More than 180 genes have been identified as contributing to autism spectrum disorder (ASD). Historically, however, researchers have focused on diagnosing the condition rather than understanding the spectrum of phenotypes these gene interactions can express.
In a recent study published in Nature Medicine, scientists conducted the largest meta-analysis to elucidate the phenotypes of ASD-diagnosed individuals and hitherto understudied undiagnosed counterparts.
Study: Phenotypic effects of genetic variants associated with autism. Image Credit: SewCreamStudio/Shutterstock.com
Background
This research revealed genetics' direct influences on ASD diagnosis and the consequences of social, economic, and environmental contexts in the externally observable behaviors of individuals on the ASD spectrum.
'Autism' is the abbreviation for 'autism spectrum disorder (ASD)', a complex neurological and developmental condition that impairs an individual's social interaction, communication, intelligence, and behavior. The World Health Organization (WHO) estimates that autism affects 10% of the global population.
The phenotypic expression of ASD is unique for each individual, and the body of literature exploring the gene-environment interactions which shape these expressions is rapidly growing. Despite decades of research into the genotypic underpinnings of ASD and other neurodevelopmental disorders (NDDs), most research has focused on diagnosed patients.
Undiagnosed individuals have remained largely ignored, even when they possess one or more of the same loss-of-function (LoF) gene alleles found in ASD patients.
This presents a gap in our understanding of the inter-individual phenotypic variability of members of the ASD spectrum and its impact on the socioeconomic status of the undiagnosed many.
About the study
In the present study, researchers reviewed four studies within a meta-analytic framework to elucidate genes present in patients with ASD and their relative contributions to cognitive impairment in undiagnosed individuals with shared European ancestry.
They began by collecting 226,649 whole exome sequences (WES) comprising 13,091 diagnosed ASD patients, 19,488 of their immediate relatives, and 194,070 undiagnosed others.
One hundred and eighty-five autosomal genes were selected based on their having LoF-mutated alleles with confirmed associations with NDDs.
They then collected and compiled a set of 2,492 genes without direct NDD association, which research has concluded are 'intolerant' to ASD LoF variants. Intolerant genes are those in which protein-truncating variants (PTVs), or other mutations, are absent or kept at very low frequencies, i.e., natural selection actively filters out or selects against their persistence in a population.
Relationships between autism and biological function were analyzed by evaluating the quantity of gene expression at four different regions of the brain and across eight different periods of fetal development.
Autism diagnoses based on genome-wide association studies (GWAS) were simultaneously condensed to arrive at a polygenic score (PGS) to evaluate if genetic composition could accurately predict ASD's presence and intensity in previously undiagnosed humans.
Phenotypic cognition information in the forms of social and communication questionnaires (SCQ), intelligence quotients (IQ), and metrics of repetitive behaviors, social interactions, difficulties in communication, and self-injurious behaviors was compiled for participants.
This was done to establish conventional diagnostic cutoffs in an otherwise near-continuous spectrum of non-ASD to severely ASD-prone.
Finally, magnetic resonance imaging (MRI) of 21,040 individuals from the UK Biobank, including 1,675 ASD patients, was analyzed to test previous claims that some genes associated with autism could alter brain anatomy.
Study findings
Corroborating previous research, all 185 selected LoF genes were found in various combinations in individuals diagnosed with ASD, with gene prevalence matching odds ratios (ORs), i.e., higher OR corresponded with wider prevalence in the ASD group.
Of these, 134 LoF genes were found in at least one undiagnosed individual, potentially indicative of the complex environmental, societal, and genetic interplays which underpin behavior and diagnosis.
Presence and ORs of genes in undiagnosed individuals were correlated with reduced qualifications and income and reduced intelligence, all of which were statistically significant and highlight the lack of support provided by society not only to individuals at the extreme of the ASD spectrum but also those with shallow genetic dispositions to the condition.
Surprisingly, high autism PGS scores in diagnosed individuals, while directly proportional to reduced income, are inversely associated with fluid intelligence – the more autistic the individual, the more intelligent they are.
MRI imagery did confirm that significant alterations in brain anatomy were present in individuals positive for ASD LoF genes when compared to the individuals devoid of these mutations confirming and corroborating previous research.
Conclusions
This research represents the largest meta-analysis of autism to date and is the first to focus on the undiagnosed carriers of autism-associated genes. Results indicate that the presence and ORs of ASD genes can predict an individual's diagnosis of the condition.
However, societal factors and developmental environment can result in carriers (individuals having autism-associated genes) living their whole lives with the economic and cognitive demerits of ASD while remaining undiagnosed.
Sex was identified as a confounding factor for some genes – mothers were found to transmit some LoF genes more than fathers. While the gene load in both female and male offspring was found to be a near-identical, phenotypic expression for autism was preferential or exclusive in males, implying that females inherently resist low ORs of rare ASD genes better than their male counterparts.
"Social environments also influence whether people with autistic traits receive a diagnosis, and there is still progress to be made on a societal level to enable people with all different neurological and developmental diversities to thrive."
Finally, this research highlights the need for education systems to be tailored to individuals with varying needs across the NDD spectrum, which could aid them later.
Additionally, the presence of ASD genes in an individual remains more often than not hidden in its phenotype, and future research exploring the biological underpinning of multiple such gene interactions is necessary to improve future diagnosis.
Such research, when paired with data considering environmental, societal, economic, and genetic factors, could help benefit humanity at large, given the added contributions of its gifted many.
Machine learning system offers new hope for diagnosis of rare genetic disorders