The list of high-confidence autism-linked genes has expanded to 253, according to a new preprint, which also offers new insights into how the genes relate to other conditions and the biological pathways they share.
Genes harboring rare, de novo variants linked to autism, or to neurodevelopmental conditions more broadly, differ in their biological roles and associated cell types, a companion preprint suggests. And a small subset of these variants have a large effect on a person’s genetic susceptibility to autism, unlike common variants—which individually have tiny effects on autism likelihood, another accompanying preprint shows.
Together, the work builds on a landmark 2022 study by the same team, which identified 72 genes by searching for rare coding variations in approximately 20,000 autistic people. Those genes were then categorized as either autism-specific or associated with neurodevelopmental conditions more broadly, with the two groups linked to different stages of neuronal development. Genes tied to both autism and broader neurodevelopmental conditions show stronger expression in immature neurons, whereas autism-predominant genes are enriched in mature excitatory neurons, the earlier study found.
The new study identified 253 genes containing de novo autism variants among exome sequences from 62,429 autistic people in the Autism Sequencing Consortium and three other major cohorts. (Two of these cohorts, the Simons Simplex Collection and SPARK, are funded by the Simons Foundation, which also supports The Transmitter.) The larger dataset revealed a robust pattern in the genetic overlap between autism and other neurological conditions that the earlier analysis had hinted at, says Michael Talkowski, professor of neurology at Harvard Medical School and an investigator on the preprints. “It really motivates us to try to understand exactly what is driving those associations and resulting in these genetic similarities across these disorders.”
Talkowski and his colleagues looked for associations between these autism-linked genes and schizophrenia, bipolar disorder, epilepsy and other neurodevelopmental conditions. The genes fell into six clusters depending on their overlap with other conditions, the study found. Genes associated with autism alone, or autism and schizophrenia, were less likely to harbor de novo variants than those also implicated in neurodevelopmental conditions more generally.
T
he six clusters could also be considered within two broader categories based on their specificity to autism. These two groups were enriched for different functions: Synapse assembly was more common among autism-specific genes, whereas transcriptional regulation was enriched among those linked to a wider range of conditions.These two groups also differed in developmental outcomes: Children carrying a rare variant in a gene connected to many conditions were more likely to show developmental delays and less likely to reach developmental milestones than those carrying a variant in an autism-specific gene.
Those clusters are probabilistic but not predictive, says study investigator Mark Daly, co-director of the Program in Medical and Population Genetics at the Broad Institute of MIT and Harvard. “We are not yet at a point where individual genetic results can accurately predict patient trajectories.”
Even so, the findings provide “concrete information on the different [autism] subtypes,” says Carrie Bearden, professor of psychiatry and biobehavioral sciences at the University of California, Los Angeles, who was not involved in the work. Understanding how these genetic clusters track with different medical outcomes “is a really important contribution.”
Broadly, the 253 autism-linked genes appear to converge on three biological processes: genomic regulation, neuronal development and synaptic function, the second preprint found. The genes are most strongly expressed in the posterior cortex and certain types of developing neurons, including newborn excitatory neurons, immature interneurons, and maturing intratelencephalic neurons.
Autism-specific genes tend to be associated with immature interneurons, whereas those linked to the broader range of neurodevelopmental conditions show stronger enrichment in excitatory neurons.
The findings offer a coherent picture of how a diverse set of genes contribute to autism, says Aaron Besterman, medical director at the Laura Rodriguez Research Institute of Family Health Centers in San Diego, who was not involved in the study. Though previous studies had recognized gene regulation and synaptic biology as key functions among implicated genes, the new work “develops a more detailed picture of how developmental regulators might connect those processes—for example, by controlling genes that later help neurons communicate.”
I
n the third preprint, the team modeled the effects of de novo coding variants using genetic data from 38,680 autistic people and their parents and non-autistic siblings. De novo variants account for 3.4 percent of the variation in people’s genetic susceptibility to autism, the study found. Among people carrying spontaneous variants, half of this variance—about 1.7 percent—can be explained by just 15 genes, most of which are associated with severe neurodevelopmental conditions.About 7 percent of autistic people carry a spontaneous variant predicted to increase the likelihood of having the condition, the study also found. Variants with the strongest effects tend to occur in constrained genes—those that show little variation in the general population—with the 20 percent most constrained genes accounting for about 90 percent of the variation in genetic likelihood attributable to de novo variants.
Many autism-linked genes remain undiscovered, including roughly 25 percent of genes strongly linked to the condition, the model predicted. The pace of future discovery also depends on the source of the data: New autism-linked genes can be discovered more quickly if analyses include autistic people with intellectual disability, the forecasting analysis found.
Two of the papers were posted on bioRxiv in August, and one in September.
The findings suggest that researchers could maximize gene discovery by optimizing the sources of new genetic data, says Maria Chahrour, associate professor of neuroscience at the University of Texas Southwestern Medical Center. “It’s a practical, useful finding for how future studies should be designed.” Chahrour did not contribute to the preprints but has studied the genetics of autism in an East African cohort.
And the fact that most autistic people still lack a rare-variant explanation, despite the growing list of implicated genes, highlights the need for continued investment in compiling large-scale genetic databases and greater attention to gene-environment interactions, Chahrour says. “We should also focus on increasing ancestral diversity in the cohorts to close the gap in identifying new variants.”