Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2025 Jun 1.
Published in final edited form as: Am J Med Genet A. 2024 Feb 17;194(6):e63569. doi: 10.1002/ajmg.a.63569

Common epilepsy variants from the general population are not associated with epilepsy among individuals with tuberous sclerosis complex

Melissa A Richard 1, Philip J Lupo 1, Erik A Ehli 2, Mustafa Sahin 3, Darcy A Krueger 4,5, Joyce Y Wu 6,7, Elizabeth M Bebin 8, Kit Sing Au 9, Hope Northrup 9, Laura S Farach 9, TACERN Study Group
PMCID: PMC11060940  NIHMSID: NIHMS1971665  PMID: 38366765

Abstract

Common genetic variants identified in the general population have been found to increase phenotypic risks among individuals with certain genetic conditions. Up to 90% of individuals with tuberous sclerosis complex (TSC) are affected by some type of epilepsy, yet the common variants contributing to epilepsy risk in the general population have not been evaluated in the context of TSC-associated epilepsy. Such knowledge is important to help uncover the underlying pathogenesis of epilepsy in TSC which is not fully understood, and critical as uncontrolled epilepsy is a major problem in this population. To evaluate common genetic modifiers of epilepsy, our study pooled phenotypic and genotypic data from 369 individuals with TSC to evaluate known and novel epilepsy common variants. We did not find evidence of enhanced genetic penetrance for known epilepsy variants identified across the largest genome-wide association studies of epilepsy in the general population, but identified support for novel common epilepsy variants in the context of TSC. Specifically, we have identified a novel signal in SLC7A1 that may be functionally involved in pathways relevant to TSC and epilepsy. Our study highlights the need for further evaluation of genetic modifiers in TSC to aid in further understanding of epilepsy in TSC and improve outcomes.

Keywords: tuberous sclerosis complex, epilepsy, genetic modifiers, common variants

Introduction

Tuberous sclerosis complex (TSC, OMIM: 19110; 613254) is a genetic disorder caused by pathogenic variants in either of two tumor suppressor genes, TSC1 or TSC2, and subsequent over-activation of the mammalian target of rapamycin (mTOR) signaling pathway. Individuals affected by TSC develop benign hamartomas throughout the body and often experience neurological problems (Cusmai et al., 2011). The highest morbidity and mortality in TSC stems from associated neurological problems such as epilepsy (Asato & Hardan, 2004; Cusmai et al., 2011; Parthasarathy et al., 2021). Up to 90% of individuals with TSC experience some type of epilepsy (Asato & Hardan, 2004; Wong & Khong, 2006), including infantile spasms and intractable epilepsy that is difficult to control with antiseizure medication. Early control of epilepsy has been shown to be one of the best measures to reduce neurologic complications in TSC (Bombardieri et al., 2010; Gomez et al., 1982; Jansen et al., 2008; Jóźwiak et al., 2011) and thus the need for earlier screening and treatment has been emphasized in recent guidelines from the World TSC Conference (Northrup et al., 2021). Due to this need, studies on biomarkers of epilepsy in TSC are being conducted to help identify subsets of patients that should be followed more closely with more frequent EEGs or initiation of medication prior to clinical seizures.

One opportunity for a biomarker that can be detected as early as prenatally is the identification of modifier genes, that is, variants outside of TSC1 or TSC2, which alter epilepsy risk (Au et al., 2008; Dabora et al., 2002). Recent work demonstrates that common variants related to complex traits in the general population may have stronger effects among individuals with genetic conditions. Examples of common genetic modifiers include enhanced genetic penetrance of well-replicated risk variants in IKZF1, GATA3, ARID5B, and CDKN2A for leukemia risk among individuals with trisomy 21 (Brown et al., 2019) and enhanced prediction by polygenic risk scores for schizophrenia among individuals with 22q11.2 deletion syndrome (Davies et al., 2020). Epilepsy is a complex trait in the general population that demonstrates heritability with both common and rare variants contributing to epilepsy risk (Hani et al., 2015). Common variants, or single nucleotide polymorphisms (SNPs), have been found to alter the risk of epilepsy in the general population (ILAECCE, 2014; Koeleman, 2018; Speed et al., 2014); however, the modification of epilepsy risk by these variants among individuals with TSC has not been evaluated.

Our study aimed to characterize the role of common epilepsy-associated variants identified in the general population in the context of epilepsy in a TSC population. The ability to identify modifiers of epilepsy phenotypes in TSC could ultimately inform prognosis, treatment, personalized care, and better understanding of the underlying pathophysiology.

Methods

Editorial Policies and Ethical Considerations

This retrospective study underwent ethical review and received approval through The University of Texas Health Institutional Review Board (HSC-MS-20-0033). Written informed consent for research participation was obtained through each parent study of this assembled retrospective cohort.

Study population

We assembled a genetic cohort of individuals who met clinical or genetic diagnostic criteria for TSC through multiple data sources. All participants had genetic testing for TSC1 and TSC2, with reliable phenotypic information related to epilepsy. Participants in the TSC Alliance Natural History Study are aged 12 months or older with retrospective demographic, clinical, and imaging information. The TSC Autism Center of Excellence Research Network (TACERN, NCT 01780441) collects longitudinal prospective data on individuals with TSC beginning at 3 months of age. The Rare Disease Clinical Research Network (RDCRN) cohort collects three years of longitudinal prospective data on individuals with TSC beginning at 3 years of age. Finally, the UTHealth TSC Biorepository cohort collects retrospective clinical information on referral patients from TSC clinics. Through these data sources, all individuals had reliable demographic and phenotypic information including sex, seizure history, and presence or absence of cortical tubers. Information on seizures was obtained through ongoing follow-up of enrolled individuals; individuals with documented seizures had a median of 36 months of follow-up (range: 0-516 months) and individuals without any seizures through at least 12 months of age and the duration of follow-up were considered seizure-free (median: 48 months, range: 12-684 months). Genetic testing including sequencing and duplication/deletion analysis of TSC1 and TSC2 was performed for all participants and was used as a covariate with individuals stratified into groups of having TSC1 causative variant, TSC2 causative variant, or no mutation identified (NMI). Variants in TSC1 and TSC2 were classified using American College of Medical Genetics standards and guidelines (Richards et al., 2015). Pathogenic or likely pathogenic variants were considered causative. Individuals with benign, likely benign, variants of uncertain significance, or no variants were classified as no mutation identified (NMI).

Genotype data

All DNA samples were obtained from biobanked samples in the respective sub-studies and genotyped on the Illumina Global Screening array at the Avera Institute for Human Genetics. Raw intensity files were clustered and called using Genome Studio and filtered for quality control using PLINK (Chang et al., 2015). Markers were filtered to have call rate >98%, minor allele frequency (MAF) >1%, and non-significant departure from Hardy-Weinberg equilibrium (P>10−5); samples were filtered to have >95% call rates and assessed for heterozygosity. Array data were imputed to the Haplotype Reference Consortium r1.1 reference panel using the Michigan Imputation Server (Das et al., 2016) and filtered for imputation quality > 0.5 and MAF >5%. Imputed data were assessed for cryptic relatedness using PRIMUS (Staples et al., 2014) and principal components of ancestry were estimated using eigenstrat (Price et al., 2006). Analyses were restricted to unrelated individuals and adjusted for population stratification.

Identification of common epilepsy variants in the general population

We aimed to contextualize previously identified epilepsy-associated variants in the general population for their role in TSC-associated epilepsy. To identify variants in the general population, we referenced multiple epilepsy and epilepsy subtype genome-wide association studies from mega-analyses in the International League Against Epilepsy Consortium on Complex Epilepsies (ILAECCE) (ILAECCE, 2018). These analyses included up to 15,212 individuals with epilepsy (>2:1 controls:cases) and were conducted among individuals of European and trans-ancestral backgrounds. From public summary data available in the NHGRI-EBI GWAS Catalog, we included variants associated with all epilepsy types (GCST007343), childhood absence epilepsy (GCST007345), juvenile absence epilepsy (GCST007346), juvenile myoclonic epilepsy (GCST007347), generalized epilepsy with tonic-clonic seizures alone (GCST007348), focal lesion negative epilepsy (GCST007349), focal epilepsy with hippocampal sclerosis (GCST007350), focal lesion with lesion other than hippocampal sclerosis (GCST007351), focal epilepsy (GCST007352), and generalized epilepsy (GCST007353).

Evaluation of common genetic variants and epilepsy among individuals with TSC

We used logistic regression to assess the association of additively-modeled genetic variants with epilepsy in our TSC cohorts. Models were adjusted for TSC-causal genotype (TSC1, TSC2, or NMI), sex, presence of tubers, and principal components of ancestry, and filtered at MAF >5%. We then used three approaches to evaluate variants identified in the general population in the context of TSC-associated epilepsy. First, we aimed to evaluate the effect of top epilepsy index variants in the general population for their effect in TSC. In our TSC analyses, we directly evaluated the estimated odds ratio (OR), 95% confidence interval (CI), and P-value of 17 epilepsy-associated index variants with P-value<5x10−8 in the ILAECCE GWAS. Most variants reported in the ILAECCE GWAS have minor allele frequencies > 0.2; we estimated our TSC study has >80% power to detect an association of these variants with P<0.05 and OR > 1.9 (conversely, OR < 0.53). Second, we aimed to evaluate variants suggestively associated with multiple epilepsy phenotypes in the general population in the context of TSC-associated epilepsy. We generated stratified quantile-quantile (QQ) plots to evaluate the enrichment of variants associated with epilepsy in the general population in the context of TSC, with the underlying hypothesis that variants associated with epilepsy in the general population will be above the expected P-value levels in a TSC population. To achieve this, we selected variants from each respective ILAECCE GWAS associated with epilepsy at three P-value thresholds: P<0.05, P<0.001, and P<0.0001. For each variant selected at these thresholds, we then plotted the estimated P-value for association of that variant with TSC-associated epilepsy against the expected P-value based on median of the observed distribution of P-values. In our third approach, we sought to assess the power of our study design to identify common variants associated with epilepsy in TSC. To do this, we selected the top variant associated with epilepsy in our genome-wide analyses and functionally annotated the locus driving this signal as a positive control.

Results

Our study included 369 individuals affected by TSC who had phenotypic and genotypic information to evaluate common genetic modifiers of epilepsy. The study included slightly more males than females (51.5% male) and, as expected, the most frequent TSC genotype was TSC2 (65.9%, Table 1). As expected, epilepsy affected a large proportion of the TSC genetic cohort, and was present in 307 individuals (83.2%), and there was an even greater burden of cortical tubers in the cohort (90.8%).

Table 1.

Demographics of individuals in the assembled TSC genetic cohort (N=369).

n percent
Sex
 Male 190 51.5
 Female 179 48.5
TSC genotype
 TSC1 57 15.4
 TSC2 243 65.9
 NMI 69 18.7
Cortical tubers
 Present 335 90.8
 Absent 34 9.2
Epilepsy
 Present 307 83.2
 Absent 62 16.8
Open in a new tab

Abbreviations: no mutation identified (NMI), Tuberous Sclerosis Complex (TSC)

We identified 17 index variants reported in the GWAS Catalog as associated with any epilepsy phenotype (P-value < 5 x 10−8) in the ILAECCE GWAS. We assessed the association of these 17 index variants for association with epilepsy in TSC (Table 2) and found that rs6432877 achieved nominal association in our cohort (P-value < 0.05) but did not remain significant after Bonferroni correction for multiple testing (P-value < 0.0029, 0.05/17). The rs6432877 variant is located in the first intron of SCNA1, a voltage-dependent sodium channel gene highly expressed in the brain. However, the rs6432877-G risk allele was associated with increased epilepsy risk in ILAECCE, yet reduced epilepsy risk in TSC. Likewise, the direction of estimated epilepsy effects for these 17 risk alleles was consistent for only 7 of the variants in our TSC cohort, regardless of significance.

Table 2.

Index variants associated with epilepsy in the general population, as estimated in the TSC genetic cohort.

Accession / trait Index SNP Reported genes Region Risk
allele		 ILAECCE
GWAS		 TSC GWAS
Dir P-
value		 OR Dir P-
value
GCST007343 / epilepsy rs4671319 BCL11A, FANCL 2p16.1 G + 8E-09 1.19 + 0.41
rs6432877 SCN3A, SCN2A, TTC21B, SCN1A 2q24.3 G + 2E-13 0.61 − 0.030
rs4638568 HEATR3, BRD7 16q12.1 A − 4E-08 2.45 + 0.16
GCST007353 / generalized epilepsy rs1402398 FANCL, BCL11A 2p16.1 G + 1E-11 1.24 + 0.32
rs4665630 − 2p24.1 C + 4E-08 0.74 − 0.32
rs11890028 SCN3A, SCN2A, TTC21B, SCN1A 2q24.3 G − 5E-08 1.68 + 0.061
rs887696 STAT4 2q32.2 C + 3E-08 1.31 + 0.24
rs11943905 GABRA2 4p12 T + 4E-08 1.42 + 0.18
rs1044352 PCDH7 4p15.1 T + 2E-09 0.66 − 0.055
rs4596374 KCNN2 5q22.3 C + 7E-10 1.07 + 0.73
rs68082256 ATXN1 6p22.3 A − 2E-09 1.03 + 0.91
rs13200150 − 6q22.33 G − 6E-09 1.11 + 0.65
rs4794333 PNPO 17q21.32 C − 7E-09 0.91 − 0.66
rs2833098 GRIK1 21q22.11 G − 2E-08 1.19 + 0.42
GCST007352 / partial epilepsy rs2212656 SCN2A, SCN3A, TTC21B, SCN1A 2q24.3 A + 7E-09 0.64 − 0.051
GCST007350 / partial epilepsy rs1991545 C3orf33, SLC33A1, KCNAB1 3q25.31 A + 1E-11 2.09 + 0.098
rs1318322 GJA1 6q22.31 G + 7E-09 0.86 − 0.58
Open in a new tab

Abbreviations: direction of estimated effects (Dir), genome-wide association study (GWAS), International League Against Epilepsy Consortium on Complex Epilepsies (ILAECCE), odds ratio (OR), single nucleotide polymorphism (SNP), Tuberous Sclerosis Complex (TSC)

We next assessed the enrichment of suggestive epilepsy-associated variants in ILAECCE for association with epilepsy in TSC using stratified QQ plots (Figure 1). QQ plots are used to assess whether two data sets come from the same distribution and are useful to determine if two populations share genetic determinants. Inflation of the plotted data above the reference line would suggest an enrichment of ILAECCE variants in TSC epilepsy, whereas equal distribution or deflation suggests lack of variant enrichment. In total, although ILAECCE conducted GWAS for ten epilepsy phenotypes, we did not find evidence of enrichment for variants suggestively associated with any ILAECCE epilepsy phenotype in the context of TSC-associated epilepsy. In fact, the most deflation was observed for the two types of juvenile epilepsy assessed in ILAECCE, juvenile absence epilepsy and juvenile myoclonic epilepsy.

Figure 1. Stratified quantile-quantile (QQ) plots for genetic variants associated with epilepsy, estimated in the context of TSC-associated epilepsy.

Figure 1.

Open in a new tab

QQ plots are used to compare an observed and expected distribution to evaluate enrichment of traits within a population. The pattern of points above a reference line indicates inflation and the pattern of points below a reference line indicate deflation. The plotted variants are estimated in TSC, but selected selected based on three P-value thresholds for suggestive genome-wide association in the International League Against Epilepsy Consortium on Complex Epilepsies (ILAECCE). The QQ plots do not show strong inflation for ILAECCE variants in TSC-associated epilepsy.

Abbreviations: childhood absence epilepsy (CAE), juvenile absence epilepsy (JAE), juvenile myoclonic epilepsy (JME), generalized epilepsy with tonic-clonic seizures alone (GE-TC), focal lesion negative epilepsy (FE-LN), focal epilepsy with hippocampal sclerosis (FE-HS), focal lesion with lesion other than hippocampal sclerosis (FE-NHS), generalized epilepsy (GE).

Our top signal from genome-wide analyses revealed 38 variants mapped to 11 loci associated with epilepsy in TSC at P<10−5. There was no evidence of genomic inflation or deflation (lambda=1.04). The top variant rs1984030 maps to the first intron of SLC7A1 and linkage in the region further supports the association of this locus with epilepsy in TSC (Figure 2). The rs1984030-T risk allele (minor allele) is associated with a reduced risk for epilepsy (OR=0.19, 95% CI=0.10-0.37, P-value=1.7x10−6).

Figure 2. Regional association plot of SLC7A1 for association with epilepsy in TSC.

Figure 2.

Open in a new tab

The – log10(P-value) is plotted along the hg19 coordinates of a +/−400kb region around the index variant, rs1984030 (purple diamond). Points are color coded based on linkage patterns with the index variant using a 1000Genomes European reference panel. Variants in strong linkage with the index variant lend further support to common variant associations with epilepsy in this region.

Discussion

Overall, we found common variants that contribute to epilepsy risk in the general population did not explain epilepsy risk in our cohort of individuals with TSC. To our knowledge, this is the first study to evaluate the role of common, epilepsy-associated variants in a single gene disorder to determine whether common genetic risk factors in non-syndromic epilepsy contribute to epilepsy risk in the context of a syndromic type of epilepsy. Seizures result from an excitation/inhibition imbalance in the brain where excitation increases beyond an individual’s seizure threshold. Non-syndromic epilepsy has been demonstrated to be polygenic and a result of both common and rare variants (Hani et al., 2015), related both to the number of genes involved in the excitation/inhibition pathways and multiple small changes to the seizure threshold by each of these variants. Pathogenesis of epilepsy in TSC includes increased mTOR activation, altering synaptogenesis, causing an imbalance between excitation and inhibition, and ultimately lowers the seizure threshold (Curatolo et al., 2016). The effect of an underlying pathogenic variant in TSC1 or TSC2 on the epilepsy threshold appears to be substantial as epilepsy is present in up to 90% of individuals (Asato & Hardan, 2004; Wong & Khong, 2006); however, it is not understood why 10% of individuals with TSC do not develop epilepsy. Even more importantly, up to two thirds of individuals with TSC and epilepsy have the severe phenotype of drug resistant epilepsy, and the cause is unknown (Chu-Shore et al., 2010; Curatolo et al., 2012; Vignoli et al., 2013). The variability of the epilepsy phenotype may be due to additional factors, such as tuber burden or modifier genes, either lowering or increasing the epilepsy threshold, similar to the polygenic nature of non-syndromic epilepsy. Interestingly, in this study we determined that the common SNPs associated with non-syndromic epilepsy are not playing a role in modifying the epilepsy phenotype in TSC.

Instead, other common SNPs appear to be associated with epilepsy risk in TSC. This differs from other genetic disorders where the same common variants that increase risk for certain phenotypes in the general population also increase risk for that phenotype in the population with the genetic disorder. For instance, the same common variants associated with increased risk for leukemia in the general population are also found to increase risk for leukemia in the trisomy 21 population, partially explaining the variable expressivity of leukemia in trisomy 21 (Brown et al., 2019). Similarly, SNPs associated with increased risk of psychiatric disorders in the general population are also found to increase risk of psychiatric disorders in individuals with 22q11 (Davies et al., 2020).

Correspondingly, previous pharmacologic evidence has demonstrated that the underlying pathophysiology for the most prominent early seizure type in TSC, infantile spasms, may be different than the general population. Some of the first line medications for infantile spasms in the general population, such as adrenocorticotropic hormone and corticosteroids (Nelson, 2015), are only recommended as second line therapy as vigabatrin is known to be efficacious (Northrup et al., 2021). Likewise, vigabatrin is 95% effective at treating infantile spasms in patients with TSC, but only 50% effective in the general population (Curatolo et al., 2001). Vigabatrin works by inhibiting GABA transaminase (encoded by ABAT) to increase GABA concentrations in the brain; thereby stifling the excitatory process that can initiate seizure activity (Ben-Menachem, 2011) and likely inhibits mTOR signaling (Zhang et al., 2013).

Other reasons for lack of common variant generalization include the forms of epilepsy included in the general population study. Likely infantile spasms did not comprise a substantial proportion of the generalized forms of epilepsy included in the ILAECCE study, as they affect 1.5 to 2 per 10,000 children in the general population (Cowan & Hudson, 1991), but affect 2 out of 3 children with TSC (Northrup et al., 2021). In addition, other epilepsy subtypes may differ between these two populations and epilepsy subtypes have different genetic profiles, yet there are no known common variant profiles for infantile spasms in the general population and rare variants have been implicated as the cause in 29% (Peng et al., 2022).

However, our genome-wide analyses suggest that variants in SLC7A1 may alter epilepsy risk in TSC. We pooled phenotypic and genomic data from some of the largest studies of TSC to date, and therefore will need to replicate these findings as more resources become available to the TSC research community. Yet, our finding in SLC7A1 may have functional implications. A de novo microdeletion on 13q12.3 has been reported in a case with a novel epilepsy phenotype, and this 2.324 Mb microdeletion encompasses SLC7A1, KATNAL1, and other genes (Wang et al., 2020). The case report authors focused on the high expression of KATNAL1 in central nervous system tissue, yet other lines of evidence suggest a role of SLC7A1 in epileptogenesis. The rs1984030 variant identified in our study is associated with SLC7A1 and UBL3 expression when analyzed in whole blood and thought to affect expression in many other tissues (Westra et al., 2013). SLC7A1 encodes the high-affinity cationic amino acid transporter 1 (CAT-1) protein, a solute carrier for arginine, lysine, and ornithine that can competitively transport these amino acids at low extracellular concentrations. Arginine potentiates GABA synaptic transmission as a substrate of nitric oxide production (Shen et al., 1997) and is expressed in microendothelial cells that make up the blood brain barrier (Watson et al., 2016). Notably, arginine function has been linked to TSC1 and TSC2 function (van Slegtenhorst et al., 2004), and arginine regulated through SLC7 transporters is a key regulator of mTOR signaling by altering mTORC1 activity (Carroll et al., 2016; Iffland et al., 2019; Jungnickel et al., 2018; Tang et al., 2019). Recently, in vivo CRISPR deletion of Slc7a1 greatly reduced mTORC1 signaling, demonstrating the reprogramming of memory T cells through this mechanism (Huang et al., 2021). Further, the role of arginine in the mTOR pathway has been directly tied to TSC and epilepsy, with alternate amino acids pathways now termed “GATORopathies” (Iffland et al., 2019). Pre-clinical drosophila models have demonstrated Tsc1 and Tsc2 function together to antagonize insulin signaling (Potter et al., 2001), and in human fetal cells insulin was determined a driver of SLC7A1 expression and the transport of arginine by CAT-1 to regulate nitric oxide in vascular endothelium (González et al., 2011). Together, these findings suggest further studies of the role of SLC7A1 in epileptogenesis are warranted.

Our modeling approach accounted for clinical information we previously identified as associated with epilepsy in TSC (Farach et al., 2020), but had some limited detail in data availability to harmonize information across the cohorts included in our study. Specifically, we were not able to examine individual epilepsy subtypes as some cohorts collected detailed epilepsy phenotyping, while others provided information only on the presence or absence of epilepsy. We adjusted models for the presence of cortical tubers. Both the presence of cortical tubers and the cortical tuber load correlate with degradation of the underlying white matter integrity and neurologic functioning (Im et al., 2016; Peters et al., 2012). However, the specific site of tubers within the cortex is largely unrelated to individual risk for epilepsy (Wong & Khong, 2006), so we did not account for tuber location in our study. Given the complex nature of TSC-associated epilepsy, our study suggests further investigation into genetic modifiers of epilepsy subtypes is greatly needed as the population does not appear to share common modifiers associated with epilepsy in the general population. Particular focus on infantile spasms and drug-resistant epilepsy may be helpful to uncover the underlying pathophysiology and thus aid in treatment. Overall, the ability to identify common genetic modifiers could aid in earlier treatment, personalized surveillance and pharmacologic plans, and improved quality of life for individuals with TSC.

Our study used multiple approaches to characterize common variants that may modify epilepsy risk in TSC. We did not observe evidence of enhanced genetic penetrance of previously identified epilepsy variants and evidence suggests novel common variants may increase epilepsy risk among individuals with TSC.

FUNDING INFORMATION

This work was supported by Autism Center of Excellence Network [1U01NS082320-01], the Developmental Synaptopathies Consortium [1U54NS092090-01], and the Department of Defense [W81XWH1810537]. The Developmental Synaptopathies Consortium [U54NS092090] is part of the Rare Diseases Clinical Research Network (RDCRN), an initiative of the Office of Rare Diseases Research (ORDR), National Center for Advancing Translational Sciences (NCATS). Research reported in this publication was supported by the National Institute of Neurological Disorders and Stroke of the National Institutes of Health (NINDS), Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD), National Institute of Mental Health (NIMH) and National Center for Advancing Translational Sciences (NCATS).

We are sincerely indebted to the generosity of the families and patients in TSC clinics across the United States who contributed their time and effort to this study. We would also like to thank the Tuberous Sclerosis Alliance for their continued support in TSC research.

Footnotes

CONFLICT OF INTEREST STATEMENT

The authors have no conflicts of interest to report.

REFERENCES

  1. Asato MR, & Hardan AY (2004). Neuropsychiatric problems in tuberous sclerosis complex. J Child Neurol, 19(4), 241–249. 10.1177/088307380401900401 [DOI] [PubMed] [Google Scholar]
  2. Au KS, Ward CH, & Northrup H (2008). Tuberous sclerosis complex: disease modifiers and treatments. Current Opinion in Pediatrics, 20(6), 628–633. 10.1097/MOP.0b013e328318c529 [DOI] [PubMed] [Google Scholar]
  3. Ben-Menachem E. (2011). Mechanism of action of vigabatrin: correcting misperceptions. Acta Neurol Scand Suppl(192), 5–15. 10.1111/j.1600-0404.2011.01596.x [DOI] [PubMed] [Google Scholar]
  4. Bombardieri R, Pinci M, Moavero R, Cerminara C, & Curatolo P (2010). Early control of seizures improves long-term outcome in children with tuberous sclerosis complex. European Journal of Paediatric Neurology, 14(2), 146–149. 10.1016/j.ejpn.2009.03.003 [DOI] [PubMed] [Google Scholar]
  5. Brown AL, de Smith AJ, Gant VU, Yang W, Scheurer ME, Walsh KM, … Rabin KR (2019). Inherited genetic susceptibility to acute lymphoblastic leukemia in Down syndrome. Blood, 134(15), 1227–1237. 10.1182/blood.2018890764 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Carroll B, Maetzel D, Maddocks OD, Otten G, Ratcliff M, Smith GR, … Korolchuk VI (2016). Control of TSC2-Rheb signaling axis by arginine regulates mTORC1 activity. Elife, 5. 10.7554/eLife.11058 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chang CC, Chow CC, Tellier LC, Vattikuti S, Purcell SM, & Lee JJ (2015). Second-generation PLINK: rising to the challenge of larger and richer datasets. Gigascience, 4, 7. 10.1186/s13742-015-0047-8 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Chu-Shore CJ, Major P, Camposano S, Muzykewicz D, & Thiele EA (2010). The natural history of epilepsy in tuberous sclerosis complex. Epilepsia, 51(7), 1236–1241. 10.1111/j.1528-1167.2009.02474.x [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Cowan LD, & Hudson LS (1991). The epidemiology and natural history of infantile spasms. J Child Neurol, 6(4), 355–364. 10.1177/088307389100600412 [DOI] [PubMed] [Google Scholar]
  10. Curatolo P, Bjørnvold M, Dill PE, Ferreira JC, Feucht M, Hertzberg C, … Zonnenberg BA (2016). The Role of mTOR Inhibitors in the Treatment of Patients with Tuberous Sclerosis Complex: Evidence-based and Expert Opinions. Drugs, 76(5), 551–565. 10.1007/s40265-016-0552-9 [DOI] [PubMed] [Google Scholar]
  11. Curatolo P, Jóźwiak S, & Nabbout R (2012). Management of epilepsy associated with tuberous sclerosis complex (TSC): clinical recommendations. Eur J Paediatr Neurol, 16(6), 582–586. 10.1016/j.ejpn.2012.05.004 [DOI] [PubMed] [Google Scholar]
  12. Curatolo P, Verdecchia M, & Bombardieri R (2001). Vigabatrin for tuberous sclerosis complex. Brain Dev, 23(7), 649–653. 10.1016/s0387-7604(01)00290-x [DOI] [PubMed] [Google Scholar]
  13. Cusmai R, Moavero R, Bombardieri R, Vigevano F, & Curatolo P (2011). Long-term neurological outcome in children with early-onset epilepsy associated with tuberous sclerosis. Epilepsy Behav, 22(4), 735–739. 10.1016/j.yebeh.2011.08.037 [DOI] [PubMed] [Google Scholar]
  14. Dabora SL, Roberts P, Nieto A, Perez R, Jozwiak S, Franz D, … Kwiatkowski DJ (2002). Association between a high-expressing interferon-gamma allele and a lower frequency of kidney angiomyolipomas in TSC2 patients. American Journal of Human Genetics, 71(4), 750–758. 10.1086/342718 [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Das S, Forer L, Schönherr S, Sidore C, Locke AE, Kwong A, … Fuchsberger C (2016). Next-generation genotype imputation service and methods. Nat Genet, 48(10), 1284–1287. 10.1038/ng.3656 [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Davies RW, Fiksinski AM, Breetvelt EJ, Williams NM, Hooper SR, Monfeuga T, … Vorstman JAS (2020). Using common genetic variation to examine phenotypic expression and risk prediction in 22q11.2 deletion syndrome. Nat Med, 26(12), 1912–1918. 10.1038/s41591-020-1103-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Farach LS, Richard MA, Lupo PJ, Sahin M, Krueger DA, Wu JY, … Northrup H (2020). Epilepsy Risk Prediction Model for Patients With Tuberous Sclerosis Complex. Pediatr Neurol, 113, 46–50. 10.1016/j.pediatrneurol.2020.07.015 [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Gomez MR, Kuntz NL, & Westmoreland BF (1982). Tuberous sclerosis, early onset of seizures, and mental subnormality: study of discordant homozygous twins. Neurology, 32(6), 604–611. 10.1212/wnl.32.6.604 [DOI] [PubMed] [Google Scholar]
  19. González M, Gallardo V, Rodríguez N, Salomón C, Westermeier F, Guzmán-Gutiérrez E, … Sobrevia L (2011). Insulin-stimulated L-arginine transport requires SLC7A1 gene expression and is associated with human umbilical vein relaxation. J Cell Physiol, 226(11), 2916–2924. 10.1002/jcp.22635 [DOI] [PubMed] [Google Scholar]
  20. Hani AJ, Mikati HM, & Mikati MA (2015). Genetics of pediatric epilepsy. Pediatr Clin North Am, 62(3), 703–722. 10.1016/j.pcl.2015.03.013 [DOI] [PubMed] [Google Scholar]
  21. Huang H, Zhou P, Wei J, Long L, Shi H, Dhungana Y, … Chi H (2021). In vivo CRISPR screening reveals nutrient signaling processes underpinning CD8(+) T cell fate decisions. Cell, 184(5), 1245–1261.e1221. 10.1016/j.cell.2021.02.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Iffland PH 2nd, Carson V, Bordey A, & Crino PB (2019). GATORopathies: The role of amino acid regulatory gene mutations in epilepsy and cortical malformations. Epilepsia, 60(11), 2163–2173. 10.1111/epi.16370 [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. ILAECCE. (2014). Genetic determinants of common epilepsies: a meta-analysis of genome-wide association studies. Lancet Neurol, 13(9), 893–903. 10.1016/s1474-4422(14)70171-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. ILAECCE. (2018). Genome-wide mega-analysis identifies 16 loci and highlights diverse biological mechanisms in the common epilepsies. Nat Commun, 9(1), 5269. 10.1038/s41467-018-07524-z [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Im K, Ahtam B, Haehn D, Peters JM, Warfield SK, Sahin M, & Ellen Grant P (2016). Altered Structural Brain Networks in Tuberous Sclerosis Complex. Cerebral Cortex, 26(5), 2046–2058. 10.1093/cercor/bhv026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Jansen FE, Vincken KL, Algra A, Anbeek P, Braams O, Nellist M, … van Nieuwenhuizen O (2008). Cognitive impairment in tuberous sclerosis complex is a multifactorial condition. Neurology, 70(12), 916–923. 10.1212/01.wnl.0000280579.04974.c0 [DOI] [PubMed] [Google Scholar]
  27. Jóźwiak S, Kotulska K, Domańska-Pakieła D, Lojszczyk B, Syczewska M, Chmielewski D, … Respondek-Liberska M (2011). Antiepileptic treatment before the onset of seizures reduces epilepsy severity and risk of mental retardation in infants with tuberous sclerosis complex. European Journal of Paediatric Neurology, 15(5), 424–431. 10.1016/j.ejpn.2011.03.010 [DOI] [PubMed] [Google Scholar]
  28. Jungnickel KEJ, Parker JL, & Newstead S (2018). Structural basis for amino acid transport by the CAT family of SLC7 transporters. Nat Commun, 9(1), 550. 10.1038/s41467-018-03066-6 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Koeleman BPC (2018). What do genetic studies tell us about the heritable basis of common epilepsy? Polygenic or complex epilepsy? Neurosci Lett, 667, 10–16. 10.1016/j.neulet.2017.03.042 [DOI] [PubMed] [Google Scholar]
  30. Nelson GR (2015). Management of infantile spasms. Transl Pediatr, 4(4), 260–270. 10.3978/j.issn.2224-4336.2015.09.01 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Northrup H, Aronow ME, Bebin EM, Bissler J, Darling TN, de Vries PJ, … Krueger DA (2021). Updated International Tuberous Sclerosis Complex Diagnostic Criteria and Surveillance and Management Recommendations. Pediatr Neurol, 123, 50–66. 10.1016/j.pediatrneurol.2021.07.011 [DOI] [PubMed] [Google Scholar]
  32. Parthasarathy S, Mahalingam R, Melchiorre J, Harowitz J, & Devinsky O (2021). Mortality in tuberous sclerosis complex. Epilepsy Behav, 121(Pt A), 108032. 10.1016/j.yebeh.2021.108032 [DOI] [PubMed] [Google Scholar]
  33. Peng P, Kessi M, Mao L, He F, Zhang C, Chen C, … Peng J (2022). Etiologic Classification of 541 Infantile Spasms Cases: A Cohort Study. Front Pediatr, 10, 774828. 10.3389/fped.2022.774828 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Peters JM, Sahin M, Vogel-Farley VK, Jeste SS, Nelson CA 3rd, Gregas MC, … Warfield SK (2012). Loss of white matter microstructural integrity is associated with adverse neurological outcome in tuberous sclerosis complex. Academic Radiology, 19(1), 17–25. 10.1016/j.acra.2011.08.016 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Potter CJ, Huang H, & Xu T (2001). Drosophila Tsc1 functions with Tsc2 to antagonize insulin signaling in regulating cell growth, cell proliferation, and organ size. Cell, 105(3), 357–368. 10.1016/s0092-8674(01)00333-6 [DOI] [PubMed] [Google Scholar]
  36. Price AL, Patterson NJ, Plenge RM, Weinblatt ME, Shadick NA, & Reich D (2006). Principal components analysis corrects for stratification in genome-wide association studies. Nat Genet, 38(8), 904–909. 10.1038/ng1847 [DOI] [PubMed] [Google Scholar]
  37. Richards S, Aziz N, Bale S, Bick D, Das S, Gastier-Foster J, … Rehm HL (2015). Standards and guidelines for the interpretation of sequence variants: a joint consensus recommendation of the American College of Medical Genetics and Genomics and the Association for Molecular Pathology. Genet Med, 17(5), 405–424. 10.1038/gim.2015.30 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Shen KZ, Cox BA, & Johnson SW (1997). L-arginine potentiates GABA-mediated synaptic transmission by a nitric oxide-independent mechanism in rat dopamine neurons. Neuroscience, 79(3), 649–658. 10.1016/s0306-4522(97)00024-9 [DOI] [PubMed] [Google Scholar]
  39. Speed D, O'Brien TJ, Palotie A, Shkura K, Marson AG, Balding DJ, & Johnson MR (2014). Describing the genetic architecture of epilepsy through heritability analysis. Brain, 137(Pt 10), 2680–2689. 10.1093/brain/awu206 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Staples J, Qiao D, Cho MH, Silverman EK, Nickerson DA, & Below JE (2014). PRIMUS: rapid reconstruction of pedigrees from genome-wide estimates of identity by descent. Am J Hum Genet, 95(5), 553–564. 10.1016/j.ajhg.2014.10.005 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Tang Y, El-Chemaly S, Taveira-Dasilva A, Goldberg HJ, Bagwe S, Rosas IO, … Henske EP (2019). Alterations in Polyamine Metabolism in Patients With Lymphangioleiomyomatosis and Tuberous Sclerosis Complex 2-Deficient Cells. Chest, 156(6), 1137–1148. 10.1016/j.chest.2019.05.038 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. van Slegtenhorst M, Carr E, Stoyanova R, Kruger WD, & Henske EP (2004). Tsc1+ and tsc2+ regulate arginine uptake and metabolism in Schizosaccharomyces pombe. Journal of Biological Chemistry, 279(13), 12706–12713. 10.1074/jbc.M313874200 [DOI] [PubMed] [Google Scholar]
  43. Vignoli A, La Briola F, Turner K, Scornavacca G, Chiesa V, Zambrelli E, … Canevini MP (2013). Epilepsy in TSC: certain etiology does not mean certain prognosis. Epilepsia, 54(12), 2134–2142. 10.1111/epi.12430 [DOI] [PubMed] [Google Scholar]
  44. Wang M, Li B, Liao Z, Jia Y, & Fu Y (2020). A novel phenotype of 13q12.3 microdeletion characterized by epilepsy in an Asian child: a case report. BMC Med Genomics, 13(1), 144. 10.1186/s12920-020-00801-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Watson CP, Pazarentzos E, Fidanboylu M, Padilla B, Brown R, & Thomas SA (2016). The transporter and permeability interactions of asymmetric dimethylarginine (ADMA) and L-arginine with the human blood-brain barrier in vitro. Brain Research, 1648(Pt A), 232–242. 10.1016/j.brainres.2016.07.026 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Westra HJ, Peters MJ, Esko T, Yaghootkar H, Schurmann C, Kettunen J, … Franke L (2013). Systematic identification of trans eQTLs as putative drivers of known disease associations. Nature Genetics, 45(10), 1238–1243. 10.1038/ng.2756 [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Wong V, & Khong PL (2006). Tuberous sclerosis complex: correlation of magnetic resonance imaging (MRI) findings with comorbidities. J Child Neurol, 21(2), 99–105. 10.1177/08830738060210020901 [DOI] [PubMed] [Google Scholar]
  48. Zhang B, McDaniel SS, Rensing NR, & Wong M (2013). Vigabatrin inhibits seizures and mTOR pathway activation in a mouse model of tuberous sclerosis complex. PLoS One, 8(2), e57445. 10.1371/journal.pone.0057445 [DOI] [PMC free article] [PubMed] [Google Scholar]

RESOURCES