The strong heritability of autism spectrum disorder has prompted the field to take a “genetics first” approach, emphasizing a hunt for DNA-based clues in affected individuals and their family members. And indeed, we have seen that a portion of the autisms do have roots in the gene sequence, particularly in novel disruptions, or “de novo” mutations of the genetic code seen in affected individuals but not their parents.

But those rare mutations only explain a fraction of cases—clinically about 10%—and often those cases represent rare genetic syndromes such as Angelman Syndrome or Fragile X that are not what most would consider autism as it is commonly understood. Furthermore, the search for common genetic variants in autism has so far yielded much speculation with few actual findings. A recent robust attempt to actually locate common genetic risk variants for autism, while identifying five genome-wide-significant loci, could not account for more than an insignificant fraction of the overall population risk. Further, a recent review and reevaluation of genome wide association studies found that almost no of autism risk could be predicted accurately from common variants.

Moreover, though recent studies have pegged autism’s heritability at 64 to 85 percent, this family-based pattern is idiosyncratic, based largely on increased recurrence rates among same-generation family members, mainly siblings, and to a lesser degree half-siblings and cousins. Studies have not revealed a tendency for autism to travel down family trees. Indeed, researchers regularly remark on the paradox that autism’s prevalence is increasing even as autism is subject to “negative selection” since affected individuals, particularly those like my two severely autistic children, are unlikely to reproduce.

So where does the strong heritability come from? I think we can only fully answer this question if we reject the underlying assumption that autism’s heritability is necessarily driven by “genetics” or the DNA sequence.

DNA is just one part, albeit an extremely important part, of the dizzyingly complex molecular universe within the egg and sperm and their lineage of precursor germ cells.These cells harbor billions of molecular factors that help guide the nature and timing of how the human genome is expressed over the course of development. It is not just the genes, but also the molecular programming around those genes, that enables the eventual development of a human, including the unfathomably intricate human brain.

In other words, to understand the complex spatio-temporal nature of brain development, it is not sufficient to consider glitches or variations in the DNA sequence; one must also take into account those countless chemical switches that tell the genes how and when to produce neurons and associated cells in the embryo, fetus, and child.

Though we often think of DNA as a linear code, the biological reality is rather more complicated: the three-dimensional double helix is highly wrapped, folded, looped, and decorated in ways that make some genes accessible for expression, and others not. Proteins create structural packaging for the DNA, determining how the strands are exposed or tucked away. Proteins called transcription factors bind to sequences in the DNA to control transcription of targeted genetic information. Methyl groups can attach to DNA, down-regulating expression of underlying genes. Non-coding RNAs can alter the final gene product.

In germ cells, these, an other non-genetic drivers of gene expression—often called “epigenetic,” though the terminology in this dynamic and growing segment of genetic toxicology is far from settled—can in some cases be perturbed by external factors. When the epigenetic machinery changes its instructions, genes function may go awry.

Mammal models have provided many examples of this sort of epigenetic dysregulation and, ultimately, altered brain development and/or behaviors in the offspring. For example, synthetic hormones administered to a parent (and its germ cells) can modify hypothalamic-pituitary-adrenal (HPA) function in offspring. Nicotine exposure to a parent (and its germ cells) has been shown to raise the risk for hyperactivity in the progeny generation via epigenetic mechanisms. General anesthesia (GA) gases administered to a parent (and its germ cells) have been shown to alter brain gene expression via epigenetic mechanisms and induce learning deficits in offspring. Intriguingly, GA is known to disrupt the proper expression of genes important for early brain development and known to be dysregulated in autism, including genes related to neuronal cytoskeleton, neural migration, GABAergic function, and synaptogenesis.

While studies of germline exposures are rare in human cohorts (for example, there have been no studies of potential heritable impacts of GA exposures in humans despite the ubiquity, neurotoxicity and genotoxicity of those chemicals), recent retrospective studies of certain toxicants suggest this possibility. Grandmaternal pregnancy use of the toxic synthetic estrogen drug diethylstilbestrol (DES) was found to be associated with significantly elevated odds for ADHD in grandchildren, through the exposed female line. And grandmaternal smoking in pregnancy was found to be associated with increased risk for autism traits and diagnosed autism in grandoffspring.

Despite the evidence that non-genetic heritable factors can perturb brain development, the field remains largely focused on sifting through the DNA sequence with little, if any, regard for the germ cell contents as a whole, or for the toxicological histories of those germ cells.

With autism’s prevalence continuing to rise, and genetics research yielding diminishing returns, the field would benefit from a fresh interdisciplinary approach that encompasses a more modern, biologically authentic understanding of inheritance. Paradigm shifts happen slowly in science, but when it comes to solving the the puzzles of autism we simply can no longer afford to waste any time.

Jill Escher is an autism research philanthropist and a councilor of the Environmental Mutagenesis and Genomics Society.

REFERENCES:

Escher J. Bugs in the program: can pregnancy drugs and smoking disturb molecular reprogramming of the fetal germline, increasing heritable risk for autism and neurodevelopmental disorders? Environ Epigen 2018;4(2);dvy001.

Escher J. and Robotti S. Pregnancy drugs, fetal germline epigenome, and risks for next-generation pathology: a call to action. Environ Mol Mutagen 2019;6(5):445-454.