The immune system is a zealous defender against small invaders, but it can sometimes turn against our own bodies. This possibility is especially frightening for pregnant mothers, whose immune systems may harm the fetus. We have written periodically about the building case that viruses during pregnancy can raise the risk of schizophrenia, and this month brings evidence that the immune system may contribute to another devastating mental disorder� autism.
David Amaral, research director of the UC Davis M.I.N.D. Institute, was tipped off to the danger of antibodies by the work of his colleague Judy Van de Water. Van de Water's work, which will appear next month in Neurotoxicology, demonstrated that when extracted from the mothers of autistic children, IgG is capable of damaging fetal brain proteins. IgG is the only type of antibody that can cross the placenta, making it the obvious suspect. Van de Water used a method called the "Western Blot" to figure out which IgG attack good cells. Interestingly, only 15-20% of women who gave birth to autistic children had these auto-antibodies.
Amaral collected large amounts of the dangerous IgG that Van de Water identified. When he injected these samples into pregnant monkeys, a glaring difference emerged in early childhood. All four monkeys from the experimental group displayed conspicuous stereotypies� repetitive movements like arm flapping that are common to autistics. These behaviors were especially pronounced when the monkeys faced the stress of a new environment. The group injected with normal human IgG showed no abnormal behavior.
Though his sample size was too small to allow firm conclusions, Amaral expressed confidence that larger studies will confirm the link between IgG and stereotypies. While that seems plausible, it will be much more difficult to prove that IgG is the sole cause of autism. For one thing, stereotypies are just one of three main symptom categories; social impairment was minor in the monkeys, and they did not know sign language so could not be evaluated for language impairment. Secondly, Van de Water�s experiments only showed a connection to the IgG of mothers with regressive, not early onset, autism, and not even all of these mothers had the auto-antibodies.
In an interview with us, Amaral agreed that autism is probably caused by multiple mechanisms, but he offered an intriguing way that his experiment could still be consistent with one mechanism. He explained that practical considerations limited the amount of auto-antibodies that could be extracted, requiring them to confine their injections to the last half of the second trimester. It is quite possible that this was the period of fetal development where only motor functions are vulnerable, and that social and language functions would have been impaired if injections had been sustained throughout pregnancy.
These findings have important implications for treatment. It was already known that mothers who have one autistic child have a much higher chance of having another, but that elevated risk is only around 10%-20%. Now that dangerous IgG has been identified as a possible antigen, doctors could conceivably give mothers a much more specific warning. While the warning alone will be helpful, Amaral believes that techniques like plasmapheresis, a type of blood transfusion, will eventually be able to restore the womb to the status of nurturing haven.
|