Today it is clear that ASD has a multigenic aetiology. ASD-related diseases with monogenic causes are very few; they include Rett syndrome, fragile X syndrome, tuberous sclerosis, and Schuurs–Hoeijmakers syndrome.
The disorder is highly heterogeneous, with few pathogenic variants present in a significant percentage of afflicted individuals.
Large-scale genetic studies conducted on ASD patients and their families unveiled hundreds of risk genes. While these proteins have diverse functions, studies have focused on two broad classes of proteins: those involved in synapse formation, and those involved in transcriptional regulation and chromatin-remodelling pathways.
Synapse-related risk genes include those encoding cell-adhesion proteins such as neuroligins, neurexins and cadherins; synaptic vesicle cycling proteins synapsin-1 (SYN1) and synapsin-2 (SYN2); synaptic Ras GTPase activating protein 1 (SYNGAP1), ion transport proteins such as sodium voltage-gated channel alpha subunit 2 (SCN2A), potassium voltage-gated channel subfamily Q members 3 and 5 (KCNQ3 and KCNQ5), potassium voltage-gated channel subfamily D member 2 (KCND2), calcium voltage-gated channel subunit alpha1 E (CACNA1E), calcium voltage-gated channel auxiliary subunit beta 2 (CACNB2), glutamate receptor signalling protein SH3 and multiple ankyrin repeat domains 3 (SHANK3), and gamma-aminobutyric acid type A receptor gamma3 subunit (GABRG3).
De novo mutations
De novo mutation load is also found to be very high in regulatory proteins of ASD risk genes in patients. Chromodomain helicase DNA binding protein 8 (CHD8), activity-dependent neuroprotector homeobox (ADNP), pogo transposable element derived with ZNF domain (POGZ) are among those susceptible to impact transcription and chromatin-remodelling pathways.
Another factor increasingly being recognised as playing crucial to ASD is somatic mosaicism, which is derived from a postzygotic
mutation. Somatic mosaicism is currently estimated to account for about 3–5% of ASD cases.
Copy number variations (CNVs), either inherited or arisen de novo — including duplications, deletions, translocations, and inversions in loci such as 15q11-13 as well as 16p11.2 — is seen in nearly 1% of autism cases.
Yet another set of genes consistently implicated in ASD susceptibility are those with epigenetic-modulating functions involved in methylation, such as chromatin remodelling proteins like MeCP2, KMT2C, RNA-binding/splicing proteins such as fragile X mental retardation protein (FMRP), and RNA binding forkhead box (RBFOX) genes, lysine methyltransferase 5B (KMT5B), and lysine demethylase 6B (KDM6B); mind bomb E3 ubiquitin protein ligase 1 (MIB1); or transcription factors like ADNP and additional sex combs like 3 (ASXL3).
Studies focused on major psychiatric diseases have identified extensive overlap in risk genes. Schizophrenia has been the most correlated mental disorder with ASD. Intellectual disability (ID), and schizophrenia have been found to share risk loci in FMRP targets, CHD5, CHD8, SCN2A, and neurexin 1 (NRXN1) with ASD. Increased incidence of de novo pathogenic variants in periodic circadian regulator 1 (PER1) and lysine methyltransferase 2C (KMT2C) are detected across ASD, ID, and bipolar disorder.
Significant links between autistic and ADHD traits have also been established.