Autism

Description

Autism, the prototypic pervasive developmental disorder (PDD), is usually apparent by 3 years of age. It is characterized by a triad of limited or absent verbal communication, a lack of reciprocal social interaction or responsiveness, and restricted, stereotypic, and ritualized patterns of interests and behavior (Bailey et al., 1996; Risch et al., 1999). 'Autism spectrum disorder,' sometimes referred to as ASD, is a broader phenotype encompassing the less severe disorders Asperger syndrome (see ASPG1; 608638) and pervasive developmental disorder, not otherwise specified (PDD-NOS). 'Broad autism phenotype' includes individuals with some symptoms of autism, but who do not meet the full criteria for autism or other disorders. Mental retardation coexists in approximately two-thirds of individuals with ASD, except for Asperger syndrome, in which mental retardation is conspicuously absent (Jones et al., 2008). Genetic studies in autism often include family members with these less stringent diagnoses (Schellenberg et al., 2006).

Levy et al. (2009) provided a general review of autism and autism spectrum disorder, including epidemiology, characteristics of the disorder, diagnosis, neurobiologic hypotheses for the etiology, genetics, and treatment options.

Genetic Heterogeneity of Autism

Autism is considered to be a complex multifactorial disorder involving many genes. Accordingly, several loci have been identified, some or all of which may contribute to the phenotype. Included in this entry is AUTS1, which has been mapped to chromosome 7q22.

Other susceptibility loci include AUTS3 (608049), which maps to chromosome 13q14; AUTS4 (608636), which maps to chromosome 15q11; AUTS6 (609378), which maps to chromosome 17q11; AUTS7 (610676), which maps to chromosome 17q21; AUTS8 (607373), which maps to chromosome 3q25-q27; AUTS9 (611015), which maps to chromosome 7q31; AUTS10 (611016), which maps to chromosome 7q36; AUTS11 (610836), which maps to chromosome 1q41; AUTS12 (610838), which maps to chromosome 21p13-q11; AUTS13 (610908), which maps to chromosome 12q14; AUTS14A (611913), which has been found in patients with a deletion of a region of 16p11.2; AUTS14B (614671), which has been found in patients with a duplication of a region of 16p11.2; AUTS15 (612100), associated with mutation in the CNTNAP2 gene (604569) on chromosome 7q35-q36; AUTS16 (613410), associated with mutation in the SLC9A9 gene (608396) on chromosome 3q24; AUTS17 (613436), associated with mutation in the SHANK2 gene (603290) on chromosome 11q13; AUTS18 (615032), associated with mutation in the CHD8 gene (610528) on chromosome 14q11; and AUTS19, associated with mutation in the EIF4E gene (133440) on chromosome 4q23. (NOTE: the symbol 'AUTS2' has been used to refer to a gene on chromosome 7q11 (KIAA0442; 607270) and therefore is not used as a part of this autism locus series.)

There are several X-linked forms of autism susceptibility: AUTSX1 (300425), associated with mutations in the NLGN3 gene (300336); AUTSX2 (300495), associated with mutations in NLGN4 (300427); AUTSX3 (300496), associated with mutations in MECP2 (300005); AUTSX4 (300830), associated with variation in the region on chromosome Xp22.11 containing the PTCHD1 gene (300828); AUTSX5 (300847), associated with mutations in the RPL10 gene (312173); and AUTSX6 (300872), associated with mutation in the TMLHE gene (300777).

A locus on chromosome 2q (606053) associated with a phenotype including intellectual disability and speech deficits was formerly designated AUTS5.

Folstein and Rosen-Sheidley (2001) reviewed the genetics of autism.

The DSM-IV (American Psychiatric Association, 1994) specifies several diagnostic criteria for autism. In general, patients with autism exhibit qualitative impairment in social interaction, as manifest by impairment in the use of nonverbal behaviors such as eye-to-eye gaze, facial expression, body postures, and gestures, failure to develop appropriate peer relationships, and lack of social sharing or reciprocity. Patients have impairments in communication, such as a delay in, or total lack of, the development of spoken language. In patients who do develop adequate speech, there remains a marked impairment in the ability to initiate or sustain a conversation, as well as stereotyped or idiosyncratic use of language. Patients also exhibit restricted, repetitive and stereotyped patterns of behavior, interests, and activities, including abnormal preoccupation with certain activities and inflexible adherence to routines or rituals.

In his pioneer description of infantile autism, Kanner (1943) defined the disorder as 'an innate inability to form the usual, biologically provided affective contact with people.' Kanner (1943) noted that in most cases the child's behavior was abnormal from early infancy, and he suggested the presence of an inborn, presumably genetic, defect.

In a review, Smalley (1997) stated that mental retardation is said to be present in approximately 75% of cases of autism, seizures in 15 to 30% of cases, and electroencephalographic abnormalities in 20 to 50% of cases. In addition, approximately 15 to 37% of cases of autism have a comorbid medical condition, including 5 to 14% with a known genetic disorder or chromosomal anomaly. The 4 most common associations include fragile X syndrome (300624), tuberous sclerosis (see 191100), 15q duplications (AUTS4; 608636), and untreated phenylketonuria (PKU; 261600). Significant associations at a phenotypic level may reflect disruptions in a common neurobiologic pathway, common susceptibility genes, or genes in linkage disequilibrium.

The autism spectrum disorder shows a striking sex bias, with a male:female ratio of idiopathic autism estimated at 4-10:1, and with an increase in this ratio as the intelligence of the affected individuals increases (Folstein and Rosen-Sheidley, 2001).

Lainhart et al. (2002) stated that approximately 20% of children with autism appear to have relatively normal development during the first 12 to 24 months of life. This period of relative normalcy gradually or suddenly ends and is followed by a period of regression, characterized most prominently by a significant loss of language skills, after which the full autism syndrome becomes evident.

Rarely, children with autism may exhibit hyperlexia, or precocious reading (238350). Among a group of 66 children with pervasive developmental disorder, Burd et al. (1985) identified 4 with hyperlexia.

Cohen et al. (2005) discussed several genetic disorders consistently associated with autism, including fragile X syndrome, tuberous sclerosis, Angelman syndrome (105830), Down syndrome (190685), Sanfilippo syndrome (252900), Rett syndrome (312750) and other MECP2-related disorders, phenylketonuria, Smith-Magenis syndrome (SMS; 182290), 22q13 deletion syndrome (606232), Cohen syndrome (COH1; 216550), adenylosuccinate lyase deficiency (103050), and Smith-Lemli-Opitz syndrome (SLOS; 270400).

Miles et al. (2008) presented an expert-derived consensus measure of dysmorphic features often observed in patients with autism. The goal was to enable clinicians not trained in dysmorphology to use this classification system to identify and further subphenotype patients with autism. The measure includes 12 body areas that can be scored to arrive at a determination of dysmorphic or nondysmorphic. The body areas include stature, hair growth pattern, ear structure and placement, nose size, facial structure, philtrum, mouth and lips, teeth, hands, fingers and thumbs, nails, and feet. The model performed with 81 to 82% sensitivity and 95 to 99% specificity.

Constantino et al. (2017) showed that variation in viewing of social scenes, including levels of preferential attention and the timing, direction, and targeting of individual eye movements, is strongly influenced by genetic factors, with effects directly traceable to the active seeking of social information. In a series of eye-tracking experiments conducted with 338 toddlers, including 166 epidemiologically ascertained twins (enrolled by representative sampling from the general population), 88 nontwins with autism, and 84 singleton controls, Constantino et al. (2017) found high monozygotic twin-twin concordance (0.91) and relatively low dizygotic concordance (0.35). Moreover, the most highly heritable characteristics, preferential attention to eye and mouth regions of the face, are also those that are differentially decreased in children with autism (chi squared = 64.03, p less than 0.0001). The results of Constantino et al. (2017) implicated social visual engagement as a neurodevelopmental endophenotype not only for autism, but also for populationwide variation in social information seeking.

Folstein and Rutter (1977) reported that there had been no recorded cases of an autistic child having an overtly autistic parent; however, they noted that autistic persons rarely marry and rarely give birth. Folstein and Rutter (1977) stated that about 2% of sibs are affected, and that speech delay is common in the sibships containing autistic children. In a study of 21 same-sex twin pairs, 11 monozygotic (MZ) and 10 dizygotic (DZ), in which at least 1 had infantile autism, Folstein and Rutter (1977) found 36% concordance among the MZ twins and no concordance among the DZ twins. The concordance for cognitive abnormalities was 82% for MZ pairs and 10% for DZ pairs. In 12 of the 17 pairs discordant for autism, a biologic hazard liable to cause brain damage was identified. The authors concluded that brain injury in infancy may lead to autism on its own or in combination with a genetic predisposition. An inheritance pattern was not suggested.

In 40 pairs of twins, Ritvo et al. (1985) found a concordance rate for autism of 23.5% in dizygotic twins (4 of 17 pairs) and 95.7% in monozygotic twins (22 of 23 pairs). Ritvo et al. (1985) ascertained 46 families with 2 (n = 41) or 3 (n = 5) sibs with autism. Classic segregation analysis yielded a maximum likelihood estimate of the segregation ratio of 0.19 +/- 0.07, a value significantly different from 0.50 expected of an autosomal dominant trait and not significantly different from 0.25 expected of a recessive trait. The authors rejected a polygenic threshold model and suggested autosomal recessive inheritance.

Using the Utah Genealogical Database, Jorde et al. (1990) determined kinship for all possible pairs of autistic subjects. The average kinship coefficient for autistic subjects and controls showed a strong tendency for autism to cluster in families. However, the familial aggregation was confined exclusively to sib pairs and did not extend to more distant relatives. The authors concluded that the findings excluded recessive inheritance, since the autosomal recessive hypothesis would predict several first-cousin pairs, of which none were found. The rapid fall off in risk to relatives, as well as the sib risk of 4.5%, was consistent with multifactorial causation.

By analysis of 99 autistic probands and their families, Bolton et al. (1994) found an increased familial risk for both autism and more broadly defined pervasive developmental disorders in sibs, 2.9% and 2.9%, respectively, which is about 75 times higher than the risk in the general population.

In 27 same-sex pairs of monozygotic twins and 20 dizygotic twins, Bailey et al. (1995) found that 60% of monozygotic pairs were concordant for autism compared to 0% of dizygotic pairs. When they considered a broader spectrum of related cognitive or social abnormalities, 92% of monozygotic pairs were concordant compared to 10% of dizygotic pairs. The high concordance in monozygotes indicated a high degree of genetic control, and the rapid fall off of concordance in dizygotes suggested to Bailey et al. (1995) a multilocus, epistatic model. In the nonconcordant monozygotic pairs, there was a significantly higher incidence of obstetric complications, which the authors attributed to prenatal developmental anomalies, as evidenced by the very high incidence of minor congenital anomalies in the affected twins. They also reported an association of autism with increased head circumference.

In a sample of families selected because each had exactly 2 affected sibs, Greenberg et al. (2001) observed a remarkably high proportion of affected twin pairs, both MZ and DZ. Of 166 affected sib pairs, 30 (12 MZ, 17 DZ, and 1 of unknown zygosity) were twin pairs. Deviation from expected values was statistically significant; in a similarly ascertained sample of individuals with type I diabetes (222100), there was no deviation from expected values. Greenberg et al. (2001) noted that to ascribe the excess of twins with autism solely to ascertainment bias would require very large ascertainment factors; e.g., affected twin pairs would need to be approximately 10 times more likely to be ascertained than affected nontwin sib pairs. In the extreme situation of 'complete stoppage,' a form of ascertainment bias in which parents stop having children after the birth of their first affected child, the only families to have an affected sib pair would be those with an affected twin pair, or affected triplets. The authors suggested that risk factors related to twinning or to fetal development or other factors, genetic or nongenetic, in the parents may contribute to autism. Hallmayer et al. (2002) presented information refuting the suggestion that the twinning process itself is an important risk factor in the development of autism.

Silverman et al. (2002) analyzed 3 autistic symptom domains, social interaction, communication, and repetitive behaviors, and variability in the presence and emergence of useful phrase speech in 212 multiply affected sibships with autism. They found that the variance within sibships was reduced for the repetitive behavior domain and for delays in and the presence of useful phrase speech. These features and the nonverbal communication subdomain provided evidence of familiality when only the diagnosis of autism was considered for defining multiply affected sibships.

Kolevzon et al. (2004) studied specific features of autism for decreased variance in 16 families with monozygotic twins concordant for autism. Using regression analysis, they demonstrated significant aggregation of symptoms in monozygotic twins for 2 autistic symptom domains: impairment in communication and in social interaction. Kolevzon et al. (2004) stated that selecting probands according to specific features known to show reduced variance within families may provide more homogeneous samples for genetic analysis.

Awadalla et al. (2010) hypothesized that deleterious de novo mutations may play a role in cases of ASD and schizophrenia (181500), 2 etiologically heterogeneous disorders with significantly reduced reproductive fitness. Awadalla et al. (2010) presented a direct measure of the de novo mutation rate (mu) and selective constraints from de novo mutations estimated from a deep resequencing dataset generated from a large cohort of ASD and schizophrenia cases (n = 285) and population control individuals (n = 285) with available parental DNA. A survey of approximately 430 Mb of DNA from 401 synapse-expressed genes across all cases and 25 Mb of DNA in controls found 28 candidate de novo mutations, 13 of which were cell line artifacts. Awadalla et al. (2010) calculated a direct neutral mutation rate (1.36 x 10(-8)) that was similar to previous indirect estimates, but they observed a significant excess of potentially deleterious de novo mutations in ASD and schizophrenia individuals. Awadalla et al. (2010) concluded that their results emphasized the importance of de novo mutations as genetic mechanisms in ASD and schizophrenia and the limitations of using DNA from archived cell lines to identify functional variants.

Sandin et al. (2014) examined the familial risk of autism in a population-based cohort of 2,049,973 Swedish children born from 1982 to 2006. They identified 37,570 twin pairs; 2,642,064 full-sib pairs; 432,281 maternal and 445,531 paternal half-sib pairs; and 5,799,875 cousin pairs. Diagnoses of ASD to December 31, 2009 were ascertained. Exposure refers to the presence or absence of autism in a sib. In the sample, 14,516 children were diagnosed with ASD, of whom 5,689 had autistic disorder. The relative recurrence risk (RRR) and rate per 100,000 person-years for ASD among monozygotic twins was estimated to be 153.0 (95% CI, 56.7-412.8; rate, 6,274 for exposed vs 27 for unexposed); for dizygotic twins, 8.2 (95% CI, 3.7-18.1; rate, 805 for exposed vs 55 for unexposed); for full sibs, 10.3 (95% CI, 9.4-11.3; rate, 829 for exposed vs 49 for unexposed); for maternal half sibs, 3.3 (95% CI, 2.6-4.2; rate, 492 for exposed vs 94 for unexposed); for paternal half sibs, 2.9 (95% CI, 2.2-3.7; rate, 371 for exposed vs 85 for unexposed); and for cousins, 2.0 (95% CI, 1.8-2.2; rate, 155 for exposed vs 49 for unexposed). The RRR pattern was similar for autistic disorder but of slightly higher magnitude. Sandin et al. (2014) found support for a disease etiology including only additive genetic and nonshared environmental effects. The ASD heritability was estimated to be 0.50 (95% CI, 0.45-0.56) and the autistic disorder heritability was estimated to 0.54 (95% CI, 0.44-0.64). Sandin et al. (2014) concluded that among children in Sweden, the individual risk of ASD and autistic disorder increased with increasing genetic relatedness.

Iossifov et al. (2014) applied whole-exome sequencing to more than 2,500 simplex families each having a child with an autistic spectrum disorder. By comparing affected to unaffected sibs, Iossifov et al. (2014) showed that 13% of de novo missense mutations and 43% of de novo likely gene-disrupting mutations contribute to 12% and 9% of diagnoses, respectively. Including CNVs, coding de novo mutations contribute to about 30% of all simplex and 45% of female diagnoses.

AUTS1 Locus on Chromosome 7q22

By analyzing 125 autistic sib pairs, the International Molecular Genetic Study of Autism Consortium (2001) found a maximum multipoint lod score of 2.15 at marker D7S477 on chromosome 7q22, whereas analysis of 153 sib pairs generated a maximum multipoint lod score of 3.37. Linkage disequilibrium mapping identified 2 regions of association: one was under the peak of linkage, the other was 27 cM distal. In another study, the International Molecular Genetic Study of Autism Consortium (2001) found a multipoint maximum lod score of 3.20 at marker D7S477. They also detected a multipoint maximum lod score of 4.80 at marker D2S188 on chromosome 2q.

In 12 of 105 families with 2 or more sibs affected with autism, Yu et al. (2002) identified deletions ranging from 5 to more than 260 kb. One family had complex deletions at marker D7S630 on 7q21-q22, 3 families had different deletions at D7S517 on 7p, and another 3 families had different deletions at D8S264 on 8p. Yu et al. (2002) suggested that autism susceptibility alleles may cause the deletions by inducing errors during meiosis.

In a metaanalysis of 9 published genome scans on autism or autism spectrum disorders, Trikalinos et al. (2006) found evidence for significant linkage to 7q22-q32, confirming the findings of previous studies. The flanking region 7q32-qter reached a less stringent threshold for significance.

Genetic Heterogeneity

Using findings from a family study of autism and a similar study of twins, Pickles et al. (1995) concluded that autism has a multiple locus mode of inheritance involving 3 loci. Risch et al. (1999) performed a 2-stage genomewide screen of 2 groups of families with autism: 90 families comprising 97 affected sib pairs (ASPs) and 49 families with 50 affected sib pairs. Unaffected sibs, which provided 51 discordant sib pairs (DSPs) for the initial screen and 29 for the follow-up, were included as controls. There was a slightly increased identity by descent (IBD) in the ASPs (sharing of 51.6%) compared with the DSPs (sharing of 50.8%). The results were considered most compatible with a model specifying a large number of loci, perhaps 15 or more, and less compatible with models specifying 10 or fewer loci. The largest lod scores obtained were for a marker on 1p yielding a maximum multipoint lod score of 2.15, and on 17p, yielding a maximum lod score of 1.21.

In 51 multiplex families with autism, Philippe et al. (1999) used nonparametric linkage analysis to perform a genomewide screen with 264 microsatellite markers. By 2-point and multipoint affected sib-pair analyses, 11 regions gave nominal P values of 0.05 or lower. Philippe et al. (1999) observed overlap of 4 of these regions with regions on 2q, 7q, 6p, and 19p that had been identified by the earlier genomewide scan of autism conducted by the International Molecular Genetic Study of Autism Consortium (1998). The most significant multipoint linkage was close to marker D6S283 (maximum lod score = 2.23, p = 0.0013).

Smalley (1997) reported on the status of linkage studies in autism. Lamb et al. (2000) reviewed chromosomal aberrations, candidate gene studies, and linkage studies of autism.

Liu et al. (2001) genotyped 335 microsatellite markers in 110 multiplex families with autism. All families included at least 2 affected sibs, at least 1 of whom had autism; the remaining affected sibs carried diagnoses of either Asperger syndrome or pervasive developmental disorder. Affected sib-pair analysis yielded multipoint maximum lod scores that reached the accepted threshold for suggestive linkage on chromosomes 5, X, and 19. Further analysis yielded impressive evidence for linkage of autism and autism-spectrum disorders to markers on chromosomes 5 and 8, with suggestive linkage to a marker on chromosome 19.

Yonan et al. (2003) followed up on previously reported genomewide screens for autism performed by Liu et al. (2001) and Alarcon et al. (2002) showing suggestive evidence for linkage of autism spectrum disorders on chromosomes 5, 8, 16, 19, and X, and nominal evidence on several additional chromosomes. In their analysis, Yonan et al. (2003) increased the sample size 3-fold. Multipoint maximum lod scores obtained from affected sib-pair analysis of all 345 families yielded suggestive evidence for linkage on chromosomes 17, 5, 11, 4, and 8 (listed in order of MLS). The most significant findings were an MLS of 2.83 on 17q11 (AUTS6; 609378) and an MLS of 2.54 on 5p.

The genetic architecture of autism spectrum disorders (ASDs) is complex, requiring large samples to overcome heterogeneity. The Autism Genome Project Consortium (2007) broadened coverage and sample size relative to other studies of ASDs by using Affymetrix 10K SNP arrays and 1,181 families with at least 2 affected individuals, performing the largest linkage scan to that time while also analyzing copy number variation in these families. Linkage and copy number variation analyses implicated 11p13-p12 and neurexins, respectively, among other candidate loci. Neurexins teamed with previously implicated neuroligins for glutamatergic synaptogenesis, highlighting glutamate-related genes as promising candidates for contributing to ASDs. See neuroligin-3 (NLGN3; 300336) and neuroligin-4 (NLGN4; 300427).

Wang et al. (2009) presented the results of a genomewide association study of ASDs on a cohort of 780 families (3,101 subjects) with affected children, and a second cohort of 1,204 affected subjects and 6,491 control subjects, all of whom were of European ancestry. Six SNPs on chromosome 5p14.1 between cadherin-10 (CDH10; 604555) and cadherin-9 (CDH9; 609974), 2 genes encoding neuronal cell adhesion molecules, revealed strong association signals, with the most significant SNP being rs4307059 (p = 3.4 x 10(-8), odds ratio = 1.19). These signals were replicated in 2 independent cohorts, with combined P values ranging from 7.4 x 10(-8) to 2.1 x 10(-10). The authors concluded that their results implicated neuronal cell adhesion molecules in the pathogenesis of ASDs.

In a genomewide association study of 438 Caucasian families including 1,390 individuals with autism, Ma et al. (2009) found evidence for linkage to chromosome 5p14.1. Validation in an additional cohort of 2,390 samples from 457 families showed that 8 SNPs on chromosome 5p14.1 were significantly associated with autism (p values ranging from 3.24 x 10(-4) to 3.40 x 10(-6)). The most significant linkage was with rs10038113.

Using array CGH analysis, Roohi et al. (2009) identified a chromosome 3 copy number variation (CNV) disrupting the CNTN4 gene (607280) in 3 of 92 individuals with autism spectrum disorder. Two sibs had a deletion, and another unrelated individual had a duplication; both variations were inherited from an unaffected father. A third affected sib of the familial cases did not carry the deletion, suggesting incomplete penetrance or that he had a different disorder. The changes resulted from Alu-mediated unequal recombinations.

Glessner et al. (2009) presented the results from a whole-genome CNV study on a cohort of 859 ASD cases and 1,409 healthy children of European ancestry who were genotyped with approximately 550,000 SNP markers, in an attempt to comprehensively identify CNVs conferring susceptibility to ASDs. Positive findings were evaluated in an independent cohort of 1,336 ASD cases and 1,110 controls of European ancestry. Glessner et al. (2009) confirmed known associations, such as that with NRXN1 (600565) and CNTN4 (Roohi et al., 2009), in addition to several novel susceptibility genes encoding neuronal cell adhesion molecules, including NLGN1 (600568) and ASTN2 (612856), that were enriched with CNVs in ASD cases compared to controls (P = 9.5 x 10(-3)). Furthermore, CNVs within or surrounding genes involved in the ubiquitin pathways, including UBE3A (601623), PARK2 (602544), RFWD2 (608067), and FBXO40 (609107), were affected by CNVs not observed in controls (p = 3.3 x 10(-3)). Glessner et al. (2009) also identified duplications 55 kb upstream of complementary DNA AK123120 (p = 3.6 x 10(-6)). Glessner et al. (2009) concluded that although these variants may be individually rare, they appear to target genes involved in neuronal cell-adhesion or ubiquitin degradation, indicating that these 2 important gene networks expressed within the central nervous system (CNS) may contribute to genetic susceptibility to ASD.

Weiss et al. (2009) initiated a linkage and association mapping study using half a million genomewide SNPs in a common set of 1,031 multiplex autism families (1,553 affected offspring). They identified regions of suggestive and significant linkage on chromosomes 6q27 and 20p13, respectively. Initial analysis did not yield genomewide significant associations; however, genotyping of top hits in additional families revealed an SNP on chromosome 5p15 (rs10513025) between SEMA5A (609297) and TAS2R1 (604796) that was significantly associated with autism (p = 2.0 x 10(-7)). Weiss et al. (2009) also demonstrated that expression of SEMA5A is reduced in brains from autistic patients, further implicating SEMA5A as an autism susceptibility gene.

Kilpinen et al. (2009) carried out a genomewide microsatellite-based scan of a unique extended Finnish autism pedigree comprised of 20 families with verified genealogic links reaching back to the 17th century. Linkage analysis and fine mapping revealed significant results for SNPs (rs4806893, rs216283, and rs216276) on chromosome 19p13.3 in close proximity to TLE2 (601041) and TLE6 (612399) genes. They also obtained a significant result for a SNP rs1016732 on chromosome 1q23 near the ATP1A2 gene (182340). Kilpinen et al. (2009) noted that chromosome 1q23 had been previously reported as an autism susceptibility locus in Finnish families.

Exclusion Studies

In a multicenter study in Sweden, Blomquist et al. (1985) found the fragile X repeat (309550) in 13 of 83 boys (16%) with infantile autism, but in none of 19 girls with infantile autism.

Using the UCLA Registry for Genetic Studies of Autism, Spence et al. (1985) studied 46 families with at least 2 affected children. Linkage studies in 34 families showed no evidence of linkage with HLA (142800), and close linkage with 19 other autosomal markers was excluded. The highest lod score, 1.04, was found with haptoglobin (140100) on chromosome 16q22 at recombination fractions of 10% in males and 50% in females. There was no association of the disorder with fragile X.

Using data from 38 multiplex families with autism to perform a multipoint linkage analysis with markers on the X chromosome, Hallmayer et al. (1996) excluded a moderate to strong gene effect causing autism on the X chromosome.

Using the Autism Diagnostic Instrument-Revised (ADI-R), the Autism Diagnostic Observation Scale (ADOS), and psychometric tests, Klauck et al. (1997) identified 141 autistic patients from 105 simplex and 18 multiplex families; 131 patients met all 4 ADI-R algorithm criteria for autism and 10 patients showed a broader phenotype of autism. Using amplification of the CCG repeat at the fragile X locus, hybridization to the complete FMR1 cDNA probe, and hybridization to additional probes from the neighborhood of the FMR1 gene, the authors found no significant changes in 139 patients (99%) from 122 families. In 1 multiplex family with 3 children showing no dysmorphic features of the fragile X syndrome (1 male meeting 3 of 4 ADI-algorithm criteria, 1 normal male with slight learning disability but negative ADI-R testing, and 1 fully autistic female), the FRAXA full-mutation-specific CCG-repeat expansion in the genotype was not correlated with the autism phenotype. Further analysis revealed a mosaic pattern of methylation at the FMR1 gene locus in the 2 sons of the family, indicating at least a partly functional gene. Klauck et al. (1997) concluded that the association of autism with fragile X at Xq27.3 is nonexistent and excluded this location as a candidate gene for autism.

Lopreiato and Wulfsberg (1992) described a complex chromosomal rearrangement in a 6.5-year-old boy with autism who was otherwise normal except for minimal dysmorphism. The rearrangement seen in every cell examined involved chromosomes 1, 7 and 21: 46, XY, -1, -7, -21, t(1;7;21)(1p22.1-qter::21q22.3-qter; 7pter-q11.23::7q36.1-qter; 21pter-q22.3::7q11.23-q36.1::1pter-p22.1).

Vincent et al. (2006) reported 2 brothers with autism who both carried a paracentric inversion of chromosome 4p, inv(4)(p12-p15.3), inherited from an unaffected mother and unaffected maternal grandfather. More detailed molecular analysis showed that the proximal breakpoint on 4p12 involved a cluster of GABA receptor genes, including the GABRA4 gene (137141), which has been implicated in autism (Ma et al., 2005; Collins et al., 2006). Maestrini et al. (1999) found no association or linkage to the GABRB3 gene in 94 families comprising 174 individuals with autism.

Moessner et al. (2007) identified deletions in the SHANK3 gene (606230) on chromosome 22q13 in 3 (0.75%) of 400 unrelated patients with an autism spectrum disorder. The deletions ranged in size from 277 kb to 4.36 Mb; 1 patient also had a 1.4-Mb duplication at chromosome 20q13.33. The patients were essentially nonverbal and showed poor social interactions and repetitive behaviors. Two had global developmental delay and mild dysmorphic features. A fourth patient with a de novo missense mutation in the SHANK3 gene had autism-like features but had diagnostic scores above the cutoff for autism; she was conceived by in vitro fertilization. See also the chromosome 22q13.3 deletion syndrome (606232).

Brandler et al. (2018) hypothesized that rare inherited structural variants in cis-regulatory elements of variant-intolerant genes may contribute to ASDs, and investigated this by assessing the evidence for natural selection and transmission distortion of cis-regulatory element structural variants in whole genomes of 9,274 subjects from 2,600 families affected by ASD. In a discovery cohort of 829 families, structural variants were depleted within promoters and untranslated regions, and paternally inherited cis-regulatory element structural variants were preferentially transmitted to affected offspring and not to their unaffected sibs. The association of paternal cis-regulatory element structural variants was replicated in an independent sample of 1,771 families. Brandler et al. (2018) detected paternally inherited deletions of the LEO1 (610507) promoter in 3 affected individuals, 1 trio (14-59) and 1 concordant sib pair (F0182). Fibroblast expression of both LEO1 and the neighboring MAPK6 (602904) were increased in deletion carriers. Brandler et al. (2018) noted that 2 de novo loss-of-function variants disrupting LEO1 had been observed in a combined exome data set of ASD and developmental delay from 20 studies (p = 0.0025) (De Rubeis et al., 2014; Deciphering Developmental Disorders Study, 2017). Brandler et al. (2018) concluded that rare inherited noncoding variants predispose children to ASD, with differing contributions from each parent.

Copy Number Variation

Using high-resolution microarray analysis, Marshall et al. (2008) found 277 unbalanced copy number variations (CNV), including deletion, duplication, translocation, and inversion, in 189 (44%) of 427 families with autism spectrum disorder (ASD). These specific changes were not present in a total of about 1,600 controls, although control individuals also carried many CNV. Although most variants were inherited among the patients, 27 cases had de novo alterations, and 3 (11%) of these individuals had 2 or more changes. Marshall et al. (2008) detected 13 loci with recurrent or overlapping CNV in unrelated cases. Of note, CNV at chromosome 16p11.2 (AUTS14; see 611913) was identified in 4 (1%) of 427 families and none of 1,652 controls (p = 0.002). Some of the autism loci were also common to mental retardation loci. Marshall et al. (2008) concluded that structural gene variants were found in sufficiently high frequency influencing autism spectrum disorder to suggest that cytogenetic and microarray analyses be considered in routine clinical workup.

Cusco et al. (2009) analyzed 96 Spanish patients with idiopathic ASD by array CGH analysis. Only 13 of the 238 detected CNVs (range, 89 kb-2.4 Mb) were present specifically in 12 (12.5%) of 96 ASD patients. The CNVs consisted of 10 duplications and 3 deletions. In 5 patients with parental samples available, CNVs were inherited from a normal parent. Two CNVs mapped to regions with previously reported ASD candidates, KIAA0442 (607270) on chromosome 7q11.22 and GRM8 (601116) on chromosome 7q31.3. Of the 24 genes in the CNVs, some act in common pathways, most notably the phosphatidylinositol signaling and the glutamatergic synapse, both known to be affected in several genetic syndromes related with autism and previously associated with ASD. Cusco et al. (2009) hypothesized that functional alteration of genes in related neuronal networks is involved in the etiology of the ASD phenotype.

Sebat et al. (2007) tested the hypothesis that de novo CNV is associated with autism spectrum disorders. They performed comparative genomic hybridization (CGH) on the genomic DNA of patients and unaffected subjects to detect copy number variants not present in their respective parents. Candidate genomic regions were validated by higher-resolution CGH, paternity testing, cytogenetics, fluorescence in situ hybridization, and microsatellite genotyping. Confirmed de novo CNVs were significantly associated with autism (p = 0.0005). Such CNVs were identified in 12 of 118 (10%) of patients with sporadic autism, in 2 of 77 (3%) of patients with an affected first-degree relative, and in 2 of 196 (1%) of controls. The authors stated that most de novo CNVs were smaller than microscopic resolution. Affected genomic regions were highly heterogeneous and included mutations of single genes. Sebat et al. (2007) concluded that their findings established de novo germline mutation as a more significant risk factor for autism spectrum disorders than previously recognized.

Pinto et al. (2010) analyzed the genomewide characteristics of rare (less than 1% frequency) CNVs in ASD using dense genotyping arrays. When comparing 996 ASD individuals of European ancestry to 1,287 matched controls, cases were found to carry a higher global burden of rare, genic CNVs (1.19-fold, p = 0.012), especially so for loci previously implicated in either ASD and/or intellectual disability (1.69-fold, p = 3.4 x 10(-4)). Among the CNVs there were numerous de novo and inherited events, sometimes in combination in a given family, implicating many novel ASD genes such as SHANK2 (603290), SYNGAP1 (603384), DLGAP2 (605438), and the X-linked DDX53-PTCHD1 locus (see 300828). The authors also discovered an enrichment of CNVs disrupting functional gene sets involved in cellular proliferation, projection and motility, and GTPase/Ras signaling.

Levy et al. (2011) studied 887 families from the Simons Simplex Collection of relatively high functioning ASD families. They identified 75 de novo CNVs in 68 probands (approximately 8% of probands). Only a few were recurrent. Variation at the 16p11.2 locus was detected in more than 1% of patients (10 of 858), with deletions present in 6 and duplications in 4. In addition, the duplication at 7q11.2 of the Williams syndrome region (609757) was also seen as a recurrent CNV. Levy et al. (2011) noted that the finding of 8% of ASD probands with de novo events compared with 2% in unaffected sibs was in keeping with other reports. They speculated that the slightly lower incidence of 8% rather than the 10% reported by Sebat et al. (2007) related to a higher functioning autism group in this cohort or to smaller families in this study. Using their study and the previous literature, Levy et al. (2011) proposed a list of 'asymmetries,' or observed biases, in simplex families with autism. There is a higher incidence of de novo copy number mutation in children with ASDs from simplex families than in their sibs. There is a higher incidence of de novo copy number mutation in children with ASDs from simplex families than in children with ASDs from multiplex families. For transmitted rare events, duplications greatly outweigh deletions; deletions outweigh duplications among de novo events in children with ASDs. There is evidence of transmission distortion for ultrarare events to children with ASDs; this bias arises from families in which the sib is an unaffected male. Females are less likely to be diagnosed with ASDs than are males. A higher proportion of females with ASDs have detectable de novo copy number events than do males with ASDs, and the events are larger. Levy et al. (2011) suggested that females are protected from autism but did not propose a mechanism.

Sanders et al. (2011) examined 1,124 ASD simplex families from the Simons Simplex Collection. Each of the families was comprised of a single proband, unaffected parents, and in most kindreds an unaffected sib. Sanders et al. (2011) found significant association of ASD with de novo duplications of 7q11.23. They also identified rare recurrent de novo CNVs at 5 additional regions, including 16p13.2. Overall, large de novo CNVs conferred substantial risks (odds ratio = 5.6; CI = 2.6-12.0, p = 2.4 x 10(-7)). Sanders et al. (2011) suggested that there are 130 to 234 ASD-related CNV regions in the human genome and presented compelling evidence, based on cumulative data, for association of rare de novo events at 7q11.23, 15q11.2-q13.1 (see 608636), 16p11.2, and neurexin-1 (600565). Sanders et al. (2011) found that probands carrying a 16p11.2 or 7q11.23 de novo CNV were indistinguishable from the larger ASD group with respect to IQ, ASD severity, or categorical autism diagnosis. However, they did find a relationship between body weight and 16p11.2 deletions and duplications. When copy number was treated as an ordinal variable, BMI diminished as 16p11.2 copy number increased (p = 0.02).

Vaags et al. (2012) reported 4 families in which 1 or more members had autism spectrum disorder associated with heterozygous deletions of chromosome 14q affecting the NRXN3 gene (600567). The deletions were all different and ranged from 63 to 336 kb. One deletion affected only the NRXN3 alpha isoform, whereas 3 affected both the alpha and beta isoforms. Two families were ascertained from 1,158 Canadian individuals with ASD who were screened for copy number variations across the genome. The third family was 1 of 1,368 ASD cases screened, and the fourth was 1 of 1,796 ASD cases screened. The phenotype was variable, ranging from high-functioning Asperger syndrome to full autism with some pervasive developmental and behavioral problems. In 1 family, the deletion occurred de novo. In the other families, the deletion was inherited from a parent; 1 parent had a broader autism phenotype, 1 self-reported mild autistic-like features, and 1 was normal. In 1 family, 2 of 3 trizygotic triplets with autism carried the deletion; the third unaffected child did not carry the deletion. Small deletions affecting only the alpha isoform were found in 4 of 15,122 controls. The report suggested that deletions affecting the NRXN3 gene may predispose to the development of autism spectrum disorder, but segregation patterns within the families suggested issues of penetrance and expressivity at this locus.

Loirat et al. (2010) reported 3 unrelated boys with heterozygous de novo deletions in chromosome 17q12 (see 614527) who had cystic or hyperechogenic kidneys and autism. Their 17q12 deletions ranged from 1.5 to 1.8 Mb, and included LHX1 (601999), HNF1B (189907), and 19 other genes; sequencing of the LHX1 gene in the 3 boys and 32 control patients with autism revealed no mutations. Loirat et al. (2010) concluded that autism might be an additional manifestation associated with HNF1B deletion.

Moreno-De-Luca et al. (2010) performed cytogenomic array analysis in a discovery sample of patients with neurodevelopmental disorders and detected a recurrent 1.4-Mb deletion at chromosome 17q12 in 18 of 15,749 patients, including 6 with autism or autistic features; the deletion was not found in 4,519 controls. In a large follow-up sample, the same deletion was identified in 2 of 1,182 patients with autism spectrum disorder and/or neurocognitive impairment, and in 4 of 6,340 schizophrenia (see 181500) patients, but was not found in 47,929 controls (corrected p = 7.37 x 10 (-5)). Moreno-De-Luca et al. (2010) concluded that deletion 17q12 is a recurrent, pathogenic CNV that confers a high risk for autism spectrum disorder and schizophrenia, and that 1 or more of the 15 genes in the deleted interval is dosage-sensitive and essential for normal brain development and function.

Luo et al. (2012) interrogated gene expression in lymphoblasts from 439 individuals from 244 families with discordant sibs in the Simons Simplex Collection and found that the overall frequency of significantly misexpressed genes, which they referred to as outliers, did not differ between probands and unaffected sibs. However, in probands, but not their unaffected sibs, the group of outlier genes was significantly enriched in neural-related pathways, including neuropeptide signaling, synaptogenesis, and cell adhesion. The outlier genes clustered within large rare de novo CNVs and could be used for the prioritization of rare CNVs of potential significance. Several nonrecurrent CNVs with significant gene expression alterations were identified, including deletions in chromosome regions 3q27, 3p13, and 3p26 and duplications at 2p15, suggesting these as potential ASD loci.

See SHANK1 (604999) for discussion of a possible association between heterozygous deletions involving the SHANK1 gene on chromosome 19q13 and susceptibility to high-functioning autism.

Krumm et al. (2013) searched for disruptive, genic rare CNVs among 411 families affected by sporadic autism spectrum disorder from the Simons Simplex Collection by using available exome sequence data and CoNIFER (Copy Number Inference from Exome Reads). Compared to high density SNP microarrays, the authors' approach yielded approximately 2 times more smaller genic rare CNVs. Krumm et al. (2013) found that affected probands inherited more CNVs than did their sibs (453 vs 394, p = 0.004; odds ratio = 1.19) and that the probands' CNVs affected more genes (921 vs 726, p = 0.02; odds ratio = 1.30). These smaller CNVs (median size 18 kb) were transmitted preferentially from the mother (136 maternal vs 100 paternal, p = 0.02), although this bias occurred irrespective of affected status. The excess burden of inherited CNVs was driven primarily by sib pairs with discordant social behavior phenotypes, which contrasts with families where the phenotypes were more closely matched or less extreme. In a combined model, the inherited CNVs, de novo CNVs, and de novo single-nucleotide variants all independently contributed to the risk of autism (p less than 0.05).

Poultney et al. (2013) used the eXome Hidden Markov Model (XHMM) as well as transmission information and validation by molecular methods to confirm that small CNVs encompassing as few as 3 exons can be reliably called from whole-exome data. They applied this approach to an autism case-control sample of 811 subjects (mean per-target read depth = 161) and observed a significant increase in the burden of rare (minor allele frequency (MAF) 1% or less) 1- to 30-kb CNVs, 1- to 30-kb deletions, and 1- to 10-kb deletions in ASD. CNVs in the 1 to 30 kb range frequently hit just a single gene, allowing Poultney et al. (2013) to observe enrichment for disruption of genes in cytoskeletal and autophagy pathways in ASD. Poultney et al. (2013) concluded that rare 1- to 30-kb exonic deletions could contribute to risk in up to 7% of individuals with ASD.

Girirajan et al. (2013) exploited the repeat architecture of the genome to target segmental duplication-mediated rearrangement hotspots (n = 120, median size 1.78 Mbp, range 240 kbp to 13 Mbp) and smaller hotspots flanked by repetitive sequence (n = 1,247, median size 79 kbp, range 3-96 kbp) in 2,588 autistic individuals from simplex and multiplex families and in 580 controls. The analysis identified several recurrent large hotspot events, including association with 1q21 duplications, which are more likely to be identified in individuals with autism than in those with developmental delay (p = 0.01; odds ratio = 2.7). Within larger hotspots, Girirajan et al. (2013) also identified smaller atypical CNVs that implicated CHD1L (613039) and ACACA (200350) for the 1q21 and 17q12 deletions, respectively. The analysis, however, suggested no overall increase in the burden of smaller hotspots in autistic individuals as compared to controls. By focusing on gene-disruptive events, Girirajan et al. (2013) identified several genes that were enriched for CNVs in autism cases versus controls, including DPP10 (608209), PLCB1 (607120), TRPM1 (603576), NRXN1 (600565), FHIT (601153), and HYDIN (610812). Girirajan et al. (2013) found that as the size of deletions increases, nonverbal IQ significantly decreases, but there is no impact on autism severity; as the size of duplications increases, autism severity significantly increases but nonverbal IQ is not affected. Girirajan et al. (2013) concluded