DDX3X (OMIM: 300160) locates in Xp11.4 and encodes a conserved ATP-independent DEAD-box RNA helicase, which is involved in transcription, splicing, RNA transport, and translation (Abdelhaleem, 2005; Garbelli et al., 2011). The DDX3X is composed of 622 amino acid residues containing two functional domains, namely, a helicase ATP-binding domain and a helicase C-terminal domain (Snijders Blok et al., 2015). De novo DDX3X variants account for 1–3% of unexplained intellectual disability (ID) or developmental delay (DD) (Deciphering Developmental Disorders Study., 2017; Maulik et al., 2011) and also perform as a highly plausible pathogenic gene for childhood apraxia of speech (CAS) (Hildebrand et al., 2020). Most cases of DDX3X variants have been reported in females but very rarely in males, and three previous large cohort studies have described heterogeneous clinical manifestations of DDX3X neurodevelopmental disorder, including ID or DD, dystonia, movement disorders, microcephaly, behavioral issues, feeding difficulties in infancy, and seizure (Snijders Blok et al., 2015; Wang et al., 2018; Johnson-Kerner et al., 2020; Lennox et al., 2020). However, the clinical and genetic features of DDX3X neurodevelopmental disorder in the Chinese cohort have not been described yet.

In this study, we elaborated on clinical manifestations of pathogenic variants of DDX3X in 23 patients (i.e., 22 female and 1 male) in the Chinese cohort and explore the association between genotypes and phenotypes.

In this report, we described a Chinese cohort of patients with DDX3X variants (n = 23, 22 confirmed de novo), and we are the first study to pinpoint clinical and genetic characteristics in the Chinese population. Among the 23 DDX3X variants in our cohort, 17 were novel variants and 14 variants were located in two functional domains of DDX3X (9 were amino acid variants and 5 were truncating variants). Significant differences in sex composition of DDX3X neurodevelopmental disorder have been noted (22 female and 1 male). ID was considered as a universal feature of DDX3X neurodevelopmental disorder in our study, followed by tone abnormalities, microcephaly, feeding difficulties, and seizures. Major aspects of neural development were assessed and quantified using the GDDS-C, and mean scores of five domains were significantly lower than the critical value of 75 (all p-value < 0.05), and language domain was impaired strikingly. Hypothyroidism was reported in 6 patients with DDX3X variants for the first time. Altogether, our study expanded the clinical and genetic spectrum associated with DDX3X neurodevelopmental disorder and evaluated the degree of developmental delay by a standardized scale. It highlighted that WES was necessary for those unexplained ID individuals.

DDX3X, an RNA-binding protein of the DEAD-box family encoded by the DDX3X gene (Abdelhaleem, 2005), acts as a translational regulator (Lai et al., 2008; Lee et al., 2008), particularly for mRNAs with highly structured 5′ untranslated regions (UTRs) (Chen et al., 2018) and for repeat-associated non-AUG translation (Cheng et al., 2019; Linsalata et al., 2019). DDX3X is also the key component of ribonucleoprotein (RNP) granules composed of mRNA and protein (Huang et al., 2019), a pathological hallmark of many neurodegenerative diseases (Ramaswami et al., 2013). Dd3x plays an indispensable role in mouse embryogenesis, synaptogenesis, and brain development (Chen et al., 2016; Lennox et al., 2020). Dd3x neural stem cells knockout at embryonic day (E) 9.5 in mice hampered brain growth, accompanied by seizures and ataxia (Patmore et al., 2020). Boitnott et al. (2021) generated a Ddx3x haploinsufficient mouse (Ddx3x^± female) with construct validity for DDX3X loss-of-function mutations. The Ddx3x^± mice showed global development delay and evolved into behavioral anomalies in adulthood (Boitnott et al., 2021).

Comparing with three previously reported cohort studies of DDX3X neurodevelopmental disorder, we found a certain degree of phenotypic overlap but some special points (Snijders Blok et al., 2015; Wang et al., 2018; Lennox et al., 2020; Table 3). ID was still a general phenotypic feature of patients with DDX3X variants in both our study and previous reports. Similarly, compared with healthy controls, the mean score in GDDS-C of the cohort could reflect global DD. The worst performance in the language subdomain of GDDS-C consolidated language impairments as the most prominent clinical feature of DDX3X neurodevelopmental disorder (Johnson-Kerner et al., 2020). Ophthalmologic problems including refractive errors, nystagmus, strabismus, and amblyopia were presented in 11/23 (47.8%) patients. Previous studies have shown that a variety of eye phenotypes including hypoplasia of the eye or absence of one or both eyes in functional studies of ddx3x knockdown in zebrafish, suggesting deleterious variants in DDX3X, may hamper eye function (Snijders Blok et al., 2015; Kellaris et al., 2018). The SCQ-Lifetime version scores indicated an increased risk for ASD in DDX3X neurodevelopmental disorder, suggesting the necessity of screening for behavior problems by trained behavioral pediatricians.

Furthermore, we noticed hypothyroidism in 6 patients. This novel or rare clinical feature was not previously reported in the original description of DDX3X neurodevelopmental disorder. Poor nutrient absorption and feeding difficulty could make them at increased risk of iodine deficiency, which could be an extrinsic factor causing hypothyroidism in these younger children (Bauer and Wassner, 2019). THs play an essential role in the growth and metabolic homeostasis in humans as well as in animals (Prezioso et al., 2018). Triiodothyronine (T3), the active form of thyroid hormone, acts on its nuclear receptor and modulates target gene transcription (Kumar et al., 2015). Even a 25% reduction in DDX3X levels strongly perturbs neurogenesis, suggesting the high dose-dependency of embryonic cortical development to DDX3X (Lennox et al., 2020). Defective RNA metabolism was considered as the potential mechanism through which DDX3X missense variants hamper fetal brain cortical development (Kumar et al., 2015; Lennox et al., 2020). We speculated that TH deficiency may intensify the adverse effect on RNA metabolism caused by DDX3X missense variants. Appropriate TH supplementation in DDX3X patients with hypothyroidism could be worth trying, but overtreatment or prophylactic hormonal therapy should be avoided because the higher dose of TH supplementation could worsen the outcome (Tuhan et al., 2016). Therefore, it is crucial to have a close follow-up in DDX3X patients with hypothyroidism. Hypothyroidism has not been verified in animal models with in vivo depletion of Ddx3x.

Many study reports have tried to establish the connection between the severity of clinical phenotypes and the location and type of variants and obtained two main findings (Lennox et al., 2020). First, the same recurrent de novo variants were more likely to have similar phenotypes. Recurrent amino acids changes, including R326, I415, and T532, all could cause polymicrogyria (PMG) (Abdelhaleem, 2005; Tuhan et al., 2016; Lennox et al., 2020). Besides the 11 individuals with PMG, 10 were missense variants and one was in-frame deletion, underscoring a striking association between missense variants and severe cerebral anatomical phenotypes, like PMG or dysgyria (Lennox et al., 2020).

Regrettably, no PMG or dysgyria was observed in our cohort, nor the previously reported DDX3X variants relating with PMG. However, we found that females with missense or in-frame deletion DDX3X variants (10/11, 90.9%) were more likely to have abnormal brain structural MRI compared with those with LOF variants (4/11, 36.3%). This could be interpreted by different pathogenic mechanisms. Many aberrant truncating mRNAs (i.e., frameshift or non-sense variants) might undergo non-sense-mediated RNA decay (NMD) and resulted in a haploinsufficiency effect, while a subset of missense variants could function in a dominant-negative manner (Chen et al., 2020; Lennox et al., 2020). Further investigations about gain-of-function mechanism behind certain missense mutations will be carried out by modeling missense mutations in mice. Multiple malignancies have a solid association with somatic DDX3X variants, like malignant melanoma and medulloblastoma (Phung et al., 2019; Patmore et al., 2020). Even though no malignancy was reported in our cohort yet, regular cancer screening is still quite necessary. Finally, a small sample size and a relatively small number of novel phenotypes, such as hypothyroidism, were the main limitations of this study.

In summary, we identified 23 unrelated Chinese patients with causal variants in DDX3X and expanded the knowledge of these increasingly recognized ID disorders. Our study delineated many clinical characteristics of the Chinese cohort with DDX3X variants, largely overlapping with phenotypic spectrum in previously reported studies, but hypothyroidism was first noted as a novel clinical feature. Overall, missense variants were only slightly more common than loss-of-function variants and were mainly located in two functional subdomains. The DDX3X missense variants may have a certain association with abnormal brain anatomical structures. Given the heterogeneous clinical manifestations and involvement of the nervous system and non-nervous systems, unexplained ID in both males and females should take the use of multigene panels that include DDX3X or WES into consideration.

Sequencing data involved in the study are available through the Genbank repository (https://www.ncbi.nlm.nih.gov/bioproject/PRJNA795095). There are restrictions to the full availability of sequencing data of the research participants due to privacy and ethical/legal issues. The data that support the findings of this study are available from the corresponding author, upon reasonable request.

The studies involving human participants were reviewed and approved by the Ethics Committee of XiangYa Hospital, Central South University. Written informed consent to participate in this study was provided by the participants’ legal guardian/next of kin. Written informed consent was obtained from the individual(s), and minor(s)’ legal guardian/next of kin, for the publication of any potentially identifiable images or data included in this article.

YD, ZY, and LL: study design, analysis and revision of the manuscript. YD, ZY, and JGu: follow-up of patient’s information. HL, JGo, and YX: reanalysis of WES data and original draft preparation. YD, BX, HW, and LL: collection of clinical and WES data. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.