Autism arises from mutations in the genome, but several studies have shown that many of these mutations occur de novo — that is, they are non-inherited mutations introduced during the parents’ gamete production. Jay Gingrich at the New York State Psychiatric Institute and his colleagues plan to study a mouse model of autism to investigate whether advanced paternal age is linked to a higher frequency of genomic aberrations. Sperm cells are generated throughout life without a clear decrease in viability or health, but recent surveys have found that children of older fathers have a higher risk of developing several disorders, including autism. Gingrich and colleagues have developed a model system in which to study the effects of aging on male fertility and sperm fitness. For four years, they have bred female mice with either young or old male mice to generate distinct mouse lines. In preliminary studies, mice with older fathers show several behavioral traits, such as a fear of unfamiliar mice and objects, that are often observed in people with autism. Gingrich proposes that these mice are a natural model for autism, and aims to study how paternal age affects the brain structure and behavior of their offspring. How aging alters sperm viability remains unclear. One possibility is that, over time, sperm stem cells acquire genomic aberrations such as mutations or an abnormal pattern of DNA modifications such as methylation. DNA methylation and other epigenetic modifications change the activity of genes without altering the genetic sequence. These patterns would be maintained in the genome and passed on during stem cell division. But the many divisions over years of sperm production could weaken the fidelity of this replication, and incorrect patterns could cause abnormal gene expression in the offspring. To examine this possibility, Gingrich and colleagues plan to use an unbiased, high-throughput method to study the methylation patterns in genomes isolated from the brain tissue of mice with young or old fathers. They hope to find specific regions in the genome that are more susceptible to age-related mis-methylation and to identify nearby genes that could be responsible for the abnormal behavior in the pups of older male mice. The researchers then plan to determine whether these loci are atypically methylated — and so misregulated — in people with autism.