Understanding Autism Spectrum Disorder

Methylation clues
Genetics may be heavily involved in ASD etiology – but now, researchers are also considering the role of paternal epigenetics. A team at Washington State University collected sperm samples from fathers of children with or without ASD to analyze changes in DNA methylation (1). Genome-wide analysis revealed 805 significant differential DNA methylation regions that could be used as biomarkers to predict paternal offspring autism susceptibility.

AI on autoantibodies
But what about mothers? Maternal autoantibodies of fetal brain proteins specific to ASD have previously been identified – a subtype now known as maternal autoantibody-related ASD (2). Researchers have developed a serological assay with machine learning to identify the reactivity patterns of ASD-specific maternal autoantibodies against eight proteins highly expressed in the fetal brain (3). The highly accurate test has the potential to predict a woman’s likelihood of having a child with ASD.

Beyond inheritance
Mosaic mutations that occur after conception may also be implicated in ASD. Using whole-genome sequencing to investigate mosaic mutations in the frontal cortices of participants with and without ASD, researchers have found that most brains have similar rates of “point” mosaic mutations – but, in participants with ASD, mutations were more likely to affect parts of the genome involved in brain function (4).

Copy number variants (CNVs) in mosaic patterns also differ in people with ASD (5). Larger CNVs (involving 25 percent or more of a chromosome) have been found in people with ASD and CNV size is linked to ASD severity. Surprisingly, smaller CNVs that are associated with ASD when found in all cells were no longer correlated when found in mosaic mutations.

It’s clear that no one factor causes ASD – and that only by stepping back and assessing each piece of the puzzle can we develop accurate screening and diagnostic tools.