DDX3X Syndrome/Intellectual Disability Type 102 via the DDX3X Gene

Patients with DDX3X syndrome (also known as X-linked intellectual disability type 102 (MRX102)) present in infancy or early childhood with intellectual disability (ID) and developmental delay that can range from mild to severe (Snijders Blok et al. 2015. PubMed ID: 26235985). Over 90% of identified patients are female. Brain abnormalities found on magnetic resonance imaging are a consistent feature of this disorder, with corpus callosum malformations being the most frequent (>95% of patients), followed by enlargement of the lateral ventricles, diminished white matter, and polymicrogyria (Lennox et al. 2018. bioRxiv 317974). A majority of patients also have behavior problems, which are usually characterized as autism, hyperactivity, or aggression. Some children gain the ability to speak in full sentences, although about half remain completely nonverbal by age five. Most patients have hypotonia, and a significant portion develop a movement disorder, which typically includes spasticity or unusual gait. Most patients exhibit some dysmorphic features, although a recognizable facial gestalt has not been established (Snijders Blok et al. 2015. PubMed ID: 26235985).

Other major features, observed in 20-50% of patients, include seizures, microcephaly, underweight for age, joint hyperlaxity, mosaic pigmentary changes, and vision problems (including strabismus, colobomas, and cortical visual impairment). Minor features (observed in 5-20% of patients) include precocious puberty, scoliosis, variable congenital heart defects, respiratory problems, cleft lip or palate, and hearing loss (both conductive and sensorineural). Additional features are seen in a small minority of patients (Lennox et al. 2018 bioRxiv 317974). While few affected males are described in the literature, their reported phenotypic features seem largely consistent with those seen in females.

There is no available treatment for this disorder, yet advantages of testing may include prognostic information, early identification and treatment of symptoms such as seizures and autism, and ability to join the DDX3X foundation and family support groups (https://ddx3x.org/). For families with inherited causative variants, prenatal testing or pre-implantation genetic diagnosis may be implemented for future pregnancies, whereas for families with affected females, knowledge of a de novo variant and decreased recurrence risk may ease anxiety for reproductive planning.

Pathogenic variants in the DDX3X gene cause DDX3X syndrome (Lennox et al. 2018. bioRxiv 317974). This syndrome is most commonly observed in females, who usually carry dominantly-acting de novo DDX3X variants. However, hemizygous males are also observed, with variants that are either de novo or inherited from unaffected mothers (Wang et al. 2018. bioRxiv 283598, Snijders Blok et al. 2015. PubMed ID: 26235985, Kellaris et al. 2018. PubMed ID: 29490693). None of the pathogenic variants observed in females have ever been reported in a male, suggesting these variants are embryonic lethal in the hemizygous state. Evidence suggests that pathogenic DDX3X variants account for roughly 1.5% of unexplained intellectual disability in females, making it one of the most common causes of unexplained ID in females (Wang et al. 2018. bioRxiv 283598). By contrast, DDX3X variants are estimated to account for less than 0.1% of unexplained ID in males. Unexplained intellectual disability indicates individuals whose ID is not caused by copy number variants detectable on array, fragile X triplet repeat expansion, or environmental insults. Additionally, these individuals may have one or more negative targeted gene sequencing tests (Snijders Blok et al. 2015. PubMed ID: 26235985, Lennox et al. 2018. bioRxiv 317974). DDX3X partially escapes X-inactivation, and studies suggest that X-inactivation status, including extreme skewing, is not a factor in determining the severity or type of symptoms observed in DDX3X patients (Lennox et al. 2018. bioRxiv 317974, Wang et al. 2018. bioRxiv 283598, Snijders Blok et al. 2015. PubMed ID: 26235985).

Pathogenic variant types include missense, splicing, nonsense, in-frame, and frameshift alterations. To date, no causative copy number variants have been reported. Pathogenic frameshift and nonsense variants span a majority of the gene, excepting the last exon, whereas causative missense changes cluster in the two helicase domains (amino acids 211-403 and 414-575) (www.uniprot.org/uniprot/O00571). Limited evidence suggests phenotype-genotype correlations for specific missense changes in the helicase domains (T532, I415, R326) causing the most severe form of this disorder (Lennox et al. 2018. bioRxiv 317974). Overall, this gene is highly intolerant of both missense and loss of function variation, as indicated by the paucity of these changes in large datasets of normal individuals (Snijders Blok et al. 2015. PubMed ID: 26235985). Some positions of the DDX3X sequence appear prone to recurrent de novo variants (notably amino acids R376 and A488); however a majority of reported DDX3X variants are unique. Interestingly, this gene as a whole may be particularly prone to de novo alterations, as indicated by a cohort from Baylor identifying this gene as one of the most likely genes to carry de novo changes (ranked third out of 450 genes)(Wang et. al. 2018. bioRxiv 283598).

The DDX3X gene is located at Xp11.4, and encodes a 661 amino acid DEAD-box RNA helicase protein important for a wide array of basic cellular processes including transcription, splicing, translation, and transport. It is expressed highly in the developing brain where it is implicated in neuronal differentiation and migration (Lennox et al. 2018. bioRxiv 317974).

This test is predicted to detect pathogenic variants in ~1.5% of females with unexplained intellectual disability. By contrast, the diagnostic yield of this test for males with unexplained ID is predicted to be <0.1% (Wang et al. 2018. bioRxiv 283598, Snijders Blok et al. 2015. PubMed ID: 26235985, Lennox et al. 2018. bioRxiv 317974). It is important to note the high clinical and genetic heterogeneity of intellectual disability, and improved diagnostic yields that result from testing large panels of genes as well as testing parents along with the patient using a trio approach (Vissers et al. 2016. PubMed ID: 26503795). Analytical sensitivity of this test should be high because all reported pathogenic variants in DDX3X are detectable by sequencing.

This test provides full coverage of all coding exons of the DDX3X gene, plus ~10 bases of flanking noncoding DNA. We define full coverage as >20X NGS reads or Sanger sequencing.