Session 5: Environmental Influences as Etiologic Factors in Autism
Patricia M. Rodier, Ph D
Department of Obstetrics and Gynecology, and
Rochester CPEA, University of Rochester, Rochester, NY
The autism spectrum of disorders is unique among psychiatric disorders in having a set of known environmental risk factors. Increased rates of ASDs are seen after exposure of the embryo to rubella, thalidomide, valproic acid (VPA), ethanol, and misoprostol. While these account for a tiny fraction of cases, they provide research opportunities that are not available in studies of idiopathic autism. Exposed human populations represent natural experiments, and environmental risk factors can be manipulated experimentally in animals.
Examination of subjects exposed to thalidomide in utero provided the first information on the stage of development when some cases of autism arise. The physical malformations of the thalidomide cases with autism indicate a critical period of only a few days during closure of the neural tube - days 20-24 post conception. Only a few neurons form so early. Most are motor neurons for cranial nerves, including abducens, facial, oculomotor, trigeminal, and hypoglossal. The neurological dysfunctions of the thalidomide cases with autism confirmed that neurons forming in the critical period were injured.
We have reproduced the reduction in numbers of some tegmental cranial nerve neurons by exposing rats to VPA during neural tube closure, and documented a shortening of the brain stem between the facial and glossopharyngeal nerves. Significant reductions in volume and neuron number are present in the inferior olive, which forms at the same time as the motor nuclei. The major target of the inferior olive, the cerebellum, is also hypoplastic, even though it forms after the exposure period. This suggests that early injury can alter the development of later-forming structures. A behavior reported to be abnormal in autism, compared to IQ-matched controls, is eyeblink conditioning. When tested with the same paradigm (one known to depend solely on reflex loops between the brain stem and cerebellum), VPA-exposed rats have been shown to mimic the abnormal responses reported in people with autism.
Alteration of the tegmental nuclei and shortening of the hindbrain are characteristic of the Hoxa1 knockout mouse. We identified a polymorphism of the human gene and tested families with evidence of familial ASDs. The newly discovered "G" allele was transmitted preferentially to probands in our sample, suggesting that it confers susceptibility to autism. The effect is greater when the allele is transmitted from the mother than when it is transmitted from the father.
Why do humans and animals exposed to VPA have phenotypes similar to those with familial autism? VPA has a potent effect on the expression of Hoxa1. Using real-time PCR, we found that two teratogenic versions of the compound altered Hoxa1 expression, while a non-teratogenic analog had no effect on expression. This result supports the hypothesis that VPA's effect on Hox gene expression is one of its mechanisms of teratogenicity.
A new study of Brazilian children with Moebius syndrome after misoprostol exposure answers an important question about brainstem development and autism. Moebius syndrome is a congenital disorder of CN VI and VII that is sometimes familial and sometimes environmental in origin. About 30% of idiopathic cases are co-morbid for autism. Misoprostol has a different mechanism from VPA and other environmental risk factors, and exposure occurs in the sixth week post-conception. The similar rate of autism in idiopathic and misoprostol-induced cases indicates that it is the location of injury, rather than the time of injury, that is critical to the etiology of autism.
Back to Agenda