremely variable. One influential
factor is the individual’s blood group, determined by proteins on the surface
of red blood cells, called the blood group antigens. Some of these proteins,
such as A, B, H, Lewis a and Lewis b, contain carbohydrates. Each individual
has specific genes that code for enzymes whose task it is to transfer chunks
of carbohydrates in specific sequences onto the cell’s surface and thereby
determine the blood group. The genes also determine the expression of the
same antigens on the surface of the cells lining the bladder, as well as
their secretion in body fluids such as saliva and, possibly, urine. One
of these genes, called the secretor gene, makes the lining of the bladder
rich in A, B, H and Lewis b antigens. Almost one in four women in Britain
possess the secretor gene, and could be particularly prone to urinary infections
as a result.

Researchers have also identified a second major class of fimbriae. These
are defined by their ability to stick to glycolipids that contain a particular
disaccharide, called a-galactosyl-1,4-b-galactose, or gal-gal for short.
These glycolipids are found in cells lining the urinary tract; they are
also components of the human blood group antigen known as P. This second
class of fimbriae, the P fimbriae, is found in roughly two-thirds of E.
coli infecting the bladder. These strains of bacteria are particularly likely
to produce infections further up the urinary tract, within the kidney, a
condition called pyelonephritis.

Unexpectedly, whether E. coli with P fimbriae stick on the cells lining
the bladder, and start a urinary infection, seems not to be related to the
amount of gal-gal on those cells. But studies have shown that E. coli with
P fimbriae are more likely to infect the bladder cells of women without
the gene than those of women with the gene. Researchers at the Memorial
Sloan-Kettering Center in New York have recently confirmed that this holds
true epidemiologically: women with recurrent urinary infections are more
likely to be nonsecretors (New England Journal of Medicine, 23 March, p
773). One likely explanation for this phenomenon is that, in women who are
secretors, blood group antigens may project from the lining of the bladder,
shielding gal-gal receptor sites from P fimbriae, in much the same way as
the treetops in a jungle hide the underlying shrubs from overhead view.
By contrast, the gal-gal receptors of women who are nonsecretors are relatively
exposed, making it easier for bacteria to attach themselves and cause an
infection.

Gary Schoolnik, who works at Stanford University in California, recently
outlined a three-pronged approach to treating urinary infections, without
using antibiotics (New England Journal of Medicine, 23 March, p 804). First,
doctors could assess a woman’s risk of infection by finding out whether
she is a secretor or by measuring how well E. coli adhere to the cells lining
the bladder, which are constantly shed in the urine. Secondly, it might
eventually be possible to immunise women prone to repeated infections by
injecting them with fragments of type 1 or P fimbriae. A third strategy
might be to make molecules resembling the receptors that type 1 and P fimbriae
stick to, in order to compete with (and so inhibit) bacteria trying to bind
to the surface of the urinary tract.

Work published recently (The Lancet, 10 December 1988,p 1327) suggests
that the process by which E. coli sticks to cells may be even more complex
than that outlined above. Led by Ten Feizi, members of the Medical Research
Council’s Glycoconjugate Research Section at Harrow in Middlesex revealed
additional mechanisms. They extracted short chains of sugars rich in lactose
from human milk (a convenient source of these molecules) and attached them
to segments of lipid manufactured in the laboratory. These combination molecules
bound to E. coli, including strains of the bacteria that had no fimbriae.
The results of this experiment suggest that E. coli might be binding to
an alternative carbohydrate sequence in the body, one containing lactose
rather than mannose or galactose, found deep within the membrane of the
host cell. Although this receptor is not normally exposed on the cells lining
the bladder, it might show itself if the lining is damaged for some reason.
There may well be more hitherto-unknown relationships between adhesins and
receptors. If, as seems likely, it turns out that bacteria have many different
ways of sticking to cells, then it could be a depressingly long time before
appropriately designed drugs can interfere with these molecular mechanisms.

There are other reasons for believing that such elegant solutions to
infectious disease, while ideal, are still remote. In the case of the respiratory
tract, the host defends itself against invasion with a repertoire of barriers,
all of which potential pathogens must overcome. At the Host Defence Unit
of the National Heart and Lung Institute in London, we have been investigating
how these infectious bacteria survive within the airways. We are discovering
that the attachment of bacteria to the tissue lining the airways may be
merely the end point in a series of steps that leads to infection.

Most of the cells lining the airway from the nose, almost all the way
down to the alveoli (air sacs) in the lungs, have tiny hair-like structures
called cilia projecting from their surfaces. The cilia sway back and forth
in a whip-like action up to 17 times a second. The cells on the surface
secrete mucus into the airway, which is swept upwards by synchronous beating
of the carpet of cilia. Any bacteria entering the airway become caught up
in the mucus and carried to the throat, where they are swallowed or coughed
out. The mucus is not inert. It contains lysozyme, an enzyme that kills
bacteria, and lactoferrin, which prevents bacteria from acquiring iron,
an essential nutrient. In addition, the body secretes specific antibodies
into the mucus.

Another barrier sits on the surface of the airways – a thin layer of
a gelatinous mucoid material called glycocalyx. This layer, rich in carbohydrates,
may conceal receptors on the surface to which bacteria may be able to attach.
We know that bacteria harmful to the lung do eventually penetrate the surface,
but there is a debate about the critical step in invasion. Does the pathogen
simply enter the airway and attach to the surface of normal cells, or does
it proliferate within the airway and then, somehow, perturb the defences
that normally conspire to prevent its attachment? In the past, scientists
thought that the bacteria simply entered and latched on. They took cells
scraped from the surface of the mouth, nose or throat, mixed them with bacteria
and looked for characteristics common to those bacteria that managed to
attach to the cells. They found that many of those that succeeded had fimbriae.
The fimbriated bacteria that attack the lungs include Pseudomonas aeruginosa,
an important cause of infection in patients with cystic fibrosis, and Haemophilus
influenzae, a frequent cause of pneumonia in developing countries, which
may also be responsible for bouts of chronic bronchitis in industrialised
countries. But the experiments have mainly been with isolated cells, ignoringthe
formidable barrier of cilia and mucus that normally covers the intact tissues
of the airways.

Peter Cole, professor of respiratory medicine at the National Heart
and Lung Institute, along with colleagues in the Host Defence Unit, has
developed a technique of taking strips of cells from the lining of the nose
and, using polarised light, measuring how fast the cilia beat. Robert Wilson,
lecturer in medicine at the same unit, used this method to examine whether
bacteria affected the rate at which the cilia beat. He discovered that P.
aeruginosa secrete a substance that slows the cilia down, and further work
showed that H. influenzae has the same effect.

The cell walls of bacteria contain a large molecule called lipopolysaccharide
that bacteria release in large amounts as they die. When researchers added
this to the nasal lining, they found that the substance disrupted the cells,
causing them to be shed. If this happened in the airway, it would disrupt
the cilia and the layer of protective mucus. So some soluble products of
bacteria, including lipopolysaccharide and substances that inhibit cilia,
can weaken the airways’ defences, regardless, possibly, of whether any bacteria
actually attach themselves to the surface.

Over the past 18 months, in our laboratory, we have been growing H.
influenzae in the presence of a preparation of cells that line the nose.
Using the electron microscope, we can inspect the surface for bacteria that
have become attached. We have never seen H. influenzae attaching to normal
preparations. We have, however, witnessed the bacterium damaging them. It
can, for example, cause cells to become detached, and can damage their mitochondria.
Such damage could allow this microorganism to bypass the host’s defences
and invade tissue. The conclusion is that drugs targeted at preventing bacteria
from attaching themselves to cells lining the airways, at least, may be
unsuccessful. A more promising approach might be to develop drugs that impede
the bacteria’s ability to disrupt the structure and function of the lining
cells. Along with other groups, we are currently working out ways of interfering
with this process and hope, ultimately, that novel drugs to fight bacterial
infections may result.