SLC7A3: In Silico Prediction of a Potential New Cause of Childhood Epilepsy

Keywords

epilepsy - case reports - gene panel - whole exome sequencing - SLC7A3

Introduction

Recent data underlines that genetic testing in childhood epilepsy can allow an accurate prognosis, impact management, likely limit the diagnostic odyssey, and possibly avoid further unnecessary testing. This is especially true in children with early-onset, drug-resistant epilepsy with neurodevelopmental disorders.[1] [2]

Several genetic tests are currently available and widely used in clinic. The advantages and limitations of cytogenetics (karyotyping, microarray), gene panels, whole exome sequencing (WES), and whole genome sequencing (WGS) have recently been reviewed.[3] Gene panel sequencing (most commonly by filtering WES data and including the analysis of both parents [“Trio”]) is currently the most applied, cost-effective, and rapid test in clinic for diagnosing genetic epilepsies.[4] This approach has a high “read depth” or “coverage,” meaning that this technique achieves a relatively high number of reads at a given base position. Due to this feature, nonsynonymous single nucleotide variants (SNVs), small intragenic deletions, duplications and flanking intronic sequences, as well as somatic variants have a higher probability to be detected. Conversely, WES would miss a variant that is poorly covered but has a higher “resolution” since it targets all the genes in our genome with an acceptable coverage in over 90% of the coding sequences. As a result, WES has largely contributed to the process of disease-gene identification.[1] [3] Interestingly, it has been shown that cytogenetic and sequencing analyses are complementary, and if one analysis does not yield positive results, the other one is often indicated. For example, up until now 27 genes, among those expressed in brain and belonging to the transmembrane solute carrier transporter (SLC) large family, have been related to patients with epilepsy (PWE).[5] In most epilepsy gene panels, only few of these SLC genes known to be involved in epilepsy are included. So, the majority of SLC variants would not be detected, making WES an appropriate, additional approach.

In either case, WES has entered the clinical domain,[1] and increasing research has been performed to investigate novel genes detected by WES. Even though functional studies using preclinical models are ultimately necessary to evaluate the pathogenicity of novel variants in vivo,[6] [7] [8] in silico computational methods and segregation analyses can contribute to elucidate the potential role of the variants of unknown clinical significance (VUS). To point out, the InterVar tool generates an automated interpretation of any VUS by using 18 criteria, validated by the American College of Medical Genetics and Genomics (ACMG),[9] and MutationTaster is able to further analyze SNVs, InDels and non-coding variants.[10] More details about InterVar and MutationTaster can be found in the [Supplementary Material S1] (available online only). Using the segregation analysis in three different generations and the aforementioned in silico prediction models, we propose that the X-linked, WES-identified SLC7A3 variant is the possible genetic cause contributing to the etiology of epilepsy and developmental issues (motor and language) in our patient.

To reinforce the role of SLC7A3 in neurological disease, we have used GeneMatcher[11] to unveil other patients with SLC7A3 variants. This had led to two additional patients with SLC7A3 variants who both show neurological problems including epilepsy and developmental problems.

Original Case Report (1)

An 11-year-old boy with a history of speech delay onset was referred to the University Hospitals Leuven at the age of 3 years due to difficult to treat epilepsy.

He was born by cesarean section at 37 weeks of uncomplicated gestation to nonconsanguineous parents. The birth weight was 3200 g. During the first years of life, there were no medical issues despite delay in language development and psychomotor skills. At the age of 3 years, he experienced his first symmetric tonic seizure. When he turned 6 years, he also experienced clusters of tonic seizures and atypical absences. Physical examination, metabolic screening ([Supplementary Material S1], [Table S1] [available online only), and brain imaging studies (magnetic resonance imaging [MRI]) were unremarkable. EEG analyses showed a background with high-voltage generalized spikes and spike waves bilaterally with maximum amplitude in the frontal regions. During sleep, a marked increase in epileptiform activity was shown. Although multiple distinct antiseizure medications (ASM) were used, in mono- and/or polytherapy, our patient never experienced any seizure freedom. Currently, he is on ethosuximide and levetiracetam (LEV), due to which he did not experience any tonic seizures or atypical absences since the age of 8 years. However, short series of myoclonic seizures are still occurring in the limbs when he falls to sleep (duration ± 1 second) on a daily/weekly basis. Regarding the family history, his father experienced a few seizures when he was a child. Nevertheless, he did not need ASM treatment and had a normal development and intelligence quotient (IQ).

Karyotype analysis, chromosomal microarray, investigations for GLUT1 deficiency (SLC2A1), myoclonic epilepsy with ragged red fibers (MERRF), and an epilepsy gene panel (IPG institute, Charleroi) were performed. The gene panel unraveled two VUS in the ZNF335 gene: c.3023C > G (Pro1008Arg) and c.3445A > G (Ile1149Val). The first VUS was inherited from the father and is present in the European control population (ExAC, 0.0015%). The second VUS was inherited from the mother and is also present in the general control population (ExAC, 0.08%). Even as four out of five prediction tools indicate that this latter VUS could be pathogenic ([Table 1]), compound heterozygous (or homozygous) mutations in this gene have been linked to basal ganglia abnormalities,[12] agyria with prominent extra-axial spaces, and cerebral/cerebellar hypoplasia or hypomyelination,[13] which do not overlap with our patient's phenotype.

Table 1 Variants of unknown significance reported in our patient analyzed by five computational analyses.[9] [10] Deleterious thresholds are shown between brackets in the first row[14] [15]

Gene	MutationTaster (> 0.5)	CADD (> 15)	SIFT (< 0.05)	GERP + + (> 4.4)	MetaSVM (> 0.09)	Tools indicating pathogenicity
Case report 1
ZNF335 c.3023C > G	Benign, p = 0.999	17.58	0.026	4.66	–1.037	3/5
ZNF335 c.3445A > G	Pathogenic, p = 0.999	25.1	0.001	4.86	–1.108	4/5
CACNA1H c.2264G > A	Benign, p = 0.999	ND	ND	ND	ND	0/1
IGLON5 c.8C > G	Benign, p = 0.996	14.7	0	1.24	–1.022	1/5
SLC7A3 (1) c.1229T > G	Pathogenic, p = 0.999	19.75	0.042	5.14	0.272	5/5
Case report 2
SLC7A3 c.107A > G	Pathogenic, p = 0.999	27.6	3	4.84	0.906	4/5
ST8SIA4 c.143T > C	Pathogenic, p = 0.999	22.5	0.044	5.96	–1.069	4/5
Case report 3
SLC7A3 c.244C > A	Pathogenic, p = 0.999	26.4	0.002	4.7	0.919	5/5

MutationTaster = conservation DNA sequence, splice site and mRNA stability prediction, protein features (p = probability); CADD = 63 distinct variant annotation retrieved from Ensemble Variant Effect Predictor, data from the ENCODE project and information from UCSC genome browser tracks; SIFT = Protein sequence conservation among homologs; GERP + + = conservation DNA sequence by multispecies sequence alignment compared with the neutral expectation; MetaSVM = Ensemble scores based on Support Vector Machine.

In the context of the ERN EpiCARE network, WES revealed three VUS: that is, CACNA1H (c.2264G > A), IGLON5 (c.8C > G), and SLC7A3 (c.1229T > G). Pathogenicity prediction tools are summarized in [Table 1]. The CACNA1H variant (inherited from the father) has been reported in the control population (66/265740 alleles, gnomAD). This is a synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence or the creation of a new splice site, and the nucleotide is not highly conserved. The IGLON5 mutation (de novo) has not been related to human disease and is reported in the control population (4/69228 alleles, gnomAD). The SLC7A3 variant has never been reported in the control population (gnomAD). Segregation of this SLC7A3 variant revealed that it was inherited from a healthy mother and was also present in the healthy maternal grandmother. The inheritance model of SLC7A3 is X-linked recessive (XLR), thus females can be healthy carriers and males can be affected.[6]

Additional Case Report via GeneMatcher (2)

A 13-year-old girl was referred to the Hillerød hospital (Denmark) at the age of 14 months due to febrile seizures (FS).

One month later, she had two afebrile seizures and treatment with valproic acid (VPA) was initiated. At the age of 4 years, she was seizure-free and tapered off VPA. At the age of 8 years, seizures reoccurred (clonic seizure) during a fever episode and again at age 11 years (two afebrile generalized tonic clonic seizures), for which treatment with LEV was started. Treatment was changed to eslicarbazepine (ESL) due to LEV-induced side effects. The EEG was normal during the first years of life. In contrast, at 11 years of age, her EEG showed epileptiform activity (spike-wave complexes [3 Hz]). Improvement on the EEG was seen after ASM treatment was started, and she has currently been seizure free for 2.3 years on ESL. Physical examination was normal. Metabolic screening and brain imaging studies (MRI at age 4 and 12 years) were normal. Further clinical details can be found in the [Supplementary Material S1] (available online only).

Regarding the family history, the father had one FS but was never diagnosed with epilepsy or neuropsychiatric disease.

Chromosomal microarray, fragile-X analysis, and WES were performed. WES revealed one VUS in the SLC7A3 gene (c.107A > G) and one VUS in the ST8SIA4 gene (c.143T > C) ([Table 1]). Both variants (de novo) have never been reported in the control population (gnomAD). We cannot assure that neither SLC7A3 nor ST8IA4 are causative.

Additional Case Report via GeneMatcher (3)

A 2-year-old boy with a history of laryngeal cleft, tracheomalacia, feeding difficulties, poor weight gain, and speech and fine motor delay was referred to Massachusetts General Hospital (MGH, Boston, United States) at the age of 12 months due to feeding difficulties. Further clinical details can be found in the [Supplementary Material S1] (available online only).

Physical examination was unremarkable. Metabolic screening showed an increase of several amino acids (including arginine), although the significance of this finding remains uncertain. MRI showed asymmetry in size of the lateral ventricles at 11 months of age (left lateral ventricle larger than right lateral ventricle, potentially related to anatomic variation). EEG analyses showed normal sleep and awake EEG and no epileptiform abnormalities.

Regarding the family history, no epilepsy or developmental delay have been reported.

Chromosomal microarray was normal. WES revealed one VUS in the SLC7A3 gene (c.244 C > A), which was inherited from the patient's mother who never experienced any health issues ([Table 1]).

Discussion

We report on the genetic analyses that we performed to find a possible new cause of childhood epilepsy and/or developmental issues.

In our original case report (1), gene panels and WES detected five variants possibly related to the epilepsy in this patient. Epilepsy in ZNF335-related disease involves biallelic recessive mutations affecting the zinc finger protein that is crucial for methylation and expression of several brain specific genes. Initial genetic studies in our patient revealed a compound heterozygous state of the ZNF335 gene. However, computational analyses revealed that one of the two VUS are likely benign and the patient's phenotype did not overlap with those of reported PWE and ZNF335 mutations, showing clear brain anomalies such as microcephaly and/or invisible basal ganglia.[12] [13]

Subsequently, WES revealed three VUS in three genes (IGLON5, SLC7A3, and CACNA1H). Our in silico predictions showed that the IGLON5 and CACNA1H variants are likely benign and also present in a healthy control population, in clear contrast to the SLC7A3 variant ([Table 1]). The CACNA1H variant was inherited from the father who experienced self-limiting seizures during childhood, thus we cannot exclude a modifier role of CACNA1H,[14] possibly another—yet unknown—gene with variable expression in father and son or a polygenetic cause. Whereas the SLC7A3 variant, follows a XLR inheritance pattern,[6] which explains why female carriers in this family are not affected. However, the additional case report (2), which was found via GeneMatcher, is a girl with two VUS. The first VUS (in ST8IA4) is located on chromosome 5 and might be inherited in an autosomal recessive manner. Since we did not find a VUS in the other ST8IA4 gene, it would be more likely that the second VUS (in SLC7A3) contributes to the phenotype of our patient, because this gene is located on the X-chromosome, and skewed inactivation could lead to more expression of the potential pathogenic SLC7A3 in the brain, possibly leading to epilepsy and psychiatric issues. In the same manner, other X-linked genes (e.g., NEXMIF or IQSEC2) were initially discovered in X-linked intellectual disability (XLID) families with unaffected carrier females, but later studies have shown that females could be affected as well.[15] [16]

SLC7A3, or the cationic amino acid transporter (CT-3), is almost exclusively expressed in the brain and modulates arginine transportation in neurons. Defective SLC7A3 can affect NMDA and mammalian target of rapamycin (mTOR) signaling pathways that both are crucial in neuronal development, plasticity and epilepsy. Consistently, SLC7A3 mutations have clinically been ascribed to a few males with ASD, a neurodevelopmental disorder.[6] Thus, a role of SLC7A3 also in epilepsy seems valid, which was present in 2 out 3 patients.

We have also summarized preclinical and/or clinical data of the 27 SLC genes that have linked to epilepsy ([Table 2]).[5] One of the most prevalent epilepsies related to SLC genes is Glut1 deficiency syndrome (Glut1DS), even though rare (incidence of ± 4/100 000 births).[17] Glut1DS should be considered in a patient with early onset absence epilepsy (< 4 years of age), myoclonic astatic epilepsy, absence epilepsies, or any epilepsy associated with a movement disorder.[18] Early diagnosis of Glut1DS is confirmed by a low cerebrospinal fluid (CSF) glucose and a pathogenic SLC2A1 variant. This syndrome is of high clinical importance, since a very effective treatment (ketogenic diet) can improve or even reverse symptoms, especially if started promptly after diagnosis.[17]

Table 2 Solute carrier transporters (SLC) and their preclinical and clinical relevance to epilepsy

SLC	Gene	Protein	Preclinical	Human (omim.org)	Gene panel
1	SLC1A2/EAAT2	Excitatory amino acid transporter 2 (glutamate)	Epilepsy	EE	Yes
2	SLC1A4/ASCT1	Alanine/serine/cysteine transporter 1	Schizophrenia	EE, developmental delay, microcephaly, severe ID	No[b]
3	SLC2A1/GLUT1	Glucose transporter 1	Epilepsy, metabolic dysfunction	GLUT1 deficiency syndrome, infantile seizures, ID, hypoglycorrhachia, microcephaly	Yes
4	SLC4A10/NBCn2	Na+/HCO3- cotransporter 2	Small brain ventricles, reduced neuronal excitation	Epilepsy, ID	No[b]
5	SLC6A1/GAT1	GABA transport protein 1	Epilepsy	Myoclonic-atonic epilepsy	Yes
6	SLC6A8/CrT	Creatine transporter	Cognitive deficits	Cerebral creatine deficiency syndrome [a]	Yes
7	SLC6A9/GlyT1	Na+/Cl− dependent glycine transporter 1	Epilepsy, schizophrenia	Glycine encephalopathy	No[b]
8	SLC6A11/GAT3	GABA transport protein 3	Epilepsy	Epilepsy, ID, stereotypic behavior	Yes
9	SLC6A12/BGT1	Betaine-GABA transporter 1	No	Epilepsy	No[b]
10[c]	SLC7A3/CAT3	Cationic amino acid transporter 3 (Arg, Lys, Orn)	No	ASD, epilepsy (this report)	No
11	SLC7A5/LAT1	L-type amino acid transporter (neutral amino acids)	Glioma	ASD	No
12	SLC9A6/NHE6	Na+/H+ exchanger 6	Christianson syndrome	ID, X-linked syndromic, Christianson type [a]	Yes
13	SLC12A2/NKCC1	Na+, K+, 2Cl− cotransporter 1	Epilepsy Stroke	Epilepsy	No[b]
14	SLC12A5/KCC2	K+−Cl− cotransporter 2	Epilepsy	Epilepsy	Yes
15	SLC13A5/NaCT	Na+/citrate cotransporter	No	Kohlschütter–Tönz syndrome [a], epilepsy	Yes
16	SLC16A4/MCT4	Monocarboxylate transporter 4	Glioma	Epilepsy	No
17	SLC16A7/MCT2	Monocarboxylate transporter 2	Epilepsy	Epilepsy	No
18	SLC17A6/VGLUT2	Vesicular glutamate transporter 2	Epilepsy	Gnathodiaphyseal dysplasia, tendinosis	No
19	SLC23A3/SVCT3	Na+-dependent vitamin C transporter 3	No	Epilepsy	No[b]
20	SLC25A12/AGC1	Aspartate glutamate carrier 1	Hypomyelination, neuronal defects	EE, early infantile, ASD	No[b]
21	SLC25A22/GC1	Mitochondrial glutamate carrier 1	No	EE, early infantile	No[b]
22	SLC25A39/CGI69	(Mitochondrial carrier variety of solutes)	No	Childhood absence epilepsy	No[b]
23	SLC25A42/MECREN	(Mitochondrial carrier variety of solutes)	No	EE, metabolic crises with variable encephalomyopathic features and neurologic regression	No
24	SLC30A1/ZNT1	Zinc transporter 1	Ischemic stroke Neonatal seizures	No	No
25	SLC32A1/VGAT	Vesicular GABA transporter	Epilepsy Anxiety Cortical dysplasia	No	No
26	SLC35A3/AMRS	(Nucleoside-sugar transporter family)	No	Epilepsy, ASD, Arthrogryposis, mental retardation, and seizures	No[b]
27	SLC55A1/LETM1	Leucine zipper-EF-hand containing transmembrane protein 1 (mitochondrial cation/proton exchangers Ca2+, K+, H+)	No	Epilepsy	No[b]
28	SLC62A1/ANKH	Progressive ankylosis protein homolog (ankylosis homolog) (pyrophosphate transporters)	No	Epilepsy	No[b]

Abbreviations: Arg, arginine; ASD, autism spectrum disorder; EE, epileptic encephalopathy; GABA, g-aminobutyric acid; ID, intellectual disability; Lys, lysine; Orn, Ornithine.

^a characterized by several features such as epileptic seizures

^b gene variants related to epilepsy though not part of the IPG epilepsy gene panel (12/27 genes, i.e., 44%)

^c gene variant discovered in a patient with epilepsy by our case report.

Overall, our study on SLC7A3 variants shows that all patients had developmental issues—including motor skills and language development—and two out of three patients had epilepsy. Although SLC7A3 is a potential contributor to epilepsy and developmental problems, additional functional studies on Xenopus oocytes quantifying the cationic amino acid transportation are necessary to confirm the pathogenicity of SLC7A3 mutations.[7]

• References

• 1 Symonds JD, McTague A. Epilepsy and developmental disorders: Next generation sequencing in the clinic. Eur J Paediatr Neurol 2020; 24: 15-23
• 2 Sisodiya SM, Whelan CD, Hatton SN. et al; ENIGMA Consortium Epilepsy Working Group. The ENIGMA-Epilepsy working group: Mapping disease from large data sets. Hum Brain Mapp 2020; (e-pub ahead of print)
• 3 Sands TT, Choi H. Genetic testing in pediatric epilepsy. Curr Neurol Neurosci Rep 2017; 17 (05) 45
• 4 Mei D, Parrini E, Marini C, Guerrini R. The impact of next-generation sequencing on the diagnosis and treatment of epilepsy in paediatric patients. Mol Diagn Ther 2017; 21 (04) 357-373
• 5 Hu C, Tao L, Cao X, Chen L. The solute carrier transporters and the brain: Physiological and pharmacological implications. Asian J Pharm Sci 2020; 15 (02) 131-144
• 6 Nava C, Rupp J, Boissel J-P. et al. Hypomorphic variants of cationic amino acid transporter 3 in males with autism spectrum disorders. Amino Acids 2015; 47 (12) 2647-2658
• 7 Gilles W, Vulcu SD, Liewald JF. et al. Monovalent cation conductance in Xenopus laevis oocytes expressing hCAT-3. Biochim Biophys Acta 2005; 1668 (02) 234-239
• 8 Siekierska A, Stamberger H, Deconinck T. et al; C4RCD Research Group, AR working group of the EuroEPINOMICS RES Consortium. Biallelic VARS variants cause developmental encephalopathy with microcephaly that is recapitulated in vars knockout zebrafish. Nat Commun 2019; 10 (01) 708
• 9 Li Q, Wang K. InterVar: clinical interpretation of genetic variants by the 2015 ACMG-AMP Guidelines. Am J Hum Genet 2017; 100 (02) 267-280
• 10 Schwarz JM, Cooper DN, Schuelke M, Seelow D. MutationTaster2: mutation prediction for the deep-sequencing age. Nat Methods 2014; 11 (04) 361-362
• 11 Sobreira N, Schiettecatte F, Valle D, Hamosh A. GeneMatcher: a matching tool for connecting investigators with an interest in the same gene. Hum Mutat 2015; 36 (10) 928-930
• 12 Caglayan AO, Yaghouti K, Kockaya T. et al. Biallelic ZNF335 mutations cause basal ganglia abnormality with progressive cerebral/cerebellar atrophy. J Neurogenet 2020; (November): 1-6
• 13 Stouffs K, Stergachis AB, Vanderhasselt T. et al. Expanding the clinical spectrum of biallelic ZNF335 variants. Clin Genet 2018; 94 (02) 246-251
• 14 Calhoun JD, Huffman AM, Bellinski I. et al. CACNA1H variants are not a cause of monogenic epilepsy. Hum Mutat 2020; 41 (06) 1138-1144
• 15 Mignot C, McMahon AC, Bar C. et al. IQSEC2-related encephalopathy in males and females: a comparative study including 37 novel patients. Genet Med 2019; 21 (04) 837-849
• 16 Stamberger H, Hammer TB, Gardella E. et al. NEXMIF encephalopathy: an X-linked disorder with male and female phenotypic patterns. Genet Med 2021; 23 (02) 363-373
• 17 Klepper J, Akman C, Armeno M. et al. Glut1 Deficiency Syndrome (Glut1DS): State of the art in 2020 and recommendations of the international Glut1DS study group. Epilepsia Open 2020; 5 (03) 354-365
• 18 Koch H, Weber YG. The glucose transporter type 1 (Glut1) syndromes. Epilepsy Behav 2019; 91: 90-93