ÐÓ°ÉÔ­´´

Neurological diseases are often named after the physicians who first
described them. In 1817, for instance, the British physician James Parkinson
reported the physical tremor we now know as Parkinson’s disease. In 1872,
George Huntington, an American neurologist, described a fatal hereditary
disorder – ‘Huntington’s chorea’ – that strikes middle-aged people, causing
dementia. And in 1907, the German physician Alois Alzheimer presented his
discovery of protein deposits in the brain of a 51-year-old demented woman
– a case of Alzheimer’s disease.

One exception to this trend is the illness now known as Lou Gehrig’s
disease: it is unlikely that Lou Gehrig was interested in the workings of
the brain. Instead his prowess was on the baseball field. During the 1920s
and 1930s he played for the New York Yankees, setting numerous records and
earning a place in baseball’s hall of fame. Early in 1939, however, his
team-mates noticed a sudden deterioration in his skills. He could not move
as freely as before, or strike the ball with his customary power and timing
– a form of paralysis had set in. Later that year Gehrig had to quit the
game. On July 4, he bade an emotional farewell to thousands of New York
fans, and within two years he was dead.

The illness named after him is in fact motor neuron disease, or amyotrophic
lateral sclerosis (ALS), a rare disorder that affects between 1 and 2 people
in 100 000 and for which there is no known treatment. In Britain it has
struck other well known people. The actor David Niven died of ALS in 1983,
and Stephen Hawking, the theoretical physicist and best-selling author of
A Brief History of Time, has suffered from a milder form of ALS for almost
30 years. He has been confined to a wheelchair for over half that period.

Behind their stories is a medical mystery that has baffled geneticists,
epidemiologists and physicians for more than 50 years. While about 10 per
cent of cases are definitely genetic in origin, the majority appear to happen
by chance; and it is these that lie at the heart of the biological puzzle.
Theories abound as to the causative agent, or indeed agents. Chemical toxins,
viruses and metal ions all have their advocates, but firm links between
any one candidate and the disease are lacking.

Today the search for clues continues on two fronts. Epidemiologists
want to explain the apparently random incidence of ALS in every population,
while geneticists are focusing on the smaller number of hereditary cases.
The hope is that identifying the defective gene, or genes, in these families
will prove a watershed in ALS research. As with Alzheimer’s disease, which
can also run in families, geneticists suspect that hereditary ALS holds
the key to understanding the sporadic disease. The recent news that a team
in the US has located an ALS gene means this suspicion could soon be put
to the test.

Like Parkinson’s disease, ALS damages the nervous system in a curiously
selective manner. Only one class of cells are destroyed, the brain’s motor
neurons. These provide the electrical signals that pass down the nerve fibres
connecting the body’s brain and muscles. As the disease progresses, the
neurons degenerate, losing the ability to signal muscles. Patients generally
lose dexterity and have difficulty speaking and swallowing. Their muscles
waste away and death results from the loss of key nerves at the top of the
spinal cord. Multiple sclerosis causes paralysis in a different way, by
attacking the myelin sheath that insulates nerve fibres. In the case of
hereditary ALS, there seem to be three forms of the disease which vary in
severity.

Recent successes in mapping the genes for common diseases such as cystic
fibrosis as well as neurological disorders such as Alzheimer’s disease have
been based on one approach: examining the affected families for the inheritance
of special DNA markers, referred to as restriction fragment length polymorphisms
(RFLPs – pronounced ‘rifleps’). A RFLP is not usually part of the target
gene but can act as a flag marking its close presence on the chromosome.
Any RFLP that is frequently inherited with ALS should be a close neighbour
of – or in genetics parlance, show linkage to – the disease gene. The aim
is to find such a RFLP, trace it to a specific chromosome (of which there
are 23 pairs in humans) and then look for the gene itself on the same chromosome.

After an exhaustive search lasting more than four years, a team led
by Teepu Siddique, a neurologist at Northwestern University in Chicago,
Illinois, has finally met with success. Unlike the hunt for the gene for
Huntington’s disease, where scientists were lucky enough to find linkage
to one of the first dozen RFLPs tested , Siddique’s team had to analyse
over 100 RFLPs scattered across many chromosomes. By any standards this
was a mammoth task: samples of DNA had to be collected from 150 families
with ALS.

The first breakthrough came in 1989 when Siddique and his colleagues
published a so-called ‘exclusion map’. As its name suggests, this showed
the regions of DNA that could be safely ruled out from containing the gene.
More important, it offered a glimmer of hope that an ALS gene may lie on
one of two chromosomes. One of these, chromosome 21, was a particularly
promising candidate. Its link to Down’s syndrome was well established, and
other research groups had shown that it contains the gene for at least one
hereditary form of Alzheimer’s disease . So Siddique’s team decided to focus
exclusively on this chromosome. They poured their energies into examining
the inheritance pattern of four markers from chromosome 21 in 23 families
with ALS.

Earlier this year the gamble paid off. In May, the researchers published
a paper in the New England Journal of Medicine showing that chromosome 21
does indeed contain an ALS gene. But the study also showed that this gene
does not cause the disease in all affected families. The implicaton? That
there must be at least one other gene which, when defective, causes ALS.
Given that clinicians have diagnosed at least three different forms of hereditary
ALS, this is perhaps not surprising. It does, however, serve as a reminder
of the underlying complexities of most genetic diseases. The notion of ‘one
gene, one disease’ is usually a gross simplification.

So what next? As yet Siddique and his colleagues have no idea of the
function of the ALS gene they have mapped. The first inkling of this will
come when the gene, or at least part of it, is sequenced. Nor do they have
more than a rough idea of where the gene is on the chromosome. Before they
can home in on it, they will need to generate a more accurate map of the
chromosome. So the first priority is to find better RFLP markers. Only then
can they move onto the final stage of searching through cloned chromosomal
fragments for the correct gene. The strategy follows closely that used to
such great effect in identifying the cystic fibrosis gene: First map the
gene, then find closer markers, then narrow the location of the gene down
to one particular segment of the chromosome, and finally sequence the DNA
of this segment.

Yet identifying disease genes even after they have been mapped is far
from easy. It took four years, for example, to isolate the cystic fibrosis
gene. And the search for the Huntington’s gene continues today, more than
eight years after researchers first reported the linkage to an RFLP. For
Lou Gehrig’s disease it will not be a formality either. In baseball parlance,
the disease may have received ‘strike one’ but the game is far from over.

Eventually, though, the gene will be characterised. The exciting question
then will be what light, if any, it can shed on the more common, sporadic
form of the disease. In contrast to the hereditary form, sporadic ALS is
almost as much a mystery today as it was 50 years ago. Although the disease
is still extremely rare, some figures suggest that, like Alzheimer’s disease,
it could be on the increase. A recent report put the current incidence as
high as 1 in 20 000. Although this could reflect improved diagnosis, Christopher
Martyn, of the Medical Research Council’s Environmental Epidemiology Unit
in Southampton, thinks we should take the trend seriously. He says that
similar increases in ALS have been reported in both Britain and Scandinavia,
whereas the incidence of Parkinson’s disease has remained stable. ‘If that’s
so, then something interesting is happening.’

One of the most publicised theories for the origins of sporadic ALS
implicates poliovirus. In 1988, a paper in The Lancet reported that the
prevailing distribution of ALS in England and Wales mimicked that of polio
40 years previously. Poliovirus damages the same motor neurons in the brain
stem and spinal cord as ALS does.

Proponents of this theory, who include Martyn argue that infections
that are below the threshold for causing polio may predispose an individual
to ALS. Some investigators urge caution, however, because ALS can resemble
post-polio syndrome, a different disease altogether.

And one conceptual drawback is that poliomyelitis is an acute, short-lasting
infection, whereas ALS is a chronic disease. Even so, the theory is seen
by many as one of the most attractive advanced so far.

Alternative theories focus on chemical toxins. Despite a plethora of
published reports associating sporadic ALS with one environmental toxin
or another, a convincing candidate has yet to emerge. Some epidemiologists
blame emissions of sulphur-containing toxins. The people at risk of contracting
ALS, they argue, are those with an impaired ability to metabolise these
compounds in the liver. Others point to chemicals of a different sort, for
example the solvents and glues used in the leather industry, and even lead.
Perhaps different substances cause the disease in different parts of the
world. Perhaps a similar array of toxins cause neuronal degeneration in
other motor-system diseases, such as Parkinson’s and Alzheimer’s diseases.

Behind these unresolved questions lurks another worry. How can an apparently
simple chemical cause the kind of selective damage to neurons seen in diseases
such as ALS? For that matter, how can a chemical reproduce the complex effects
on the body of a genetic defect? Some clues, at least, come from studies
of Parkinson’s disease.

In the early 1980s, US researchers discovered by chance that a chemical
called MPTP can produce Parkinsonian tremors. The alarm bell rang when heroin
users began showing up at a clinic in San Francisco, apparently, with Parkinson’s
disease; MPTP happens to be a by-product of the chemical synthesis of heroin.

MPTP, and in particular its use on rats, opened up a new line of research
on the biochemical events that kill dopaminergic neurons, the neurons lost
in Parkinson’s disease. Among other things, it appears that the body metabolises
MPTP, generating a chemical similar to dopamine which is then taken up selectively
by dopaminergic neurons.

The discovery of the effects of MPTP did much to heighten awareness
among scientists of the possible links between toxins and degenerative brain
diseases. Yet, even now, they remain split on the key question: do chemicals
such as MPTP actually cause disease in the world at large, or are they merely
useful research tools? Further clues come from a different quarter. At St
Mary’s Hospital Medical in London John Hardy and his colleagues are studying
the molecular genetics of Alzheimer’s disease . They have pinned down a
genetic defect that is common to five families with one form of the disease.
The defect is in a gene encoding a protein known as APP. Hardy and his colleagues
believe, but have yet to prove, that the defect alters APP’s response to
metabolising enzymes.

An agent, chemical or otherwise, that interfered with the same metabolic
steps could, at least in theory, produce the same end point – in this case,
a gradual build-up of incorrectly processed protein in the brain. The same
kind of biochemical model could hold for other genetic diseases with sporadic
counterparts.

The main frustration in studying sporadic ALS is its epidemiology. Its
incidence is roughly the same all over the world, unlike most diseases which
tend to afflict certain ethnic or racial groups more than others because
of variations in genetics or environmental factors. Fortunately, however,
there have been some notable exceptions. In the 1950s and 1960s, a dramatic
outbreak of ALS on a remote island in the Pacific Ocean offered what was
probably the best chance of finding a cause.

Guam is the largest and southernmost of the Mariana islands in the northern
Pacific. It lies to the east of the Philippines and southeast of Japan.
The islanders, whose population totals over 100 000, are called Chamorros.

Today the island is a rather commercialised tourist resort. But it looked
very different when Harry Zimmerman, an expert neuropathologist assigned
there shortly after US forces reclaimed the island from the Japanese in
1944, first recognised a serious malady among the population. As compellingly
described by Terence Monmaney in The New Yorker (October 29, 1990), Zimmerman
performed numerous autopsies and encountered a handful of cases with full-blown
symptoms of ALS. Although totalling less than 10, the number was still far
greater than expected for the small number of inhabitants on the island.
Follow-up studies showed that the incidence of ALS among Chamorros was almost
100 times greater than anywhere else in the world.

In Umatac, a remote village in the southwestern corner of the island,
over one quarter of adult deaths were thought to result from ALS. Even more
startling was the identification of an illness of greater severity which
was almost as common as ALS – indeed it sometimes occurred in the same family,
even the same individual. The second illness seemed to combine the physical
tremor of Parkinson’s disease with the dementia of Alzheimer’s disease.
Moreover, the illness’s characteristics – deteriorating neurons, shrunken
brains and protein deposits called neurofibrillary tangles – suggested a
possible connection between the neurological diseases on Guam and Alzheimer’s
disease, even a common causative agent.

Certain candidates for the cause of Guam’s affliction were ruled out
at an early stage. Despite its prevalence, the disease did not seem to be
genetic, nor did it seem to be due to a transmissible agent such as a virus.
The most likely possibility was proposed in the early 1960s by an epidemiologist
called Majorie Grant Whiting. She suggested that ALS on Guam was somehow
connected to the native’s habit of eating cycad seeds of the false sago
palm. Guamanians commonly ground the seeds into a flour called fadang, which
they used to make tortillas. Before doing so, they dried the seeds and then
washed them with many changes of water over a week or more to remove potentially
dangerous toxins. The smell of cycad seeds alone can make people ill, and
animals have died from drinking the water used to soak them.

Link with food shortage

Most Guamanians were fully aware of the likelihood of illness from eating
unwashed seeds. However, Whiting and Leonard Kurland, an epidemiologist
at the Mayo Clinic in Minnesota, argued that the shortage of food, especially
rice, during the war may have led many people to eat unwashed seeds in desperation.
And this, they suggested, could have led to the explosion of ALS on the
island.

Several researchers, including Daniel Gajdusek, who won a Nobel prize
in 1976 for work on the brain disease known as kuru, investigated the seed
theory thoroughly. Yet by the early 1970s it had been dropped for lack of
firm evidence. Why? On the plus side, cycad seeds contain several known
toxins, including &bgr;-methylamino-L-alanine (BMAA), a rare amino acid,
and cycasin, which is carcinogenic in animals. But all attempts to use BMAA
to induce ALS-like symptoms in laboratory rats failed, casting doubt on
the idea that it causes ALS in humans. As time passed, positive results,
even striking ones, were gradually forgotten. A notable example was a paper
published in 1964 describing neural degeneration and muscle wasting in a
rhesus monkey fed cycad flour.

Then, in the early 1980s, the theory’s fortunes changed. A British biochemist
called Peter Spencer, now at the Oregon Health Sciences University, Oregon,
noted an important similarity between ALS and a motor neuron disease called
lathyrism. Lathyrism, which is common in poor communities in Ethiopia, India
and Bangladesh, is caused by a toxin in the grass pea called &bgr;-N-oxalylamino-L-alanine
(BOAA), a close relative of BMAA. In a previous study, Spencer had found
that monkeys, when fed BOAA, succumb to a moderate form of the disease.
But, and most significantly for Spencer, his team also discovered that the
lathyrism cannot be induced in laboratory mice or rats. Perhaps the earlier
experiments on BMAA had been wrongly interpreted after all.

In July 1987, Spencer’s team reported the results of experiments involving
oral administration of BMAA to macaques (New ÐÓ°ÉÔ­´´, Science, 13 August
1987). The paper described how a dose regimen spanning several weeks had
led to a weakening of the monkeys’ muscles, considerable ageing and signs
of brain degeneration. The symptoms, though much milder than those of full-blown
ALS, suggested that BMAA might act as a slow toxin. Coupled with the normal
loss of neurons due to advancing age, reasoned Spencer, BMAA might be just
potent enough to cause ALS in humans.

Equally, the severity of the disease might depend on the amount of toxin
ingested. The Parkinsonian tremors and dementia in some Guamanian patients
might result from a particularly high intake of BMAA. Spencer also noted
that some of the BMAA-induced effects in primates were similar to those
caused by MPTP, the amino acid that induces Parkinsonian tremors. Perhaps
these chemicals worked in a similar way.

However, other scientists hold very different views about the prevalence
of ALS on Guam, largely because no one has been able to fully reproduce
Spencer’s findings. ‘It’s been very disappointing that the BMAA result has
not been replicated,’ admits Kurland. Spencer’s attention has now shifted
towards another seed chemical, cycasin.

Gajdusek believes that the key lies elsewhere, with a mineral imbalance
– specifically, decreased calcium and magnesium uptake coupled with excess
aluminium. As evidence, he cites the fact that researchers have detected
abnormally high levels of aluminium in the brains of dead ALS patients,
just as they have in Alzheimer’s patients. The protein deposits seen are
also strikingly reminiscent of those found in the brains of Alzheimer’s
patients. But – again, as with Alzheimer’s disease – the evidence for a
causal link between ALS and minerals such as aluminium is only circumstantial.

The Guam riddle remains intractable. The publicity afforded cycad seeds
has alerted Guamanians to the possible danger of eating them, and the incidence
of ALS on Guam is now much less than in the 1940s and 1950s. But many Guamanians
still swear by fadang, claiming to have eaten it for years with no ill effects.
And, according to Kurland, the incidence of the Parkinsonism-dementia complex,
is growing.

Perhaps the assumption that Guam’s neurological diseases have no genetic
basis is wrong after all. So far no one has uncovered any evidence to support
such a contention, but the necessary tools have been lacking. One thing
is certain: the recent of discovery the ALS gene on chromosome 21 provides
the means to test the notion of a genetic link more thoroughly. Siddique
thinks this may be a worthwhile exercise. He is contemplating a comparative
study of genetic markers from the people of Guam.

* * *

Neurological disorders and the new genetics

Lou Gehrig’s disease, or ALS, is one of a growing number of neurological
disorders that are slowly yielding their secrets to molecular genetics.

The first breakthrough came in 1983, when Jim Gusella’s group at the
Massachusetts General Hospital in Boston mapped the gene for Huntington’s
disease. Not only was the gene much sought after in its own right, but the
study was also the first successful attempt to localise an autosomal disease
gene on any chromosome. Gusella and his team looked at just 12 markers scattered
throughout the DNA, and found that one of them, called G8, was close to
the gene, which they thereby assigned to chromosome 4.

Sadly, progress towards the final goal of isolating the abnormal gene
itself has been tantalisingly slow. One problem is that the gene appears
to lie close to the very end of the short arm of the chromosome, where the
DNA is particularly prone to shuffling its structure.

But the news is not all bad. Researchers have isolated large chunks
of the DNA from this region. And recently Gusella’s team and a group led
by Hans Lehrach of the Imperial Cancer Research Fund in London announced
that they have pinned down the position of the gene to within 2.5 million
bases, a stretch of DNA representing roughly 3 per cent of the whole chromosome.

Elsewhere the search for genes linked to Alzheimer’s disease continues
to gather momentum. Although most cases of the disease are sporadic, as
many as 20 per cent run in families. Geneticists has been mainly interested
in those who succumb to the disorder at a relatively young age, often in
their forties. In 1987, teams from the US and Europe, led by Gusella, studied
genetic markers in four large families with this ‘early onset’ disease,
assigning the disease gene to chromosome 21.

Last year John Hardy of St Mary’s Hospital in London collaborated with
Peter St George-Hyslop, now at the University of Toronto in Canada, to compare
the inheritance of genetic markers in families with both early-onset and
late-onset Alzheimer’s disease. According to their results, the two conditions
are caused by completely different genetic defects. This conclusion was
bolstered by findings from Allen Roses and colleagues at Duke University
in North Carolina, which point to a defective gene on chromosome 19 as the
cause of the late-onset disease.

John Hardy and his team took another major step forward earlier this
year. They announced in Nature that they had pinpointed a genetic mutation
on chromosome 21 that causes Alzheimer’s disease in two families. This result
has now been confirmed in three other families with the disease. The mutation
lies in a gene encoding the amyloid precursor protein (APP), which researchers
have long suspected has some role in the disease.

Hardy’s discovery has deepened the intrigue surrounding APP: how does
a tiny defect in its chemical structure contribute to the formation of protein
deposits? Further evidence implicating APP came last month from experiments
on transgenic mice. Two US groups reported that mice engineered so as to
produce extra APP develop protein deposits similar to those seen in the
brains of Alzheimer’s patients.

ALS has sometimes been confused with spinal muscular atrophy (SMA),
another disease that affects motor neurons. Patients show loss of reflexes
and wasting, but, unlike ALS, SMA is entirely a gen-etic disorder.

Last year teams led by Kay Davies of the University of Oxford and Conrad
Gilliam of Columbia University in New York discovered that a number of clinically
distinct forms of SMA are caused by a defect in the same gene on chromsome
5. Efforts are well underway to find the faulty gene.

Kevin Davies is an assistant editor of Nature based in Washington DC