NEWSFOCUS

Environment counts. Despite the highly genetic

nature of autism, which researchers are now
deciphering, specialized school programs help.

Autisms Cause May Reside in

Abnormalities at the Synapse

190

13 JULY 2007

VOL 317

SCIENCE

Published by AAAS

www.sciencemag.org

Downloaded from www.sciencemag.org on November 13, 2011

The first autism genes?

Zoghbis provocative 2003 synapse hypothesis
rested partly on work that year by a group led by
Thomas Bourgeron at the Pasteur Institute in
NEUROSCIENCE
Paris, France, that found mutations in proteins
called neuroligins in two pairs of Swedish
brothers with autism spectrum disorder.
Neuroligins are proteins expressed on the
surface of the postsynaptic neuron that bind to
proteins on the presynaptic neuron called
neurexins, spanning the synapse and forming a
New genetic evidence is leading researchers to home in on the cleft separating neurons physical tether. Together, neuroligins and
neurexins are thought to play key roles in the
as the site where the disorder may originate
formation and functioning of synapses.
No one knows what causes autism, which in the term specialists prefer, is overwhelmingly
Some researchers contested the Pasteur
its broad definition affects about 1 in every genetic. Based mostly on studies of fraternal Institute findings, however, because no other
150 children. The impaired social interaction, and identical twins, University of Illinois at reports of these mutations in other individuals
communication deficits, and restricted and Chicago autism researcher Edwin Cook con- with autism followed; some even questioned
repetitive behaviors seen in people with the cludes that genetic factors contribute about whether the Swedish brothers actually had
condition have confounded scientists since it 90% to autism, with environmental factors autism. If it wasnt [autism], it was pretty damn
was first identified in 1943. Because only a contributing no more than 10%. Autism is the close, says Scherer.
minority of autistic persons have severe intel- most heritable of neurodevelopmental disorders
These rare neuroligin mutations and other
lectual disability, and some show exceptional that are complex in origin,
suggestive evidence linking
cognitive talents, relatively subtle changes in says Scherer. (Biology is not
some neuroligin-binding prothe brain are probably responsible. Now a flurry destiny, of course, because the
teins to autism led Bourgeron
of new discoveries is pointing to one possible environment affects the form
to postulate a neuroligin
site of autisms origin: the synapse.
any genetic disorder takes,
autism pathway in which
Synapses are junctions across which and autistic children often
abnormalities in any of these
Neuroligin
neurons communicate. They are essential improve if placed in the right
dozen or more proteins could
for sensory perception, movement coordi- learning setting.)
predispose their possessors
nation, learning, and memoryvirtually
Abnormalities of chroto the disorder. Bourgeron
all brain function. The synapse is like the mosomes, many of them visbuttressed his case this Jansoul of the brain, says Huda Zoghbi, a ible under the microscope, -neurexin
uary, when his group identipediatric neurologist at the Baylor College are thought to account for
f ied mutations in one of
of Medicine in Houston, Texas. Its at the 10 to 20% of autism cases.
these proteins, Shank3, in
root of everything.
The effect of multiple genes
three autistic individuals. In
Zoghbi was the first to propose, in 2003, that acting in combination probasuch rare cases, mutations in
altered synapses might be responsible for bly accounts for most of the
this single gene seem to be
autism. But direct evidence was thin. Now rest. Two groups recently
sufficient to cause autism.
there seems to be a confluence of data flow- reported that many autism
Other groups, according to
ing, says Stephen Scherer, a geneticist at the patients have novel deletions
Scherer, are also reporting
Hospital for Sick Children in Toronto, Ontario. and duplications in their Critical connection. Neurexins Shank3 mutations in autistic
Until the mid-1980s, experts considered genomes (Science, 20 April and neuroligins coming together in patients. Its being repliautism a strictly environmental disorder, with 2007, p. 445), probably the synapse. Alterations in these cated for sure, he says. In the
most of the blame falling on faulty parenting. arising when chromosomes proteins could change how neurons one published study so far,
Now we know that autistic spectrum disorder, cross over during meiosis. communicate and lead to autism.
Shank3 mutations appear to

CREDITS (TOP TO BOTTOM): STEPHEN MCGEE/MCT/LANDOV; UNIVERSITY OF CALIFORNIA, SAN DIEGO/SKAGGS SCHOOL OF PHARMACY AND PHARMACEUTICAL SCIENCES

Researchers are honing in on the individual

genes responsible.
Because autism is a spectrum of disorders,
different gene combinations will play a role in
different individuals. Whats generating excitement now is the discovery of mutations in single genes that, in rare instances, seem able to
cause autism. These genes may be pointing
directly at a general mechanism for the disorder, at the synapse.

CREDIT: P. HUEY/SCIENCE

NEWSFOCUS

www.sciencemag.org

SCIENCE

VOL 317

Published by AAAS

13 JULY 2007

Downloaded from www.sciencemag.org on November 13, 2011

account for about 1% of autism cases.

later became clear that they also promote the the brain develops. Studies in animals to
Then, in March 2007, the Autism Genome recruitment of neurotransmitter receptors and understand the different components of the
Project Consortium, a group of over 50 institu- various structural molecules to the synapsein synapse and to determine mutation effects are
tions in North America and Europe, reported fact, orchestrating a complete synapse. Neuroli- just beginning.
results of a 5-year study on the genetics of gins and neurexins are unbelievably important
Many research groups are now focusing on
autism in 1600 families. In addition to several proteins, says Sdhof. Theyre life or death. finding links between synapse genes and
new chromosomal regions implicated in the
Clues to their possible role in autism are now autism. Cook argues for a broader approach,
disorder, the researchers found the neurexin-1 appearing. One theory is that an abnormal neu- including whole genome scans for other genes
gene associated with autism. Since neurexins roligin pathway upsets the balance of excitatory that might have less individual effect but may
bind to neuroligins at the synapse, this finding and inhibitory synapses in neurons, thereby account for more autism cases. (Some such
boosted the neuroligin autism pathway idea, affecting learning and memory, and thus lan- studies are in progress.) To say one or the other
although the studys authors did not look for guage and social communication. Broadly approach is the right way to go is, I think, at
specific neurexin mutations. (Several groups speaking, synapses can be either excitatory, this point nave, Cook says.
are now sequencing the gene.) Shank3 abnor- when the neurotransmitter glutamate is
malities also turned up in some Autism released, or inhibitory, with release of the Few genes or many?
Genome Project families, reports Scherer, the neurotransmitter gamma-aminobutyric acid The hope is that most cases of autism are caused
studys coprincipal investigator, again implicat- (GABA). The ratio of excitatory and inhibitory by just a few strongly acting genes, rather than
ing the neuroligin pathway.
synapses on a neuron determines whether it will many weak genes in concert. Simpler genetics
Bourgeron now feels vindicated. People in fire in any given situation. In the 21 June issue would accelerate understanding of the disorder,
the field are really accepting that
as well as facilitate early diagnosis
this is a pathway which is associand genetic counseling, and proPresynaptic
Presynaptic
ated with autism, he says. When
vide more discrete targets for therwe published the neuroligin [report
apy. Bourgeron notes that a single
in 2003], nobody believed it.
abnormal geneor even a single
Glutamate
Mutations in single synaptic
gene copy, as with Shank3can, in
genes, including neuroligins,
rare instances, cause autism. But
GABA
neurexins, and Shank3, will probaeven Bourgeron doubts that altered
-Neurexin
bly explain only a small number of
synapses by themselves are enough
Neuroligin-1,3
autism cases5% at most, Scherer
to cause most cases. Autism is not
Glutamate
Synapse
estimates. In the most convincing
a single entity, he stresses.
receptor
-Neurexin
case so far, Shank3, a single gene
He speculates that a combinacould cause this complex disease
tion of abnormal synapses and
Neuroligin-2
type, says Scherer. Thats
altered neural networksthe
PSD-95
GABA
GKAP
tremendously important, Scherer
complex circuitry involving the
receptor
explains, because it could provide
billions of neurons that permits
SHANK
clues to cellular defects underlying
language and social interaction
Gephyrin
all autism. In Alzheimers disease,
could combine to cause most
Postsynaptic
Postsynaptic
for example, mutations in the
cases of autism. Factors that could
-amyloid precursor protein (APP)
alter neural networks include a
account for a tiny fraction (less than Autisms origin? Neuroligins and neurexins, proteins crucial for aligning and global, as opposed to neuron0.1%) of all cases yet were crucial activating synapses, have now been implicated in autism, along with the Shank3 level, shift in the excitatoryin revealing the likely disease scaffolding protein. An altered balance between excitatory synapses (left) and inhibitory balance, increased
mechanism: the abnormal deposit inhibitory (right) could affect learning and memory during development.
neuron numbers (many autistic
of amyloid plaques in the brain.
children have large heads), or
This field, autism, is probably about 7 years of Neuron, Sdhof reported that in experiments high levels of the neurotransmitter serotonin,
behind the Alzheimers story, says Scherer.
in cells, overexpressing neuroligin-1 leads to seen in about a quarter of autistic patients.
excitatory transmission at synapses, whereas
Besides synapse abnormalities, many
Orchestrating the synapse
neuroligin-2 overexpression leads to inhibition. causes of autism have been postulated, from
Now the race is on to figure out how neuroli- Sdhof speculates that an alteration in either altered neuron migration during early developgins and their binding proteins are contribut- neuroligin could change the excitatory- ment to chronic inflammation in the brain.
ing to autism. What exactly do these proteins inhibitory balance, subtly changing the number Imaging and post-mortem studies suggest that
do at synapses? asks Thomas Sdhof, a neu- of neurons that are firing during brain develop- underconnectivity between brain regions is at
roscientist at the University of Texas South- ment. Such disruptions could eventually pro- the heart of the disorder (Science, 24 June 2005,
western in Dallas. Thats crucial for duce the lasting symptoms of autism, he p. 1856). Underconnectivity and altered
understanding autism.
explains, because synapses change with use, synapses are not mutually exclusive. If you
Sdhofs lab discovered neurexins in 1992 becoming more or less sensitive to stimuli have regionally different synapse dysfunction,
and neuroligins in 1995. They have been stud- depending on experience. This synaptic plas- youre going to have differences in connectivity
ied intensely ever since, because they seemed to ticity is the basis of learning and memory.
between different brain regions, says Sdhof.
hold the key to how synapses form, and thus to
Thats just one possibility. The synapse is Thats exactly what you would predict.
brain development. At first their pairing was extraordinarily complex, both chemically and
In the end, it may all come down to
thought to physically tether the synapse, but it structurally, and a lot could go wrong there as the synapse.
KEN GARBER

191