Epilepsy Phenotypic Spectrum of NUS1‐Related Disorder: A Case Series

ABSTRACT

Background

Epilepsy with myoclonic and atonic seizures (EMAtS), also known as Doose syndrome, accounts for 1%–2% of childhood epilepsies, and various genes have been implicated in causing this epilepsy syndrome. NUS1 encodes for Nogo‐B receptor (NgBR), which stabilizes the dehydrodolichyl‐diphosphate synthase complex in the endoplasmic reticulum, promoting its enzymatic activity (cis‐IPTase) and thereby regulating cholesterol biosynthesis. Pathogenic variants in NUS1 have been associated with movement disorder and epilepsy; however, the spectrum of epilepsy and electroencephalogram (EEG) phenotype has not been well characterized.

Methods

We describe a single‐center case series of five patients with NUS1‐related disorder.

Results

In our cohort, three patients met the diagnostic criteria of EMAtS, and the remainder had a milder form of generalized epilepsy. Four patients had a pathogenic variant in NUS1 on one allele, and one patient had a missense change of unclear significance but fit the phenotype of NUS1‐related disorder. All patients bearing the pathogenic variants in NUS1 had normal to mild developmental delay at the onset of epilepsy, with normal brain magnetic resonance imaging. Age of seizure onset in these patients was 1–7 years, and patients responded to levetiracetam and/or valproic acid. The EEG findings for these patients included the presence of spike and slow wave discharges, as well as the presence of generalized, invariant monomorphic theta range activity in the awake state, which was seen in four out of the five patients.

Conclusion

Taken together, NUS1 variants are associated with generalized epilepsy phenotype and an invariant EEG pattern of monomorphic theta activity.

1. Introduction

Genetic generalized epilepsies of childhood have a highly variable phenotype and clinical course, in part due to the effect of multiple gene variants associated with epilepsy in isolation or together [1]. Among these generalized epilepsies, there exists a spectrum first described by Doose and colleagues in 1988, wherein patients develop multiple seizure types, and at the onset of epilepsy, patients have normal development; however, later there may be developmental stagnation or regression, which is proportional to seizure burden [1, 2]. The definition of Doose syndrome or epilepsy with myoclonic‐atonic seizures (EMAtS) has evolved over the past four decades, and this epilepsy accounts for up to 2% of all childhood‐onset epilepsies [3]. EMAtS is defined by the International League Against Epilepsy (ILAE) as having multiple generalized onset seizures, including myoclonic‐atonic, myoclonic, absence, atonic, and generalized tonic‐clonic seizures [1]. Several candidate genes have been associated with EMAtS or Doose syndrome, including SLC6A1 [4], SYNGAP1, SCN1A, SCN1B [4], GABRG2 [5], GABRB3, SLC2A1 [6], CHD2, STX1B [7], WDR45 [8], NEXMIF (formerly KIAA2022) [9, 10], HNRNPU, and STS [11]. Many of these genes are also implicated in several different generalized epilepsy syndromes, including Lennox–Gastaut syndrome, developmental and epileptic encephalopathies (DEEs), and childhood absence epilepsy, highlighting the highly heterogeneous phenotypes associated with variants in these genes, as well as the heterogeneous natural history of patients with EMAtS [2]. The variation in the epilepsy phenotype is also accompanied by a spectrum of neurological comorbidities, including neurodevelopmental disorders, autism spectrum disorder, and developmental delay/intellectual disability [2].

From an electroencephalography (EEG) standpoint, interictal epileptiform abnormalities of 2–6 Hz generalized spike and wave discharges are considered characteristic of EMAtS [1]. Most patients with EMAtS have normal background EEG features at onset, including the presence of a well‐formed posterior dominant rhythm (PDR) [1, 2]. Some patients with EMAtS can also have theta rhythms on their EEG background. The correlation between biparietal theta rhythms on EEG and prognosis in generalized epilepsies was first described by Doose and Baier [12] as “theta trait,” which was thought to be characteristic of the cortico‐reticular seizure network [12].

Variants in NUS1 have only recently been associated with EMAtS or Doose syndrome [13] but are known to result in epilepsy. NUS1 encodes for Nogo‐B receptor (NgBR), an endoplasmic reticulum (ER) membrane resident protein important for stabilization of the dehydrodolichyl‐diphosphate synthase complex, increasing its enzymatic activity (cis‐IPTase) and thereby regulating cholesterol biosynthesis [14]. Loss‐of‐function variants in NUS1 are associated with lysosomal cholesterol accumulation [14]. Monoallelic variants in NUS1 are associated with myoclonus, epilepsy, ataxia, and intellectual disability (MEAID) syndrome; however, biallelic variants in NUS1 are associated with congenital disorders of glycosylation, type IAA [15].

In this study, we describe five patients with novel heterozygous NUS1 variants, all presenting with a relatively homogeneous generalized epilepsy phenotype and EEG characteristics.

2. Materials and Methods

2.1. Patient Cohort and Ethics

This is a single‐center retrospective case series of five patients followed at Washington University in St. Louis, Department of Neurology. The study was approved by the local institutional review board (IRB).

2.2. Molecular Genotyping

Genetic testing for all patients was performed as a clinical standard of care, and data were retrospectively collected using an electronic health record database filtered for patients with de novo variants in NUS1.

2.3. Phenotypic Data

Medical records were retrospectively obtained for the five individuals with NUS1 variants. Pertinent clinical, EEG, and imaging data were assessed by two board‐certified clinical epileptologists. EEG was performed as clinically indicated using the Nihon Kohden digital EEG system (Tokyo, Japan).

3. Results

3.1. Phenotype

This is a single‐center cohort of five patients bearing heterozygous variants in the NUS1 gene. All patients developed epilepsy with an age of onset for seizures between 15 months and 7 years of age. Before developing unprovoked seizures, two of five patients had febrile seizures. At the onset of epilepsy, developmental delay was present in four of five patients (details in Table 1). Of the four patients with developmental delay, one patient had autism, two had mild global developmental delay, and one had moderate intellectual disability with a composite intelligence quotient (IQ) of 35 on the Wechsler Adult Intelligence Scale. Congruent with recent findings [13], two patients had dysarthria and one patient had speech apraxia (Table 1).

Table 1.

Clinical and EEG characteristics of patients with NUS1‐related epilepsy.

Pt	Variant	Seizure types	Seizure onset age	Tremor/ataxia	Dev delay	ASMs	Febrile seizures	EEG characteristics				Age at last seizure	Follow‐up	Family history
								PDR	Excessive theta	IEDs	Photic activation
#1	c.335C>G (p.Pro112Arg)	Myoclonic, atonic, absence	15 months	None	Mild language/speech delay after the onset of seizures. Dysarthria present.	VPA, LCM	Yes	Yes	Present	Generalized spike and polyspike (3 yo)	No	7 (1)	6 (8)	Dad: febrile seizures until 6 yo Brother: night terrors Maternal cousin: trisomy 21 Maternal cousin: ASD and OCD Paternal cousin: neurofibromatosis
#2	c.695C>G (p.S232*)	Absence, GTC	7 years	Tremor	Mild autistic features, mild language/speech delay, no gross or fine motor delay	LEV	No	Yes	Present	Generalized spike wave (9 yo)	Yes	11 (1)	5 (12)	None
#3	c.792‐2A>G (splice)	Absence, myoclonic	2.5 years	Tremor, Ataxia	Moderate intellectual disability with an IQ of 35, no gross or fine motor delay	LEV	No	Yes	Present	Generalized spike and slow wave (2.5 yo) None (22 yo). Generalized slowing.	No	5 (22)	25 (27)	None
#4	c.378del (p.Val127Trfs*22)	Myoclonic, atonic	14 months	Tremor	Speech apraxia, mild fine motor, gross motor, speech delays	LEV, CLB, VPA	Yes	Yes	Present	Generalized spike and slow wave (15 mo)	Yes	3 (1)	3 (4)	Mother: papillary thyroid cancer Maternal grandmother: multiple sclerosis Mothers's paternal aunt's daughter with adult‐onset seizures Mother's maternal aunt: limb‐girdle muscular dystrophy (deceased) Paternal grandmother: breast cancer, SLE, and scleroderma Paternal aunt's son with autism
#5	c.719T>G (p.Leu240*)	Myoclonic, atonic	2.5 years	None	Speech delay, mild fine motor delay, no gross motor delay	LEV, OXC	No	Yes	Absent	Generalized spike and slow wave (2.5 yo)	Yes	3 (1.5)	2 (5)	Mother, father, and sister with a learning disability Brother with epilepsy, Fabry disease, and a learning disability

Note: Age at last seizure is the age of each patient when they had their last seizure while on anti‐seizure medication(s), and the number in parentheses indicates the number of years the patient has been seizure‐free. Follow‐up refers to the number of years of follow‐up from seizure onset, with the number in parentheses representing the age at the time of the study. Abbreviations: ASM = antiseizure medications, VPA = valproic acid, CLB = clobazam, dev delay = developmental delay, LCM = lacosamide, LEV = levetiracetam, mo = months old, OXC, oxcarbazepine, yo = years old, IED = interictal epileptiform discharges, PDR = age‐appropriate posterior dominant EEG discharges ("Yes" means present).

Each patient had more than one seizure type, with the presence of myoclonic‐atonic seizures (4/5), absence seizures (3/5), and generalized tonic‐clonic seizures (1/5). All patients were diagnosed with generalized epilepsy. Most patients were followed for 2–6 years from the first clinical seizure, with one patient being followed for 25 years. On retrospective chart review of seizure semiology, electrographic pattern of seizures, and grapho‐elements on EEG background (like age‐appropriate PDR) at onset of epilepsy, it was determined that all patients were on a spectrum of a milder form of generalized epilepsy [2], with three patients meeting the recent ILAE criteria of EMAtS [1], and the remaining patients had a generalized epilepsy phenotype (Table 1).

Movement disorders in NUS1‐related disorder, including ataxia and tremors, have been previously reported [15, 16]. In our cohort of five patients, ataxia was present in one patient, and tremor with arm extension was present in three patients.

3.2. EEG Characteristics

The EEG background for all patients at the onset of epilepsy consisted of an age‐appropriate PDR. Generalized excessive monomorphic invariant theta rhythms were present in four of five patients. Generalized interictal epileptiform discharges (IEDs) were present in all patients, with one patient (patient #3) having resolution of IED at 22 years of age, with the presence of generalized slowing (Table 1). Photoparoxysmal response was present in three of five patients. A typical EEG in bipolar longitudinal montage is shown in Figure 1a, and quantitative EEG visualization of excessive theta rhythm is shown in Figure 1b.

Figure 1.

Electroencephalography (EEG) characteristics of NUS1‐related epilepsy. (a) Typical awake EEG of one patient with NUS1‐related epilepsy (patient #2, 3 years of age at the time of this EEG). Tracing depicts a bipolar longitudinal montage. An invariant monomorphic theta rhythm is seen on the EEG background. A typical myoclonic‐atonic seizure is also seen, as shown with the red arrow (sensitivity of 10 microvolts, low‐frequency filter set at 1 Hz, high‐frequency filter set at 70 Hz). (b) Quantitative EEG over a 12‐h epoch using Persyst 14, of one patient with NUS1‐related epilepsy. The top panel represents left hemisphere electrodes, and the bottom panel represents right hemisphere electrodes. Prominent theta range activity is shown with red arrows.

3.3. Response to Treatment

All patients responded to anti‐seizure medications, with levetiracetam being effective in four of five patients, with one patient discontinuing levetiracetam because of side effects. Levetiracetam monotherapy resulted in full seizure control in two patients, and three patients required valproate, clobazam, lacosamide, and/or oxcarbazepine for resolution of seizures for at least 1 year. None of the patients has experienced resolution of epilepsy at the time of this study (Table 1).

3.4. Genetic Variants

As detailed in Table 1, variants in NUS1 included frameshift (1/5), premature stop codon (2/5), splice variant (1/5), and missense variant (1/5). Four patients (of five) had heterozygous de novo variants in NUS1. One patient (patient #5) has a half‐brother with a NUS1‐related disorder (not included in our cohort), but both parents do not have the NUS1 variant, indicating possible germline mosaicism. All variants were classified as pathogenic except for one patient with the missense variant, which was classified as having a variant of unknown significance (VUS). However, this missense variant (c.335C>G) has previously been reported on ClinVar by multiple groups as associated with autosomal recessive congenital disorder of glycosylation, type IAA (rs540846579). We included this patient in the cohort given the phenotypic similarities to other patients. Pathogenic variants c.695C>G, c.792‐2A>G, and c.378del have not been previously reported. Pathogenic variant c.719T>G has been reported on ClinVar (rs2114693896) for its association with autosomal recessive congenital disorder of glycosylation, type IAA.

4. Discussion

NUS1‐related disorder represents a wide phenotypic spectrum associated with movement disorders, including myoclonus, cerebellar ataxia, and tremors; generalized epilepsy with a variety of different seizure types, including myoclonic, myoclonic‐atonic, generalized tonic‐clonic, and absence seizures; and developmental delay/intellectual disability [15, 16]. The epilepsy phenotype for NUS1‐related disorder has not been well characterized, and a recent case series reported that 68% of patients have a diagnosis of epilepsy [13]. Here, we describe the clinical and EEG characteristics of patients with NUS1‐related disorder. The overall epilepsy phenotype in three of the five patients is most consistent with EMAtS or Doose syndrome, with worsening of the developmental trajectory after onset of epilepsy and partial improvement with anti‐seizure medications. The overall phenotype fits a mild DEE, which includes EMatS [1]. The remaining two patients had generalized epilepsy with mild developmental delay and did not fulfill the ILAE criteria for EMAtS [1].

This is a single‐center case series with three patients meeting the ILAE definition of EMAtS and two patients having generalized epilepsy with normal development to mild developmental delay at the onset of epilepsy, multiple seizure types, and characteristic EEG findings. Generally, patients were responsive to levetiracetam and valproate (Table 1), and there was a variable movement disorder phenotype. In terms of development, all patients had speech delay, which worsened or appeared after the onset of seizures, and one patient (patient #4) had mild gross and fine motor delay after the onset of seizures. Taken together, developmental delay is variable in patients with NUS1‐related disorder and mainly affects language and speech.

The EEG background in our NUS1‐related epilepsy patients demonstrated a characteristic generalized invariant theta range activity (four of five patients), which is in contrast to the more restricted theta rhythms in the biparietal regions for patients with EMAtS or Doose syndrome [12]. Additionally, all patients had age‐appropriate PDR and a normal thalamo‐cortical rhythm seen during relaxed wakefulness and associated with acute cognitive changes [17].

Currently, there are no approved therapies for patients with NUS1‐related disorder; however, zebrafish studies have demonstrated that loss‐of‐function variants in NUS1 result in lysosomal cholesterol accumulation, and this buildup of cholesterol could be ameliorated by 2‐hydroxypropyl‐ß‐cyclodextrin [14]. To our knowledge, no gene therapy or antisense oligonucleotide (ASO) trials are currently ongoing.

Given the unique neurological phenotype of this rare condition, it is vital to foster societal support among families globally with children who have a NUS1‐related disorder. Toward this end, two mothers of children with NUS1‐related disorder founded the NUS1 Foundation with the goal of increasing collaborative research, maintaining a patient registry, and providing useful resources for families of patients with NUS1‐related disorder (www.nus1foundation.org).

Author Contributions

Saumel Ahmadi: conceptualization, methodology, supervision, data curation, writing – original draft, writing – review and editing. Natalie Fulton: data curation, methodology. Michael Morrissey: methodology, data curation, writing – review and editing. Rebekah Landre: methodology, data curation, writing – review and editing. Stuart Tomko: methodology, data curation, writing – review and editing. Fumihiko Urano: writing – review and editing. Réjean M. Guerriero: methodology, project administration, resources, writing – review and editing.

Conflicts of Interest

S.A. has clinical trial support and is a consultant for Amylyx Pharmaceuticals Inc., which is unrelated to this work. F.U. has a sponsored research agreement and received material support from Prilenia Therapeutics and National Institutes of Health (NIH) grants; royalties from Novus Biologicals and Sana Biotechnology; licensing/consulting fees from Opris Biotechnologies; consulting fees from Emerald Biotherapeutics; and travel fees from Wolfram France, Wolfram UK, and the Snow Foundation; has had unpaid advisory roles for the Snow Foundation and Be A Tiger Foundation; holds patents (9 891 231, 10 441 574, and 10 695 324); and was president and shareholder of now‐dissolved CURE4WOLFRAM. None of the other authors has any conflicts of interest related to this work.