Diagnosing autism spectrum disorders (ASD) depends heavily on neurological and psychological analysis. Research has identified some genetic risk factors for the disorder, including a high heritability of ASD. However, more research is needed to characterize the genetic basis of ASD and identify risk factors. Genome-wide studies could potentially address diagnostic and treatment limitations.
What Is ASD?
Autism Spectrum Disorders are a complex group of neurodevelopmental disorders. They can cause challenges in social interaction, communication, and behavior. An estimated 2.3% of American children have ASD.
ASD is diagnosed via a medical examination of the child’s development and behavior. Parents and caregivers sometimes notice differences in the child’s interactions with other children.
In adults, ASD diagnosis is achieved through observations of social interaction, behaviors, and other social interactions. Diagnosis is not straightforward, often because observed symptoms may be similar to other neurological disorders.
Investigating the Causes of ASD
The debate between genetic and environmental factors in autism is a long-standing one. Most experts now agree that a combination of factors lead to the development of ASD.
It’s clear that autism spectrum disorders have a genetic component, and often are associated with other genetic disorders. Rett syndrome is an autism spectrum disorder with mutations in the MECP2 protein. Similarly, fragile X syndrome is often present in individuals with ASD.
Epigenetic studies also indicate an environmental influence on ASD. A 2014 study conducted genome-wide DNA methylation studies on brain samples. Results showed two hypomethylation regions in individuals with ASD. Genome-wide association studies (GWAS) have also identified an impacted region on chromosome 5p14.1.
In a study published in Nature, researchers conducted whole-exome sequencing on blood samples from parents and children with and without ASD. They found that 13% of de novo missense mutations contribute to 12% of ASD diagnoses, while 42% of de novo mutations contribute 9% of diagnoses. Mutations vary based on sex and IQ level.
Next-Generation Technologies in ASD Research
NGS has revolutionized the study of medical disorders and conditions. Unprecedented speed and lower cost allow the collection, analysis, and archive of data. This data can improve early diagnosis and provide clues to creating personalized therapy for those with ASD.
Whole-genome sequencing identifies coding and noncoding variants throughout the sample genome. Past studies have worked to create more valuable reference genomes for the study of ASD.
In a Molecular Psychiatry study, WGS of 85 quartet families addressed the need for a comprehensive data resource for ASD. Their results show that 69.4% of affected siblings carried different ASD-related mutations. This correlated to differential phenotypic presentation in these siblings. The investigators concluded that WGS is crucial to better identify all variants associated with ASD risk.
NGS studies can also help classify ASD for medical intervention. A study of 325 Canadian children calculated a score for each child’s genome-wide rare variants. They found that children with dysmorphic ASD had substantially more rare variants than children with nondysmorphic ASD.
Common variants were also more often observed in nondysmorphic ASD samples. Their results support classifying ASD by morphology, rather than by symptoms, for more targeted treatments.
Advances in Genomic Prediction
ASD research has many goals, one of which is to predict the risk of ASD occurring in children. The median diagnosis age for ASD is 52 months of age, even though biomarkers are present as early as the first trimester. Because of the heterogeneity of the disorder, a reliable genetic screening has yet to be developed. A study from Molecular Psychiatry was able to classify 88% of typically developing and ASD toddlers via leukocyte RNA information.
Protein-altering genetic variants are present in many ASD cases and in samples from people with related diagnoses. Still, we only have strong evidence for about 65 causal genes out of an estimated several hundred. Princeton researchers developed a machine learning approach to tackle this problem.
Their approach yielded a genome-wide prediction of autism risk genes. By creating a disease-gene classifier, this study was able to identify genes that may be related to ASD, which could then be sequenced and verified. They also found that, while there are hundreds of ASD-linked genes, many of those genes are concentrated on specific neurodevelopmental stages and brain pathways.
As we learn more about the heritability of autism, noninvasive prenatal testing (NIPT) may be possible in the near future. Prospective parents could better understand their likelihood of having children with autism. With this come ethical considerations, as with other current NIPT applications. However, if parents understood their children’s risk of having autism, early detection would allow them to begin therapeutic and social skills interventions to benefit the child’s development.
The heterogeneity of ASD is the biggest challenge in diagnosis and treatment. Further progress requires comprehensive genetic research approaches to link genetic variants to functional pathways. Molecular and bioinformatic strategies can help make targeted therapies possible to provide a higher quality of life and social functionality for people with ASD.
