Researchers know that defects in brain development are to blame for autism, but pinpointing the likely genetic culprits has remained an elusive goal. To put it in mobster terms, there is no single “Tony Soprano gene” corrupting brain development. Instead, scientists believe that multiple “small-time thug genes” gang up to undermine the developing brain. Because the individual effects of these crooked genes may be subtle, it has been hard to get the goods on them.

To help crack the case, Matthew W. State, M.D., Ph.D. ’01, the Irving B. Harris Associate Professor of Child Psychiatry in the Child Study Center and of genetics and director of the Program on Neurogenetics, sought to flush out autism-associated genes by focusing on clues from certain affected individuals. Children with autism—or any of a spectrum of related disorders—have difficulty communicating and interacting with others, exhibit stereotyped behaviors and often suffer from mental retardation and seizures. A small percentage of these children also have a visible chromosomal abnormality. In one such patient State found that the abnormality disrupted the Contactin Associated Protein-Like 2 (CNTNAP2) gene, which encodes a protein that helps brain signals pass from one neuron to another. Based on prior work by himself and others linking contactin proteins to autism spectrum disorders, mental retardation and seizures, State grew suspicious of CNTNAP2.

With the help of colleagues in clinical medicine, neurobiology, biochemistry and genetics, State has collected a body of evidence that strongly incriminates CNTNAP2 as one of perhaps many autism accomplices. First, CNTNAP2 is present at the scene of the crime, including all layers of the cerebral cortex within the temporal lobe and within the limbic system, a brain circuit involved in social behavior. Second, CNTNAP2 is found with its binding partner, contactin 2, at synaptic plasma membranes—the gates of communication between neurons. The third and strongest line of evidence against CNTNAP2 is that sequencing of the gene from 635 autistic patients and 942 controls turned up 13 rare, unique changes to the encoded protein that were found only in autistic individuals. Eight of these mutations are predicted to disrupt the proper functioning of CNTNAP2. One particular mutation was identified in four autistic children in three unrelated families, but not in more than 4,000 chromosomes from controls. “This is strong but not definitive evidence linking this gene with autism,” according to State.

Unbeknownst to each other and to State, two other medical research laboratories—the labs of Daniel H. Geschwind, M.D., Ph.D., at the University of California, Los Angeles (UCLA), and of Aravinda Chakravarti, Ph.D., at Johns Hopkins University—also fingered CNTNAP2 as causative of autism. Both Geschwind and Chakravarti independently homed in on CNTNAP2 after surveying the genomes of hundreds of individuals and identifying a particular chunk of genetic material that appeared to surface in families with autism. State and Geschwind, longtime friends who met as residents at UCLA, learned of the other’s discovery while catching up during one of their regular phone conversations. Soon after, Geschwind caught wind of Chakravarti’s work through the research grapevine. When the three scientists compared notes, they decided to co-publish their findings to build the strongest case possible against CNTNAP2. “There’s a reason we’re all landing on this gene,” said State. All three papers were published in the January issue of the American Journal of Human Genetics.

State thinks that identification of CNTNAP2 may give him the traction he needs to begin to understand the complex biology of autism. “Our hope is that our continued work on understanding the biology of CNTNAP2 will lead to real opportunities for novel approaches to treatment,” said State.