Address correspondence to Thomas Sander, Cologne Center for Genomics, University of Cologne, Weyertal 115b, 50931 Cologne, Germany. E-mail: firstname.lastname@example.org
Structural variations disrupting the gene encoding the neuron-specific splicing regulator RBFOX1 have been reported in three patients exhibiting epilepsy in comorbidity with other neuropsychiatric disorders. Consistently, the Rbfox1 knockout mouse model showed an increased susceptibility of seizures. The present candidate gene study tested whether exon-disrupting deletions of RBFOX1 increase the risk of idiopathic generalized epilepsies (IGEs), representing the largest group of genetically determined epilepsies.
Screening of microdeletions (size: >40 kb, coverage >20 markers) affecting the genomic sequence of the RBFOX1 gene was carried out by high-resolution single-nucleotide polymorphism (SNP) arrays in 1,408 European patients with idiopathic generalized epilepsy (IGE) and 2,256 population controls. Validation of RBFOX1 deletions and familial segregation analysis were performed by quantitative polymerase chain reaction (qPCR).
We detected five exon-disrupting RBFOX1 deletions in the IGE patients, whereas none was observed in the controls (p = 0.008, Fisher's exact test). The size of the exonic deletions ranged from 68 to 896 kb and affected the untranslated 5′-terminal RBFOX1 exons. Segregation analysis in four families indicated that the deletions were inherited, display incomplete penetrance, and heterogeneous cosegregation patterns with IGE.
Rare deletions affecting the untranslated 5′-terminal RBFOX1 exons increase risk of common IGE syndromes. Variable expressivity, incomplete penetrance, and heterogeneous cosegregation patterns suggest that RBFOX1 deletions act as susceptibility factor in a genetically complex etiology, where heterogeneous combinations of genetic factors determine the disease phenotype.
The idiopathic generalized epilepsies (IGEs) affect up to 0.3% of the general population and account for 30% of all epilepsies (Jallon et al., 2001). Genetic factors play a predominant role in the etiology of common IGE syndromes. Heritability estimates are >80% and recurrence risk for first-degree relatives varies between 4% and 9% depending on the IGE subtype (Helbig et al., 2008). The genetic architecture is likely to display a biologic continuum, in which a small fraction follows monogenic inheritance, whereas the majority of IGE patients presumably display an oligogenic/polygenic predisposition. Molecular genetic studies have identified causative gene mutations in mainly rare monogenic forms of genetic epilepsies. Most of the currently known genes for human genetic epilepsies encode voltage-gated or ligand-gated ion channels (Reid et al., 2009; Pandolfo, 2011). Despite extensive research, the majority of genetic variants predisposing to common IGE syndromes remain elusive.
Large-scale analysis of structural genomic variations using high-resolution whole-genome oligonucleotide arrays suggests that copy number variations (CNVs) play a substantial role in about 3% of patients with idiopathic epilepsies (de Kovel et al., 2010; Heinzen et al., 2010; Mefford et al., 2011). Recurrent microdeletions on 15q11.2, 15q13.3, and 16p13.11 increase risk of IGE and a wide range of neurodevelopmental disorders (Helbig et al., 2009; de Kovel et al., 2010; Heinzen et al., 2010; for review see Hochstenbach et al., 2011). The genes deleted by these microdeletions are thought to play a key role in the regulation of neuronal excitation and cortical synchronization.
Structural variations disrupting the gene encoding the neuronal splicing regulator RBFOX1 (also assigned as A2BP1, HRNBP1, or FOX1) have been reported in three patients exhibiting epilepsy in comorbidity with autism, intellectual disability, or pontocerebellar hypoplasia (Bhalla et al., 2004; Martin et al., 2007; Gallant et al., 2011). The RBFOX1 gene is located in the chromosomal region 16p13.3 to which a linkage locus for photoparoxysmal response in families of IGE subjects has been mapped (Pinto et al., 2005). The RBFOX1 gene plays a key role in the regulation of neuronal excitation and influences susceptibility of epilepsy (Gehman et al., 2011; Voineagu et al., 2011). The RBFOX1 protein regulates splicing of many neuronal transcripts by binding the sequence (U)GCAUG in introns flanking alternative exons (Jin et al., 2003; Auweter et al., 2006; Voineagu et al., 2011; Fogel et al., 2012). A number of RBFOX1 target transcripts (e.g., SNAP25, SCN8A, GRIN1, GABRG2, DCX, GAD2, KCNQ2, SLC12A5, SV2B, SYN1) have been implicated to play a role in epileptogenesis (Barnby et al., 2005; Corradini et al., 2009; Papale et al., 2009; Pandolfo, 2011; Fogel et al., 2012; Veeramah et al., 2012) and show differentially spliced RNA transcripts in Rbfox1 knockout mice (Gehman et al., 2011) Notably, brain-specific homozygous and heterozygous Rbfox1 knockouts in mice do not alter brain morphology but display spontaneous seizures and a dramatic epileptogenic response to kainic acid resulting in status epilepticus (Gehman et al., 2011). Consistent with the splicing alterations in mice, a RNA interference–mediated 50% knockdown of RBFOX1 transcripts in human neurons changes the alternative splicing pattern and expression of primarily neuronal genes involved in synapse formation and function (Voineagu et al., 2011; Fogel et al., 2012).
The present candidate gene association study tested whether exon-disrupting deletions of RBFOX1 increase risk of common IGE syndromes. We found a significant excess of exon-disrupting deletions of the RBFOX1 gene in IGE patients compared to population controls. Familial cosegregation analysis implicates that exon-disrupting RBFOX1 deletions represent susceptibility factors that increase risk of IGE but are not sufficient for the expression of IGE in most of the families.
Subjects and Methods
The study protocol was approved by the local institutional review boards of the participating centers, and all study participants gave informed consent. The patients with common IGE syndromes were recruited as a concerted effort of epilepsy genetics programs integrated in the European EPICURE Project (http://www.epicureproject.eu; EPICURE Consortium et al., 2012). Phenotyping and diagnostic classification of IGE syndromes were carried out according to standardized protocols (available at: http://portal.ccg.uni-koeln.de/ccg/research/epilepsy-genetics/sampling-procedure). Patients with IGE exhibit unprovoked generalized seizures but are typically otherwise normal and have no anatomic brain abnormalities (Commission on Classification & Terminology of the International League Against Epilepsy, 1989; Nordli, 2005). Accordingly, the ascertainment scheme applied in this multicenter study did not include IGE patients with severe intellectual disability (no basic education, permanently requiring professional support in their daily life). All subjects of the case–control association cohorts were typed by the Affymetrix Genome-Wide Human SNP Array 6.0 (Affymetrix, Santa Clara, CA, U.S.A.). To ensure highly confident CNV calls, we excluded all individuals (351 of 4,015) carrying a genome-wide excess of more than 50 microdeletions (size >40 kb, coverage >20 markers) prior to the assessment of RBFOX1 microdeletions (Elia et al., 2012).
The case–control sample included in this candidate gene CNV study comprised 1,408 unrelated patients with IGE of self-identified Northwestern European ancestry (869 females/539 males; childhood absence epilepsy [CAE] n = 413, juvenile absence epilepsy [JAE] n = 207, unspecified idiopathic absence epilepsy [IAE] n=7, juvenile myoclonic epilepsy [JME] n=557, epilepsies with generalized tonic–clonic seizures alone [EGTCS] n=224) and 2,256 German population controls (1,077 females/1,179 males). Array data of 2,256 German control subjects were obtained from the PopGen biobank (University Hospital Schleswig-Holstein, Campus Kiel, Kiel, Germany) and the KORA (Cooperative Health Research in the Region of Augsburg) research platform representing epidemiologically recruited cohorts from the Northern (Schleswig-Holstein, PopGen) and Southern (Augsburg, KORA) regions of Germany. The control subjects have not been screened for epilepsy or other neurodevelopmental disorders. EIGENSTRAT principal component analysis (Price et al., 2006) was applied to remove ancestry outliers and to match the European ancestry of the case–control cohorts (EPICURE Consortium et al., 2012).
RBFOX1 microdeletion screening
For all DNA samples of the case–control cohorts, we assessed the signal intensities of 1.8 million probe sets on the Affymetrix Genome-Wide Human SNP Array 6.0. CNV analysis of all Affymetrix SNP 6.0 arrays was performed at the Cologne Center for Genomics, using the algorithm implemented in the Affymetrix Genotyping Console version 4.1.1. To achieve high accuracy in CNV calling across Affymetrix SNP 6.0 arrays processed at different laboratories, microdeletion screening was restricted to deletions covered by at least 20 probes and spanning >40 kb in size (Pinto et al., 2011).
Array-based screening of RBFOX1 deletions captured all deletions affecting the genomic sequence of the RBFOX1 gene (chr16:6,069,131-7,763,339, human genome build 37/hg19). The RBFOX1 gene is located in the chromosomal region 16p13.3 and consists of 21 exons (NM_018723.3), which form six mRNA transcripts encoding five known protein isoforms (Fig. 1; RBFOX1 exon annotation adapted from Fogel et al., 2012). Notably the 5′-terminal exons 1, 2, 3, 1B and part of exon 4 are untranslated (Fig. 1). All potential RBFOX1 microdeletions were manually inspected for the regional SNP heterozygosity state and log2 ratios of the signal intensities to exclude technical artifacts. Subsequently, the copy number state of all RBFOX1 microdeletions identified by the array-based CNV analysis was examined by real-time quantitative PCR (qPCR), using seven TaqMan CNV assays covering the 5′-terminal RBFOX1 exons 1-4 (Life Technologies, Carlsbad, CA, U.S.A.;Fig. S1).
Case–control association analysis was carried out using a two-sided Fisher's exact test.
Detection of RBFOX1 deletions in patients with IGE and controls
Microdeletions (size >40 kb, coverage >20 markers) affecting the genomic sequence of the RBFOX1 gene were found in 8 (0.6%) of 1,408 individuals with IGE, whereas two deletions were observed in 2,256 controls (p = 0.017, Fisher's exact test; odds ratio [OR] 6.4, 95% confidence interval [CI] 1.2–62.35; Figs 1 and S1). The size of the deletions ranged from 41 to 896 kb. All 10 RBFOX1 deletions were located in the 5′-terminal RBFOX1 region encompassing the untranslated exons 1–4. Remarkably, the two RBFOX1 deletions observed in the controls were both located in intronic sequences and were smaller (41 and 56 kb) than the deletions observed in the IGE patients (68–896 kb; Fig. S1). Specifically exon-disrupting RBFOX1 deletions were present in 5 (0.35%) of 1,408 individuals with IGE (Table 1) and none was detected in 2,256 controls (p = 0.008, Fisher's exact test; Figs 1 and S2). The hemizygous copy number state of all RBFOX1 deletions detected by the array-based CNV scan could be confirmed in the IGE patients by TaqMan qPCR assays. DNA samples of the control subjects were not available for qPCR validation.
Table 1. RBFOX1 exon-disrupting deletions in IGE index patients
Survey on RBFOX1 exon-disrupting deletions in index patients with idiopathic generalized epilepsies.
Values indicate the age-of-onset in years.
Developmental delay, LD, sudden death
No neuropsychiatric disorders
ASD, LD, myopia
No neuropsychiatric disorders
No neuropsychiatric disorders
Familial segregation and comorbidity analysis of the exonic RBFOX1 deletions
The segregation of exonic RBFOX1 deletions identified in the IGE index cases was tracked in four families (Fig. 2). The copy number status of the RBFOX1 was assessed in 20 available family members using TaqMan qPCR assays (Table S1). In total, 12 family members carried an exon-disrupting RBFOX1 deletion (6 females, 6 males). All RBFOX1 deletions identified in the IGE index patients were inherited. Overall, the deletions were transmitted five times maternally and one time paternally. Seven of 12 deletion carriers were affected by IGE, and five carriers were clinically unaffected. Seven of nine IGE patients investigated carried an exonic RBFOX1 deletion. In families 1 and 2, exonic RBFOX1 deletions were detected in all investigated family members with IGE. In contrast, the RBFOX1 deletion identified in the IGE index patient did not cosegregate in family 3. The phenotypic features of the IGE syndromes of the seven affected exonic deletion carriers did not differ from those IGE patients lacking a RBFOX1 deletion. Notably, six of seven IGE patients with RBFOX1 exon-disrupting deletions exhibited typical absence seizures. Comorbidity with other neuropsychiatric disorders was observed in families 1 and 3. In family 1, the index patient EG0369 was affected by a classical CAE but also exhibited neurodevelopmental problems with delayed speech and attention and memory problems resulting in a learning disability that required special education. Learning disability also occurred in the IGE-affected siblings of family 3 (D07u680 and D06519). However, only D07u680 carried a RBFOX1 deletion, whereas the RBFOX1 deletion was missing in sibling D06519, who was affected by IGE since the age of 15 years but also had pervasive developmental disorder, which is part of the diagnostic group of autism spectrum disorders. Moreover, vision impairment due to a strong myopia was present in the IGE-affected mother and all three siblings in family 3. Comorbidity with neuropsychiatric disorders was not reported in families 2 and 4. Magnetic resonance imaging scans of three IGE patients carrying an exonic RBFOX1 deletion (family 1, EG0340; family 2, EPI613; family 4, EG0395) did not reveal any structural abnormalities of the brain, other than for bifrontal lesions in patient EG0340 due to a traumatic brain contusion occurring 16 years after the onset of the IGE.
The present candidate gene CNV study revealed a significant excess of intronic and exonic deletions affecting the neuron-specific RBFOX1 gene in patients with IGE compared with population controls. Specifically, we found RBFOX1 exon-disrupting deletions in 5 (0.35%) of 1,408 IGE patients, whereas none was detected in 2,256 controls. Considering the key role of the splicing regulator RBFOX1 in the control of neuronal excitation and seizure susceptibility (Gehman et al., 2011), the present findings suggest that rare microdeletions affecting the RBFOX1 gene increase the risk of common IGE syndromes.
The four exonic RBFOX1 deletions tested for familial segregation were all inherited. They differed considerably in size, ranging from 68 to 896 kb, and were all located in the 5′-terminal RBFOX1 region encompassing the untranslated exons 1–4 (Fig. S1). The RBFOX1 5′-terminal exons represent highly conserved genomic sequences (Fig. S3) and are predominantly expressed in brain, suggesting that the 5′-terminal RBFOX1 region contains important regulatory elements (Damianov & Black, 2010). Consistent with our findings, the structural genomic variations of the RBFOX1 gene reported previously in three single patients with neurodevelopmental disorders and epilepsy also disrupted the 5′-terminal RBFOX1 exons (Bhalla et al., 2004; Martin et al., 2007; Gallant et al., 2011). Moreover, a female with autism carrying a deletion of RBFOX1 exon 1 due to a de novo translocation t(15p;16p) displayed a significantly reduced RBFOX1 mRNA expression in lymphocytes (Martin et al., 2007). Accordingly, a similar reduction in RBFOX1 mRNA expression can be expected in the members of IGE-multiplex family 1 carrying the large 896 kb microdeletion that deletes the RBFOX1 exons 1–2. IGE-multiplex family 1 is of particular interest because of the consistent cosegregation of the IGE trait with the RBFOX1 deletion (Fig. 2). Notably, none of the previously identified IGE-associated microdeletions at 15q11.2, 15q13.3, and 16p13.11 (Helbig et al., 2009; de Kovel et al., 2010; Heinzen et al., 2010; for review see: Hochstenbach et al., 2011) was found in the IGE index patients carrying a RBFOX1 deletion.
The potential functional alterations of the four smaller deletions involving the RBFOX1 5′-terminal exons 2–4 and exons 1B and 4 remain elusive (Figs 1 and S1). In particular, IGE-multiplex family 3 does not show a co-segregation of the IGE-trait with the 163 kb spanning deletion affecting exon 4. Similar heterogeneous cosegregation patterns, incomplete penetrance, and variable phenotypic expressivity have been observed for the recurrent 15q13.3 microdeletion, representing the strongest genetic risk factor for IGE (OR 68; 95% CI 29–181) identified so far (Dibbens et al., 2009; Helbig et al., 2009; de Kovel et al., 2010; Mefford et al., 2011; Mulley et al., 2011). Together these lines of evidence support an oligogenic/polygenic heterogeneity model for the genetic architecture of the majority of common IGE syndromes and other common neurodevelopmental disorders. Accordingly, the effect of each genetic risk factor alone is not sufficient to express IGE phenotypes, but the interactive effects of heterogeneous sets of rare and low frequency susceptibility factors together promote ictogenesis and epileptogenesis (Dibbens et al., 2007). Depending on the heterogeneous composition of genetic risk factors, the phenotypic expression of exonic RBFOX1 microdeletions is likely to exhibit extensive phenotypic variability as observed for the large recurrent microdeletions at 1q21.1, 15q11.2, 15q13.3, 16p11.2, 16p13.11, and 22q11.2 (Coe et al., 2012). These pathogenic microdeletions seem to affect normal neurodevelopment, resulting in an impaired homeostatic regulation of neuronal networks. In combination with other genetic susceptibility factors, a set of genomic structural variations may contribute to the genetic variability and phenotypic overlap of a wide spectrum of common neuropsychiatric diseases sharing a neurodevelopmental pathogenesis (Coe et al., 2012). With regard to the pivotal role of RBFOX1 in regulating both splicing and transcriptional networks in human neurodevelopmental processes (Fogel et al., 2012), the highly variable spatiotemporal expression of the RBFOX1 gene in differentiating human neurons (Gehman et al., 2011; Fogel et al., 2012; Lin et al., 2012) and the variable expressivity of the large number of downstream gene transcripts may also contribute to the pleiotropic effects of exon-disrupting RBFOX1 deletions. In line with the oligogenic/polygenic heterogeneity model, we observed a familial comorbidity with other neurodevelopmental disorders, such as learning disability and autism spectrum disorder, in two families (1 and 3) with RBFOX1 exon-disrupting deletions (Table 1, Fig. 2). Taking into account that the ascertainment scheme for subjects with IGE applied in this study leads to an exclusion of individuals with severe intellectual disability or predominant neuropsychiatric disorders, comorbidity of generalized seizures with other neurodevelopmental disorders should be more common.
In summary, the present candidate gene CNV study of the neuron-specific splicing regulator gene RBFOX1 suggests that microdeletions affecting the untranslated 5′-terminal RBFOX1 exons increase risk of common IGE syndromes. The present findings warrant further studies to replicate an involvement of RBFOX1 in the genetic predisposition of IGE syndromes and other common neurodevelopmental disorders and to elucidate the pathogenic mechanisms of epileptogenesis resulting from RBFOX1-mediated alterations of the splicing process of neuronal genes.
We thank all participants and their families for participating in this study. We gratefully acknowledge the excellent technical assistance of Sebastian Fey, Christian Becker, Ingelore Bäßmann, and Kerstin Wodecki in array processing and TaqMan qPCR analysis. The Dutch NIGO program (Dutch IGE Genetics Research) thanks all participants and contributing neurologists. This work was supported by grants from the European Community [FP6 Integrated Project EPICURE, grant LSHM-CT-2006-037315 to B.P.C.K., H.L., and T.S.], the German Research Foundation (DFG) within the EUROCORES Programme EuroEPINOMICS [grants NE416/5-1 to B.N., Le1030/11-1 to H.L., SA434/5-1 to T.S.]; the German Federal Ministry of Education and Research, National Genome Research Network [NGFNplus: EMINet, grants 01GS08120 to T.S., and 01GS08123 to H.L.; IntenC, TUR 09/I10 to T.S.]; The Netherlands National Epilepsy Fund [grant 04-08 to B.P.C.K.]; The Netherlands Organization for Scientific Research [grant 917.66.315 to B.P.C.K. and C.G.F.d.K.]; the PopGen biobank [grant to A.F.]. The PopGen project received infrastructure support through the German Research Foundation excellence cluster “Inflammation at Interfaces.” The KORA research platform (KORA, Cooperative Research in the Region of Augsburg) was initiated and financed by the Helmholtz Zentrum München – German Research Center for Environmental Health, which is funded by the German Federal Ministry of Education and Research and by the State of Bavaria; this research was supported within the Munich Center of Health Sciences (MC Health) as part of LMUinnovativ.
None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.
EMINet (Epilepsy and Migraine Integrated Network) Consortium (participating centers): Cologne Center for Genomics, University of Cologne, Weyertal 115b, 50931 Cologne, Germany (Costin Leu, Holger Trucks, Peter Nürnberg, Thomas Sander). Department of Epileptology, University Clinics Bonn, Sigmund Freud Strasse 1, 53105 Bonn, Germany (Christian E. Elger, Ailing A. Kleefuß-Lie, Wolfram S. Kunz, Rainer Surges). Abteilung Neurologie mit Schwerpunkt Epileptologie, Hertie Institut für klinische Hirnforschung, University of Tübingen, Hoppe-Seyler-Str. 3, 72076 Tübingen, Germany (Yvonne Weber, Felicitas Becker, Holger Lerche).
EPICURE Integrated Project (contributing centers listed by country): Department of Clinical Neurology (Fritz Zimprich) and Department of Pediatrics and Neonatology (Martina Mörzinger, Martha Feucht), Medical University of Vienna, Währinger Gürtel 18-20, A-1090 Vienna, Austria. VIB Department of Molecular Genetics, University of Antwerp, Universiteitsplein 1, 2610 Antwerpen, Belgium (Arvid Suls, Sarah Weckhuysen, Lieve Claes, Liesbet Deprez, Katrien Smets, Tine Van Dyck, Tine Deconinck, Peter De Jonghe). Research Department, Danish Epilepsy Centre, Dianalund Kolonivej 7, 4293 Dianalund, Denmark (Rikke S Møller, Laura L. Klitten, Helle Hjalgrim). Wilhelm Johannsen Centre for Functional Genome Research, University of Copenhagen, Blegdamsvej 3, 2200 Copenhagen N, Denmark (Rikke S Møller). Department of Neuropediatrics, University Medical Center Schleswig-Holstein (Kiel Campus), Schwanenweg 20, 24105 Kiel, Germany (Ingo Helbig, Hiltrud Muhle, Philipp Ostertag, Sarah von Spiczak, Ulrich Stephani). Cologne Center for Genomics, University of Cologne, Weyertal 115b, 50931 Cologne, Germany (Costin Leu, Peter Nürnberg, Thomas Sander, Holger Trucks). Department of Epileptology, University Clinics Bonn, Sigmund Freud Strasse 1, 53105 Bonn, Germany (Christian E. Elger, Ailing A. Kleefuß-Lie, Wolfram S. Kunz, Rainer Surges). Department of Neurology, Charité University Medicine, Campus Virchow Clinic, Humboldt University of Berlin, Augustenburger Platz 1, 13353 Berlin, Germany (Verena Gaus, Dieter Janz, Thomas Sander, Bettina Schmitz). Epilepsy-Center Hessen, Department of Neurology, Philipps-University Marburg, 35043 Marburg, Germany (Karl Martin Klein, Philipp S. Reif, Wolfgang H. Oertel, Hajo M. Hamer, Felix Rosenow). Abteilung Neurologie mit Schwerpunkt Epileptologie, Hertie Institut für klinische Hirnforschung, Universität Tübingen, Hoppe-Seyler-Str. 3, 72076 Tübingen, Germany (Felicitas Becker, Yvonne Weber, Holger Lerche). Netherlands Section Complex Genetics, Department of Medical Genetics, University Medical Center Utrecht, Str. 2.112 Universiteitsweg 100, 3584 CG Utrecht, The Netherlands (Bobby P.C. Koeleman, Carolien de Kovel, Dick Lindhout). SEIN Epilepsy Institute in the Netherlands, P.O. Box 540, 2130AM Hoofddorp, Netherlands (Dick Lindhout).