The first case of intellectual disability caused by novel compound heterozygosity for NUDT2 variants

The proband was a 5-year-old female to healthy non-consanguineous parents of Chinese ethnicity. She was born at week 40^+ 5 by caesarian section with a birth weight of over 3 kg and a length of 52 cm. No history of neonatal complications or feeding difficulties was observed. Milestones of motor development were delayed as she sat at 9 months, stood at 2 years, and walked unsupported at 2 years 4 months. For her speech, she vocalized “mama, baba” at 10 months and began using single words at 1 year. On current examination at 6 years, her growth parameters weight (30 kg, 97th centile), and height (123 cm, 90th centile) were within normal range. She had subtle facial dysmorphism, such as slightly upslanting palpebral fissures. Her language expression lagged behind that of other children of the same age and slightly poor logic of speech with unclear articulation. Testing with the Wechsler Intelligence Scale revealed 84 for the verbal IQ, 63 for the performance IQ, 71 for the full-scale IQ, showing cognitive skills mildly impaired. She presented with ADHD and deficits in executive function, including slow processing speed, poor learning, and working memory performance. Muscular hypotonia in all extremities and motor impairments with an unsteady gait, climbing stairs difficulties were also noted. Brain magnetic resonance imaging (MRI) revealed thinning but fully formed corpus callosum (Fig. 1A), without other abnormalities. Electroencephalogram (EEG) was normal. Electromyography (EMG) examination demonstrated sensorimotor neurogenic lesions in the distal lower extremities with axonal features in a length-dependent manner, and a normal nerve conduction velocity and SNAP amplitudes.

Trio-WES revealed the proband harboring compound heterozygous variants at NUDT2 NM_001244390: c.34 C > T (p.R12*) and c.194T > G (p.I65R), which were inherited from his mother and father, respectively (Fig. 1B). Sanger sequencing further confirmed both asymptomatic parents to be heterozygous carriers, following the rule of autosomal recessive inheritance (Fig. 1C). Following the ACMG guidelines, the c.34 C > T (NM_001244390) variant in exon 2 is classified as pathogenic, documented in dbSNP with rs number rs148119952. The variant has a frequency of 0.0024 in the 1000 Genomes Project and is absent in gnomAD. It is a nonsense mutation (PVS1), expected to result in premature protein termination. The pathogenic p.R12* has been reported as a homozygous LOF variant in seven GDD/ID patients (PM3) and showed genotype-phenotype segregation in one family (PP1-Moderate). The missense variant of uncertain significance (VUS) p.I65R is absent in the population databases such as 1000 genomes, ExAC, ESP and gnomAD or our internal database of exomes. So it is considered a rare variant although this variant has been recorded in the dbSNP database with the rs number rs765777597 (PM2). Various protein damage prediction software tools, including MutationTaster and SIFT, suggest that this variant is deleterious (PP3). Of the six missense variants in ClinVar, c.174G > T (p.Glu58Asp) was likely pathogenic in intellectual developmental disorder with or without peripheral neuropathy (autosomal recessive inheritance), submitted by the University Hospital Tübingen. Furthermore, residue I65 is positioned near the Nudix box domain of NUDT2 and is highly evolutionary conserved across species (Fig. 2A and B). In molecular modeling, I65 localizes in the loop between the first and second helixes, if it mutates to R65, which replaces hydrophobic isoleucine with a basic polar arginine, will generate unfavorable torsion angles and lead to a change in protein thermodynamic stability (ΔΔG) and thus destabilize the structure of NUDT2, indicating a likely impact on the protein function (Fig. 2C). No other rare variants were identified by Trio-WES in the patient that could explain the observed phenotype. With her clinical profile, especially the presence of ID and delayed motor development resembling the reported cases’ clinical presentations, we considered the identified NUDT2 variants as the likely cause explaining our patients’ disease phenotypes.