Recurrent NUS1 canonical splice donor site mutation in two unrelated individuals with epilepsy, myoclonus, ataxia and scoliosis - a case report

Background We encountered two unrelated individuals suffering from neurological disorders, including epilepsy and scoliosis. Case presentation Whole-exome sequencing identified the same recurrent, de novo, pathogenic variant in NUS1 [NM_138459.4:c.691 + 1C > A] in both individuals. This variant is located in the conserved cis-prenyltransferase domain of the nuclear undecaprenyl pyrophosphate synthase 1 gene (NUS1), which encodes the Nogo-B receptor, an essential catalyst for protein glycosylation. This variant was confirmed to create a new splice donor site, resulting in aberrant RNA splicing resulting in a 91-bp deletion in exon 3 in both individuals. The mutant mRNA was partially degraded by nonsense mediated mRNA decay. To date, only four de novo variants and one homozygous variant have been reported in NUS1, which cause developmental and epileptic encephalopathy, early onset Parkinson’s disease, and a congenital disorder of glycosylation. Seven patients, including our two patients, have presented with epileptic seizures and intellectual disabilities. Conclusions Our study strongly supports the finding that this recurrent, de novo, variant in NUS1 causes developmental and epileptic encephalopathy with involuntary movement, ataxia and scoliosis.


Background
The NUS1 (nuclear undecaprenyl pyrophosphate synthase 1) gene encodes the Nogo-B receptor (NgBR) [1,2], which interacts with dehydrodolichyl diphosphate synthase complex subunit (DHDDS) and promotes cisprenyltransferase (cis-PTase) activity. NgBR is an essential catalyst of the dolichol monophosphate (Dol-P) biosynthetic machinery in eukaryotic cells [3,4]. The wellconserved C-terminus domain of cis-PTase in NgBR has intrinsic effects for protein structure stabilization, in association with N-glycans. To date, four de novo NUS1 variants have been reported in association with developmental and epileptic encephalopathy (DEE) and earlyonset Parkinson's disease, and one homozygous NUS1 variant has been associated with a congenital disorder of glycosylation. Interestingly, two de novo variants [5] and a pair of compound heterozygous variants [6] in DHDDS have been reported in five patients with DEE seizures or congenital glycosylation defects, suggesting that pathogenic variants in the NgBR-DHDDS pathway may cause neurological disorders. Here, we report two unrelated Japanese patients with a novel, recurrent, de novo NUS1 variant, who presented with epileptic seizures with involuntary movement, ataxia, intellectual disability and scoliosis.

Case presentation
The patient 1 was the second child born to nonconsanguineous, healthy parents. Her elder brother had febrile seizures during childhood. She was born spontaneously, at full term, with no asphyxia. Her birth weight was 2826 g (− 0.44 SD). She gained head control at 4 months of age and sat without support at 7 months of age. She experienced febrile seizures at 9 months of age and generalized tonic-clonic convulsions without fever at 14 months of age, at which time valproic acid (VPA) was administered. Tremulous myoclonus of the extremities was also observed. She walked without support at 20 months of age, spoke a meaningful word at 10 months of age and two-word phrases at 24 months of age. Her developmental quotient was 77 at 2 years of age. Her seizures occurred once per year until the age of 6 years and 5 months; however, an increase in the VPA dosage lessened an episode of convulsive attack, after which her seizures disappeared. Her electroencephalograms (EEG) showed 3-Hz, diffuse, spike-and-slow-wave complexes with a 7-Hz slow wave background at 8 years of age (Additional file 1: Figure  S1A), which became worsened at 15 years of age. However, the treatment with VPA and LEV significantly lessened 3-Hz diffuse spike-and-wave complexes and only 3-Hz high-amplitude slow wave bursts were infrequently recorded during sleep at 17 years of age (Additional file 1: Figure S1B). Her brain magnetic resonance imaging (MRI) results were normal at 6 years of age (Additional file 1: Figure S1C, S1D) and at 15 years of age (Additional file 1: Figure S1E, S1F). At the age of 17 years, her height was 157.4 cm (− 0.12 SD) and her weight was 41.9 kg (− 1.42 SD). She had no dysmorphic features except for scoliosis which needed a surgical correction at 15 years of age. She showed dysgraphia, due to tremulous myoclonus of the bilateral extremities (Additional file 4: Movie S1). She showed no behavioral disorders, such as autistic spectrum disorders or attention deficit/hyperkinetic disorder.
In patient 1, trio (sequencing with parents) WES was performed, and 5 de novo variants were detected (Additional file 5: Table S1 and Additional file 6: Table S2 and Additional file 7: Supplemental method This variant was absent from public databases (allele frequency was 0 in ExAC, gnomAD, ESP6500, HGVD, ToMMo and in-house 575 Japanese exome controls). Multiple in-silico evaluation scores for predicting the pathogenicity of DNA sequence alternations suggest that this variant is deleterious: Mutation Taster (http://www. mutationtaster.org/) returned a value of disease causing; CADD (https://cadd.gs.washington.edu/) returned a value of 25.9; and Fathmm (http://fathmm.biocompute. org.uk/) returned a value of deleterious. This splice variant is predicted [7] to create a new splice donor site, which could change the reading frame and introduce a premature termination codon (PTC).
The patient 2 was born normally to nonconsanguineous, healthy parents. His birth weight was 3550 g. At the age of 6, it was noticed he had speech delay, clumsiness of the hands, and involuntary movements of the hands when he used chopsticks. At the age of 8 years, he was suspected to have a cerebellar atrophy, with seizures. Occasionally, jerky movements of the limbs also appeared. From 14 years of age, he gradually developed a gait abnormality. At the age of 37 years, he was admitted to hospital. On physical examination, he showed flat foot and limb ataxia. Clonazepam was remarkably effective for treating his gait disturbance. Laboratory examinations were all normal, including lactic acid, pyruvic acid, vitamins, the thyroid gland, ceruloplasmin, copper, lipoproteins, amino acid analysis and leukocyte lysosome enzyme activities (α-galactosidase, β-galactosidase, βhexosaminidase and arylsulfatase). Genetic testing for dentatorubral-pallidoluysian atrophy was negative. At the age of 40 years, myoclonic jerks of the limbs developed, in addition to ataxia. He was diagnosed as progressive myoclonus epilepsy with an unknown cause. At the age of 42 years, scoliosis became apparent. Starting at the age of 48 years, he began to require assistance with walking. At the latest examination (59 years), he showed intellectual disability (equivalent to that of a 6-year-old), excessive blinking due to tenseness, and profound action myoclonus of the limbs, which could be referred to "hyperkinésie volitionnelle". His eye pursuit was saccadic, and his speech was explosive. Tendon reflexes were slightly increased, and no sensory disturbances were observed. An EEG analyzing jerk-locked back averaging potentials suggested that the myoclonus emerged from the cortex. Examinations of MRI, nerve conduction studies, conventional EEGs, and laboratory examinations of the cerebrospinal fluid and blood were almost within normal ranges (Additional file 2: Figure S2A, S2B). Increasing the dosage of clonazepam up to 12 mg (0.5 mg × 24 tablets/day) did not alleviate neurological symptoms; however, the oral administration of 50 mg baclofen remarkably lessened myoclonus and slightly improved gait disturbance.
We performed proband-only WES in patient 2 and de-  (Fig. 1a and Additional file 7: Supplemental method). SPTAN1 variant was inherited from healthy mother. Therefore, its pathogenicity should be minimal in the patient.
To test whether this variant causes aberrant splicing, we examined the cDNA from both individuals' lymphoblastoid cell lines, which revealed that the mutant allele has a 1-base alternation in the splice donor site (chr6: 118,015,344) creating a new splice donor site of GT in exon 3 (chr6:118,015,253), resulting in a 91-bp deletion in the NUS1 exon 3 (Fig. 1b). Electropherograms of both individuals' cDNA showed that a 91-bp region of exon 3 is missing. TA-cloning of the short RT-PCR product confirmed the same event in the mutant allele (Fig. 1b, Additional file 3: Figure S3A and Additional file 7: Supplemental method). This variant creates a new reading frame [c.601_691del:p.(Arg202Glnfs*9)] and produces a PTC at chr6:118,024,795 (73-bp upstream of the 3′ exon-exon junction) (Fig. 1b). According to the major rule of nonsense-mediated mRNA decay (NMD) [8], the mRNA of the mutant allele should be subjected to NMD. However, the results of both individuals' cDNA sequencing chromatograms showed that cycloheximide (an NMD inhibitor) treatment did not recover the peak height of the electropherogram, suggesting that NMD is not involved. Thus, we performed a quantitative analysis using RT-PCR (see Additional file 7: Supplemental method). RT-PCR showed that the relative gene expression levels of NUS1 was slightly reduced in the patients' LCLs, with possible minor recovery following cycloheximide treatment (Fig. 1c). These results support the prediction that the transcribed mRNA of the mutant allele is only partially subjected to NMD.