https://www.hal.inserm.fr/inserm-00549873/document

Etiological heterogeneity in autism spectrum disorders: More than 100 genetic and genomic disorders and still counting

Since autism was first recognized as a disorder in 1943, speculation about its aetiology has ranged from biological to psychological and back again. After twin studies during the 1970s and 1980s yielded unequivocal evidence for a genetic component, aetiological research in autism began to focus primarily on uncovering the genetic mechanisms involved. The identification of chromosomal abnormalities and Mendelian syndromes among individuals with autism, in conjunction with data from genome screens and candidate-gene studies, has helped to refine the view of the complex genetics that underlies autism spectrum conditions.

Genetics of autism: complex aetiology for a heterogeneous disorder.

Angelman syndrome (AS) is a human genetic disorder characterized by mental retardation, seizures, inappropriate laughter, abnormal galt, tremor and ataxia. There is strong genetic evidence that the disorder is associated with a maternally expressed, imprinted gene mapping to chromosome 15q11-13. Affected patients demonstrate varied molecular abnormalities, including large maternal deletions, uniparental paternal disomy (UPD). Imprinting mutations and loss of function mutations of E6-associated-protein (E6-AP) ubiquitin-protein ligase (UBE3A). All of these abnormalities are associated with loss of maternal expression of UBE3A. Although mutations in UBE3A cause AS, indicating that maternal-specific expression of UBE3A is essential for a normal phenotype, evidence for maternal-specific expression of UBE3A has been lacking. Using mice with partial paternal UPD encompassing Ube3a to differentiate maternal and paternal expression, we found by in situ hybridization that expression of Ube3a in Purkinje cells, hippocampal neurons and mitral cells of the olfactory bulb in UPD mice was markedly reduced compared to non-UPD littermates. In contrast, expression of Ube3a in other regions of the brain was only moderately or not at all reduced in UPD mice. The major phenotypic features of AS correlate with the loss of maternal-specific expression of Ube3a in hippocampus and cerebellum as revealed in the mouse model.

Imprinted expression of the murine Angelman syndrome gene, Ube3a , in hippocampal and Purkinje neurons

Of the chronic mental disabilities of childhood, autism is causally least well understood. The former view that autism was rooted in exposure to humorless and perfectionistic parenting has given way to the notion that genetic influences are dominant underlying factors. Still, identification of specific heritable factors has been slow with causes identified in only a few cases in unselected series. A broad search for genetic and environmental influences that cause or predispose to autism is the major thrust of the South Carolina Autism Project. Among the first 100 cases enrolled in the project, abnormalities of chromosome 15 have emerged as the single most common cause. The four abnormalities identified include deletions and duplications of proximal 15q. Other chromosome aberrations seen in single cases include a balanced 13;16 translocation, a pericentric inversion 12, a deletion of 20p, and a ring 7. Candidate genes involved in the 15q region affected by duplication and deletion include the ubiquitin-protein ligase (UBE3A) gene responsible for Angelman syndrome and genes for three GABA(A) receptor subunits. In all cases, the deletions or duplications occurred on the chromosome inherited from the mother.

Autism and maternally derived aberrations of chromosome 15q

We undertook clinical evaluation (32 cases) and molecular evaluation (31 cases) of unrelated patients affected with Smith-Magenis syndrome (SMS) associated with an interstitial deletion of band p11.2 of chromosome 17. Patients were evaluated both clinically and electrophysiologically for peripheral neuropathy, since markers showing close linkage to one form of Charcot-Marie-Tooth disease (CMT1A) map to this chromosomal region. The common clinical findings were broad flat midface with brachycephaly, broad nasal bridge, brachydactyly, speech delay, and hoarse, deep voice. Fifty-five percent of the patients showed clinical signs (e.g., decreased or absent deep tendon reflexes, pes planus or pes cavus, decreased sensitivity to pain, and decreased leg muscle mass) suggestive of peripheral neuropathy. However, unlike patients with CMT1A, these patients demonstrated normal nerve conduction velocities. Self-destructive behaviors, primarily onychotillomania and polyembolokoilamania, were observed in 67% of the patients, and significant symptoms of sleep disturbance were observed in 62%. The absence of REM sleep was demonstrated by polysomnography in two patients. Southern analysis indicated that most patients were deleted for five 17p11.2 markers--FG1 (D17S446), 1516 (D17S258), pYNM67-R5 (D17S29), pA10-41 (D17S71), and pS6.1-HB2 (D17S445)--thus defining a region which appears to be critical to SMS. The deletion was determined to be of paternal origin in nine patients and of maternal origin in six patients. The apparent random parental origin of deletion documented in 15 patients suggests that genomic imprinting does not play a role in the expression of the SMS clinical phenotype. Our findings suggest that SMS is likely a contiguous-gene deletion syndrome which comprises characteristic clinical features, developmental delay, clinical signs of peripheral neuropathy, abnormal sleep function, and specific behavioral anomalies.

Molecular analysis of the Smith-Magenis syndrome: a possible contiguous-gene syndrome associated with del(17)(p11.2).

Six persons with the classical Angelman syndrome (AS) phenotype and de novo deletions of chromosome 15q11-q13 were studied to determine the parental origin of the chromosome deletion. Four of the 6 patients had informative cytogenetic studies and all demonstrated maternal inheritance of the deletion. These findings, together with other reported cases of the origin of the chromosome 15 deletion in AS, suggest that deletion of the maternally contributed chromosome leads to the AS phenotype. This contrasts with the Prader-Willi syndrome (PWS) in which a similar deletion of the paternally contributed chromosome 15 is observed. In deletion cases, a parental gamete effect such as genomic imprinting may be the best model to explain why apparently identical 15q11-q13 deletions may develop the different phenotypes of AS or PWS.

Maternal origin of 15q11-13 deletions in Angelman syndrome suggests a role for genomic imprinting.

A cytogenetic survey of 67 individuals previously identified as having mental retardation and autistic behaviors revealed 1 person (1.5%) with the fragile X chromosome (fra[X]) and 3 (4.5%) with autosome abnormalities. This low prevalence of fra(X) indicates that most persons with fra(X) in this mental retardation center did not have autistic behaviors severe enough to be identified as a secondary psychiatric diagnosis. The presence of other chromosomal abnormalities is consistent with the known causal heterogeneity of autism in mental retardation populations.

J. W. Stone

Papers

Maternal origin of 15q11-13 deletions in Angelman syndrome suggests a role for genomic imprinting.

Cytogenetic Survey for Autistic Fragile X Carriers in a Mental Retardation Center.