Insights into DDX3X syndrome from a novel mouse model with construct and face validity

DDX3X syndrome is a genetic disorder accounting for 1–3% of intellectual disability (ID) and developmental delay in females (1). Individuals with DDX3X syndrome present with ID, autism spectrum disorder (ASD), language delays, motor impairments including gait abnormalities and hypotonia, and structural brain malformations, notably in the corpus callosum, with white matter abnormalities (1). The clinical phenotype is heterogenous, with ID ranging from mild to severe, and approximately half of affected females continue to be nonverbal after 5 years of age (2). DDX3X is an X-linked gene that encodes a DEAD-box ATP-dependent RNA-helicase involved in regulation of transcription, translation, and RNA metabolism (3). Although DDX3X is on the X chromosome, it escapes X-inactivation, and females express both alleles, and thus can be heterozygous for deleterious variants (4). Pathogenic DDX3X variants in females are de novo loss of function mutations, pointing to haploinsufficiency as the mechanism underlying DDX3X syndrome (5). Few affected males have been identified, but many of them carry missense mutations inherited from asymptomatic heterozygous mothers, indicating that the pathogenic mechanism(s) of DDX3X syndrome in males may differ from that in female patients (2,5). Other affected males carry de novo pathogenic variants (2).

One of the primary challenges associated with generating a mouse model of DDX3X syndrome is the necessity of Ddx3x for embryonic development and placentation, such that loss of Ddx3x leads to embryonic lethality (6). A Ddx3x germline knockout leads to early embryonic lethality, and Ddx3x null male embryos do not survive to later stages of embryonic development (6). Deletion of Ddx3x from neural stem cells mitigates the embryonic lethality issue and the mice exhibit impaired growth of the cerebral cortex and cerebellum, as well as severe ataxia and seizures (7). Despite the prevailing theory that DDX3X syndrome stems from haploinsufficiency, a mouse model of Ddx3x haploinsufficiency in females had yet to be characterized.

In this issue of Biological Psychiatry, Boitnott et al. generate and characterize a novel Ddx3x haploinsufficient mouse with construct validity as a model for DDX3X syndrome (8). The authors performed a series of tests to assess the developmental and behavioral phenotypes of the mouse model. In addition, magnetic resonance imaging and immunostaining were utilized to evaluate brain development and organization of glutamatergic neuronal populations in the developing cortex.

Boitnott et al. utilized a Sox2-Cre driver to delete Ddx3x from the epiblast, thus allowing mice to develop with a functional placenta (6). In line with previous findings (6) and the scarcity of affected males (5), Ddx3x deletion in males led to embryonic lethality (8). The Ddx3x^+/− females, however, survived to adulthood with nearly half the levels of Ddx3x mRNA and protein compared to control females, confirming haploinsufficiency and construct validity for the heterozygous loss-of-function mutations seen in patients.

Ddx3x^+/− pups exhibited lower body weights, as well as aspiration pneumonitis, both of which align with the known patient phenotype of reduced body weight and feeding issues (1). Developmental deficits in Ddx3x^+/− pups included delayed eye opening and pinna detachment and subsequent delays in the processing of sensory cues, evidenced in visual placing, auditory startle, and cliff aversion tests. These delays in sensory processing were accompanied by motor delays, with a longer latency to righting from a supine position, delays in negative geotaxis, and decreased grip strength. The sensory and motor deficits seen in Ddx3x^+/− pups align with phenotypes frequently seen in patients, including sensory hyporeactivity and muscle hypotonia (1). However, while DDX3X syndrome patients tend to have sensory deficits more so in the tactile domain than auditory or visual (1), Ddx3x^+/− pups displayed more consistent delays in responses to auditory and visual stimuli than to tactile cues (8).

Adult Ddx3x^+/− mice showed evidence of hyperactivity and reduced nociception, both of which are commonly seen in patients (1). Boitnott et al. did not see any significant impairments in short-term working memory, but they did find a deficit specific to contextual fear memory. Strikingly, the Ddx3x^+/− mice exhibited normal sociability in the three-chamber social approach test (8), contrary to the deficits in social communication and social motivation seen in patients (1). However, the version of the three-chamber test used by Boitnott et al. primarily assesses social approach, thus it may not necessarily capture deficits in alternative forms of social behaviors. Additional testing would more thoroughly characterize any social behavior abnormalities in this novel mouse model; for example, the nestbuilding task can establish a quantitative measure of social nesting behavior and direct social interaction tests can provide a robust measure of social interaction (9). Further, Ddx3x^+/− mice showed some evidence of anxiety-like behavior in an open field test but not in an elevated plus maze test (8); anxiety is rarely reported in patients (1).

The most common clinical manifestation of DDX3X syndrome in patients is impairment in motor functioning, including reduced gross and fine motor skills, gait abnormalities, and dyskinesia (1,5). These motor deficits initially present as delays in age of first crawling and walking, then persist into adulthood as movement disorders and diminished fine and gross motor skills (1,5). Ddx3x^+/− mice displayed a transient gait abnormality during the second and third postnatal week, and adults demonstrated impaired motor coordination and reduced motor performance on a rotarod test. The same deficit in rotarod performance was seen in older Ddx3x^+/− mice, as well as a reduction in motor learning and function over multiple days of rotarod testing. Balance issues were seen in adult mice and were further exacerbated with age. Interestingly, Ddx3x^+/− mice that were exposed to the balance beam earlier in adulthood showed no balance impairments at an older age.

DDX3X is known to play a role in cortical development through neural progenitor differentiation, and DDX3X syndrome patients present with structural brain abnormalities, most frequently in the cerebral cortex, including microcephaly (1,5,10). While Ddx3x^+/− pups show a reduction in overall brain volume, this phenotype is no longer present by adulthood, pointing to a growth delay, as opposed to a microcephaly phenotype. However, the neocortex and parts of the olfactory and limbic systems all exhibited volumetric reductions.

Along with reduced cortical volume, Ddx3x^+/− mice showed significant alterations in the localization and specification of populations of glutamatergic projection neurons in the cortex. Boitnott et al. saw an increase in the number of cells in the secondary motor cortex, mostly due to an increase in the number of subcerebral projection neurons (ScPNs), marked by CTIP2. Part of this population also localized deeper than its expected position in layer V, and the number of corticothalamic projection neurons (CThPNs), marked by TBR1, increased in deeper layers. The primary motor cortex also had fewer deeper layer CTIP2⁺SATB2⁺ cells, and more CTIP2⁺BRN1⁺ cells. Overall, the motor cortices of Ddx3x^+/− mice had altered localizations of CTIP2⁺ populations, with a larger population of CTIP2⁺BRN1⁺ neurons and a smaller population of CTIP2⁺SATB2⁺ neurons. The primary somatosensory cortex also had fewer CTIP2⁺ neurons and more BRN1⁺ neurons, though subpopulations were unchanged. These changes to CTIP2⁺ populations persisted throughout adulthood. Ddx3x haploinsufficiency thus altered glutamatergic projection neurons in the somatosensory and motor cortices of both developing and adult brains, a finding that aligns with the cortical irregularities and substantial sensory symptoms associated with DDX3X syndrome (1,5).

Boitnott et al. present the first characterization of a Ddx3x haploinsufficient mouse model (Figure 1), a development that is especially important for studying DDX3X syndrome, since haploinsufficiency is considered the most prevalent pathogenic mechanism (2,5). The delays in sensory response and motor development in Ddx3x^+/− mice, along with the abnormal cortical lamination seen in the somatosensory and motor cortices, point to the potential for a better understanding of the mechanisms underlying the neurodevelopmental phenotypes seen in DDX3X syndrome. Most of the brain regions and cortical circuits identified to be perturbed in Ddx3x^+/− mice are also linked to the observed behavioral phenotypes, further highlighting how this mouse model may provide greater insight into disease mechanisms. The construct validity of Boitnott et al.’s model is one of its greatest advantages, offering confidence that data generated from this model will have significant relevance for DDX3X syndrome. The model is also promising with its face validity, as many of the identified mouse phenotypes parallel those observed in patients with DDX3X syndrome. This combination of construct and face validity creates exemplar conditions for future translational studies. Perhaps the most intriguing finding from Boitnott et al. is the improvement of motor function in adult Ddx3x^+/− mice following previous exposure to the task; this suggests that there could be a critical window in the patient population when therapeutic interventions could be used to successfully mitigate pathology and improve prognosis. While this mouse model is limited in its scope as far as exploring other pathogenic mechanisms for DDX3X syndrome, like gain-of-function mutations (10), it is nevertheless valuable in its potential to inform future studies into DDX3X syndrome mechanisms and possible interventions.

Figure 1.

The Ddx3x^+/− mouse model demonstrates construct validity for the haploinsufficient mechanism of DDX3X syndrome. Boitnott et al. generated this model by deleting Ddx3x from the epiblast, thus circumventing the requirement of Ddx3x for placental development. Notably, the mice exhibit developmental delays, hyperactivity, motor deficits, and abnormalities in brain morphology that align with the clinical manifestations of the syndrome, indicating that the model also has face validity for DDX3X syndrome. The model offers insight into neuronal circuits potentially underlying DDX3X syndrome, as well as the possibility of a critical window for successful therapeutic intervention. Findings from the Boitnott et al. study promise to inform future therapeutic strategies.