It has become one of the most controversial and feared medical diagnoses of
modern times. Autism was barely spoken of a generation ago but it has been
forced into public consciousness by the row over the MMR vaccine and the
growing realisation that it is much more common than doctors had imagined.
The suggestion that the developmental disorder can be triggered by the MMR
vaccine has been shown to be scientifically unfounded, but it prompted
thousands of parents to agonise over the cruel condition that seems to leave
children walled off in a social and emotional world of their own, apparently
beyond their love.
Their concerns have also been fed by reports of an autism epidemic. A disorder
that was once rare has become alarmingly common, with as many as one in 100
children now thought to be affected in some way.
Even if much of this is explained by better diagnosis, the condition retains a
brutal mystery. What is it that makes children who seem normal at birth
regress suddenly a year or two into life? Now a change in science’s ability
to decipher how genes influence health is promising to pin down what autism
owes to inheritance.
Within the next year a new study is expected to identify many of the genes
that underlie autism for the first time. At the same time, two new theories
are challenging established thinking about autism genetics in ways that
could ultimately transform diagnosis and treatment.
“The medics tell me we are at a tipping point,” said Dame Stephanie Shirley,
the millionaire computer entrepreneur and philanthropist, who is the
chairman of the research charity Autism Speaks and the mother of an autistic
son.
That genetics are the chief cause of autism has been known for three decades.
It was in 1977 that Professor Michael Rutter, of the Institute of Psychiatry
at King’s College London, published a twin study that transformed the
understanding of its origins.
Twin studies are one of the mainstays of genetics. Because identical twins
share all of their genes while fraternal twins share only half, and both
share broadly similar environments, comparisons can tease out the relative
contributions of nature and nurture.
Professor Rutter found that if an identical twin was autistic, it was highly
likely that the other twin was autistic too. Fraternal twins, however, were
no more likely to share the diagnosis than ordinary siblings. This made it
certain that genes played a large role and it is now thought that autism is
among the most heritable of all psychiatric disorders. Genetics account for
most of the variance and, although environmental factors matter too, they
are less important.
The condition, however, has remained a genetic paradox. For all the certainty
that genes are heavily involved, it has proved impossible to discover which
ones are guilty. In the 30 years since Professor Rutter’s study, hundreds of
genetic mutations that affect health have been found. Most are single-gene
disorders, where inheriting a rogue gene invariably means developing a
disease such as Hunting-ton’s, which affects the central nervous system.
Most of the others have involved very high risks: women with abnormal
variants of the BRCA1 gene, for example, have an 80 per cent risk of
developing breast cancer.
Autism does not work like that: the search for genes with such large effects
has failed. It might be influenced by dozens of genes, each of which raises
the risk by amounts too small to have been detected. Or it could be the
result of spontaneous mutations instead of more easily tracked defects that
are passed from generation to generation. Science does not yet know.
The scientific success story of 2007 has been the coming of age of a new
method of gene-hunting that can find the sort of genes with weak effects
that are thought to influence autism. These genome-wide association studies
compare the DNA of thousands of people who have a disease with healthy
controls, using tools called “gene chips” to screen the entire human genome
for hundreds of thousands of tiny genetic variations that differ between the
two groups.
In recent months, the technique has revealed scores of genes that subtly
influence common conditions such as diabetes, heart disease, breast cancer
and multiple sclerosis, often raising the risk by as little as 10 per cent.
Autism is the next target. The Autism Genome Project (AGP), an international
consortium that studies more than 1,000 families with at least two autistic
members, is about to apply the tool to its database.
“We have been waiting ten years for the technology to do this,” said Antho-ny
Monaco, of the University of Oxford, one of the project’s leaders. “We were
never likely to understand until we were able to screen very large numbers.
The probability has always been that autism is highly genetic, but highly
heterogeneous – that lots of different genes are involved. We now have a
great chance of picking them up.”
The AGP’s genome-wide association study is a classic example of win-win
science. Even if it draws a blank, it will still shed new light on the
genetic origins of the condition. No results would mean one of two things.
It could be that the effects of the genes responsible are even tinier than
suspected and bigger samples are needed. Or it could be that a radical new
theory of autism genetics is correct.
Professor Michael Wigler, of Cold Spring Harbour Laboratory in New York state,
believes that autism might be the result of single genes with big effects
after all. These mutations, however, are not quite the same as the inherited
ones that cause diseases such as Huntington’s.
According to his model, most cases of autism are caused by random, spontaneous
mutations in the sperm or eggs of parents that are passed on to individual
children. Most of these then develop the condition but some, particularly
girls, do not. They are somehow resistant and, although they carry a
potentially harmful mutation, they do not suffer its consequences.
This may explain why autism is an overwhelmingly male disorder, four times
more common among boys than girls. It fits with data showing that the
children of older parents are at higher risk: sporadic mutations of this
sort increase with age. It also points towards an intriguing explanation for
the existence of high-risk families with more than one autistic child.
Professor Wigler’s research suggests that in these families, a mutation
first occurred in one of the parents, usually the mother. While she was
immune, probably because of her gender, her sons were not so lucky: half of
them would be autistic, depending on whether they inherited the rogue gene.
“Sporadic autism is the more common form of the disease and even the inherited
form might derive from a mutation that occurred in a parent or grandparent,”
the professor said.
If mutations of this sort are responsible, they would not show up in the AGP:
they are new and unique to individuals and families, so will not surface
from large comparisons of DNA.
“That is one of the exciting things about our work,” Professor Monaco said.
“If we find genes, it is interesting and if we don’t find genes, it is
interesting too.”
What Professor Wigler’s theory does not account for is another aspect of new
thinking about autism: that it may not be a single disorder.
For autism to be diagnosed, children must meet three criteria: they must show
social impairment, communication difficulties and nonsocial problems such as
repetitive and restricted behaviour. Yet there is an emerging consensus that
these traits do not always go together and that there are people who meet
the criteria for one or two characteristics but who do not receive any
diagnosis. Autism, in short, may be the confluence of three separate
developmental conditions. Only when they occur together is the result
devastating.
Research by Angelica Ronald, Franc-esca Happé and Robert Plomin, of the
Institute of Psychiatry, has suggested that each of these three problems is
influenced by different sets of genes. The twin studies have shown that
while each trait is highly heritable, they do not often overlap.
“The label autism is something that was applied to a set of behaviours that
were first described in the 1940s,” said Dr Ronald, who is funded by Autism
Speaks. “It’s not necessarily a label for a clear biological entity and in
research it may be a misnomer to assume it’s one thing.”
This has important implications for gene-hunting. It could be that genes have
not been found because scientists have been treating autism as a whole. If
different genes affect the communication and social elements of the
disorder, finding them might involve looking at people who are not autistic,
but who have mild versions of one of the problems. “We need to tackle
whether we should look at autism as a single phenomenon, or whether it would
be better to look, for example, just at autistic social problems,” Dr Ronald
said.
Such an approach would also be valuable by shedding immediate light on what
any genes that are found actually do.
Dr Ronald added: “If we split up the symptoms, we can know that these genes
are going to be involved in social problems and those ones in nonsocial
problems. That is obviously going to be valuable when we look towards
diagnosis and treatment.”
An understanding of which genes are involved in which parts of autism should
help doctors to spot the condition earlier. It would also prepare parents
for the way their child is likely to develop and it could help with the
design of therapies.
Dame Stephanie is excited by the pace of change. “It is quite possible that in
five to ten years, we will have a real understanding of this disorder,” she
said. “That’s a timescale that means today’s children may be helped.”
Difficulties - and above-average intelligence
— Autism is a developmental disorder that first becomes apparent by the age of
3
— It is part of a group of disorders known as the autistic spectrum,
which include Asperger’s syndrome, a milder form of the condition
— 1 in 150 children is given a diagnosis of autism
— Boys are four times more likely than girls to have autism
— Autism is defined by three main impairments:
Social interaction
This ranges from a lack of intuition of social stimuli to the inability to
form attachments to carers
Communication
Autistic children show impairments such as delays in language development and
a reduced ability to initiate and sustain conversations
Restricted, stereotyped repetitive behaviour
These include obsessively arranging objects or following very specific
routines
— Other problems include phobias, sleeping and eating disturbances,
tantrums and self-directed aggression
— Many autistic children show above average intelligence
— There are no current effective means to prevent, treat or cure autism
Sources: Autism Speaks, World Health Organisation, Institute of Psychiatry
