Role of SOX2 Mutations in Human Hippocampal ... - Wiley Online Library

12 downloads 1856 Views 393KB Size Report
with data from 45 female controls (14). ... in a cohort of 102 patients and 88 controls: the only com- ..... the region of SOX genes to unravel the complex control.
Epilepsia, 47(3):534–542, 2006 Blackwell Publishing, Inc.  C 2006 International League Against Epilepsy

Role of SOX2 Mutations in Human Hippocampal Malformations and Epilepsy ∗ Sanjay M. Sisodiya, †‡Nicola K. Ragge, §Gianpiero L. Cavalleri, Ann Hever, ¶Birgit Lorenz, ∗∗ Adele Schneider, Kathleen A. Williamson, ∗ John M. Stevens, ∗ Samantha L. Free, ∗ Pamela J. Thompson, Veronica van Heyningen, and David R. FitzPatrick ∗ Department of Clinical and Experimental Epilepsy, Institute of Neurology, UCL, London, and National Society for Epilepsy, Bucks SL90RJ †Adnexal Service, Moorfields Eye Hospital, London, ‡Department of Human Anatomy and Genetics, University of Oxford, Oxford, and §Department of Biology, UCL, London, United Kingdom; MRC Human Genetics Unit, Edinburgh, United Kingdom; ¶ Department of Paediatric Ophthalmology and Ophthalmogenetics, Klinikum, University of Regensburg, Germany; and ∗∗ Department of Clinical Genetics, Albert Einstein Medical Center, Philadelphia, Pennsylvania, U.S.A.

Summary: Purpose: Seizures are noted in a significant proportion of cases of de novo, heterozygous, loss-of-function mutations in SOX2, ascertained because of severe bilateral eye malformations. We wished to determine the underlying cerebral phenotype in SOX2 mutation and to test the candidacy of SOX2 as a gene contributing to human epilepsies. Methods: We examined high-resolution MRI scans in four patients with SOX2 mutations, two of whom had seizures. We determined the Sox2 expression pattern in developing murine brain. We searched for SOX2 mutation in 24 patients with typical hippocampal sclerosis and for common variations in SOX2 in 655 patients without eye disease but with epilepsy, including 91 patients with febrile seizures, 93 with hippocampal sclerosis, and 258 with temporal lobe epilepsy. Results: Striking hippocampal and parahippocampal malformations were seen in all cases, with a history of febrile seizures

or epilepsy in two of four cases. The Sox2 expression pattern in developing mouse brain supports the pattern of malformations observed. Mutation screening in patients with epilepsy did not reveal any abnormalities in SOX2. No associations were found between any clinical epilepsy phenotype and common variation in SOX2. Conclusions: SOX2 haploinsufficiency causes mesial temporal malformation in humans, making SOX2 dysfunction a candidate mechanism for mesial temporal abnormalities associated with chronic epilepsy. However, although mutation of SOX2 in humans causes hippocampal malformation, SOX2 mutation or variation is unlikely to contribute commonly to mesial temporal lobe epilepsy or its structural (hippocampal sclerosis) or historic (febrile seizures) associations in humans. Key Words: Brain malformation—Epilepsy—SOX2—Hippocampal sclerosis.

Brain malformations are an important cause of chronic partial epilepsy (1), the commonest form of which is temporal lobe epilepsy (TLE). The most common structural cause of TLE is hippocampal sclerosis (HS). Genetic causes of many seizure disorders, including those caused by brain malformations, are now known (2), but the genetics of HS remains obscure (3). Many associations between common gene variants and TLE or HS have been described (4): none has been reliably replicated (4,5). It has been suggested that HS develops, for whatever reason, on a background of hippocampal malformation (6).

Although rare familial hippocampal malformations have been described, and in some cases associated with temporal lobe epilepsy (7,8), no underlying genetic mutation has been identified. Identification of a genetic basis to HS, whether the genetic basis led to HS through an underlying malformation or not, would be of considerable clinical importance, raising the possibility of intervention. We have developed a paradigm for the identification of candidate genes for brain malformations that may underlie epilepsy. Several fundamental anomalies of early eye development are the result of mutations in major developmental regulator genes that are expressed in both the eye and the brain. The phenotypically obvious eye malformations lead to case ascertainment but also can signal unsuspected brain malformations. For PAX6, we have shown this to be the case (9). Mutations in this gene cause not only aniridia, but also brain malformations, including epileptogenic polymicrogyria (10,11). PAX6 has thus

Accepted October 16, 2005. Address correspondence and reprint requests to Dr. S. Sisodiya at Department of Clinical and Experimental Epilepsy, Institute of Neurology, UCL, Queen Square, London WC1N 3BG, U.K. E-mail: [email protected] Current address for G.L. Cavalleri, The Institute for Genome Sciences and Policy, Center for Genomics and Pharmacogenetics, Duke University, Durham, North Carolina, U.S.A.

534

SOX2 BRAIN MALFORMATION become a candidate gene for human polymicrogyria, even in the absence of overt eye anomalies. We recently reported de novo, heterozygous, nonsense mutations in SOX2 in four individuals and one case with a deletion of the whole gene, all associated with anophthalmia (absent eye) or microphthalmia (small eye); additional significant neurologic involvement was evident in three cases (12). The ocular and extraneural phenotypes have been expanded in a larger cohort: MRI findings were not given in detail or illustrated (13). SOX2 is a highly conserved, developmentally regulated transcription factor that shows site- and stage-specific expression in the developing brain and eye of all vertebrate species studied. Therefore after our previous observation of developmental brain malformations by using MRI in a cohort of cases with PAX6 mutations (9,10,14), we gathered MRI data in four cases with SOX2 haploinsufficiency. Abnormalities of the mesial temporal region, which in general are often associated with epilepsy, were observed in the scans from all four cases. The expression pattern of Sox2 in the developing mouse brain was studied, by using immunohistochemistry, to assess consistency with the observed abnormalities. The association of SOX2 mutation with brain malformation and epilepsy in anophthalmia cases suggests that SOX2 may be a candidate gene for more common forms of epilepsy: we went on to test this hypothesis. MATERIALS AND METHODS This study was approved by local ethics committees (National Hospital for Neurology and Neurosurgery/ Institute of Neurology, Moorfields Eye Hospital, University of Regensburg).

535

Subjects studied with MRI The main clinical features of each case are summarized in Table 1. All cases were female. Additional characteristics in these individuals were as follows: Case 1, with mild learning disability requiring significant educational support within a mainstream school, has a 600kilobase deletion adjacent to a translocation breakpoint and encompassing the entire SOX2 gene (12). Case 2 has a nonsense SOX2 mutation (p.S83X); no history of other early insult was noted, such as prolonged febrile convulsion. She has a mild spastic diplegia. Case 2 had epilepsy since onset of habitual seizures at 5 years, she has continued to have stereotyped complex partial seizures without aura, characterized by arrest of activity, brief vacancy, and then rapid recovery. She has two seizures a year on treatment with a single antiepileptic drug (AED). Case 3, with mild learning disability requiring special educational support, has a two-base-pair (bp) deletion causing a frameshift in the 3’ region of the open reading frame (c.943del2). Case 4, with global developmental delay, has a 23-bp deletion causing a frameshift in the 5’ coding region of the mRNA (c.67del23). Case 4 had a seizure at age 2.9 years, consisting of right hemifacial and right arm twitching, followed by a transient right hemiparesis. Her seizures were controlled on valproate. Further details on genotype are given in Fantes et al. (12) and on nonbrain phenotype in Ragge et al. (13). Of five additional known cases with SOX2 mutations, four (three male, one female patient) have had febrile convulsions (13). These five additional cases either had no MRI or did not have an MRI of sufficient resolution to incorporate into this study.

TABLE 1. Summary of MRI and clinical findings MRI findings Hippocampal malformation Temporal periventricular heterotopia Fusiform gyrus malformation Parahippocampal malformation Elongation of anterior hypothalamus Malformation of gyrus rectus Subcallosal gyrus malformation Rostral sulcus malformation Optic nerves Chiasm Chiasmatic notch and recess Clinical phenotype Age at last assessment Prenatal growth failure Postnatal growth failure Eye phenotype Left eye Right eye Vision Neurologic phenotype Learning disability Motor coordination problems Seizures

Case 1

Case 2

Case 3

Case 4

Bilateral Absent Bilateral Bilateral Bilateral Absent Absent Absent Normal Normal Normal

Bilateral Bilateral Bilateral Bilateral Bilateral Bilateral Absent Absent Normal Normal Normal

Bilateral Absent Bilateral Bilateral Bilateral Bilateral Absent Absent Attenuated Not visible Normal

Bilateral Bilateral Bilateral Bilateral Bilateral Right-sided Right-sided Right-sided Attenuated Not visible Not visible

6 yr No Yes

18 yr No No

7 yr No Yes

8 yr No Yes

AN AN None

MI+PPM+CA AN Near normal

AN AN None

AN AN None

Yes Yes No

Yes Yes Yes

Yes Yes No

Yes Yes Yes

AN, anophthalmia; MI, microphthalmia; CA, cataract; PPM, posterior pupillary membrane; motor coordination problems, limb and orofacial abnormalities reported by referring physicians. Epilepsia, Vol. 47, No. 3, 2006

536

S. SISODIYA ET AL.

MRI High-resolution MRI images were available or acquired in four cases. Two subjects (cases 3 and 4) were scanned on one scanner, two others (cases 1 and 2) on two other scanners. Data were reformatted in multiple planes to allow careful examination of regions of interest. Additionally, for case 2, signal changes were inspected on T2 and FastFLAIR sequences (T2 and PD sequence: TE1, 30; TE2, 120; TR, 2,000; NEX, 1; acquisition matrix, 256 × 128; field of view, 24 × 18 cm; slice thickness. 5 mm contiguous; FastFLAIR sequence: TE1, 152; TE2, 2,200; TR, 10,002; NEX, 1; acquisition matrix, 256 × 128; field of view, 24 cm; slice thickness, 5 mm contiguous). For case 2, cerebral volume and corpus callosal cross-sectional area were measured and compared with data from 45 female controls (14). Hippocampal volume and hippocampal T2 values were measured and compared with control data from our routine clinical epilepsy practice. Case 2 also had standardized neuropsychological assessment including tests of verbal and visual memory. No EEG studies were undertaken. Sox2 expression studies in the mouse Mouse embryos were harvested from timed matings, at E11.5, E12.5, E14.5, and E15.5, and fixed overnight in 4% paraformaldehyde at 4◦ C. Embryos were then processed for immunohistochemistry. Fixed embryos were dehydrated through a graded ethanol series to 70% for immunohistochemistry. Embryos were embedded in paraffin wax and cut into 6-μm serial sections. The sections were dewaxed in xylene and rehydrated through a graded ethanol series. Pretreatment was by boiling in 10 mM citrate buffer, pH 6.0, for 2 min. After cooling for 20 min at room temperature (RT), sections were washed several times in phosphate-buffered saline (PBS) and rinsed in milliQH2 O. All subsequent steps were performed in a humidity chamber. Nonspecific binding was reduced by blocking in 10% inactivated sheep serum (Sigma) in PBS for 1 h at RT. Primary antibody, rabbit anti-Sox2 (Chemicon, Temecula, CA, U.S.A.) was diluted to 1:500 in 10% inactivated sheep serum and applied to the sections overnight at 4◦ C. The sections were washed twice in PBS and once in PBS containing 0.1% Tween 20 for 5 min. Biotinylated anti-rabbit immunoglobulin G (IgG; Vector Laboratories) was diluted 1:500 in 10% inactivated sheep serum and applied to all sections for 1 h at RT. The sections were washed as previously described and then incubated in alkaline phosphatase–conjugated streptavidin (Vector Laboratories) for 1 h at RT. Staining was visualized with BCIP/NBT (Vector Laboratories, Burlingame, CA, U.S.A.) containing levamisole to reduce endogenous alkaline phosphatase. After rinsing in PBS, sections were counterstained in eosin and mounted in Histomount (Raymond A. Lamb Ltd., London, U.K.). Epilepsia, Vol. 47, No. 3, 2006

SOX2 candidate gene analysis in large cohort of epilepsy cases From an ongoing effort examining aspects of the genetics of epilepsy, we studied a cohort of 655 epilepsy patients. All subjects are of European origin. To allow a case–control examination of common variation in SOX2, we used 364 unrelated individuals from the British twin registry as controls (15). We examined idiopathic generalized (IGE, n = 96), symptomatic focal (n = 263), and cryptogenic focal (n = 208) epilepsy subgroups, corresponding approximately, although not exactly, to the to the International League Against Epilepsy classification. Eighty-eight individuals were unclassifiable according to these criteria. No cases of HS were familial. Forty-nine of the 655 individuals had at least one first-degree relative with a diagnosis of epilepsy. Subgroups of patients with TLE (n = 258) or febrile seizures (FSs; n = 91) were identified because of the known associations between FS, HS, and temporal lobe epilepsy (16). HS was diagnosed after high-resolution MRI. We characterized the SOX2 region for common variation. Fantes et al. (12) screened the coding region of the single-exon SOX2 gene in a cohort of 102 patients and 88 controls: the only common variation they detected was G976A in the gene 3’ UTR. To characterize all common variation in the SOX2 gene, we extended the screen to the full 500 bp of the 5’ UTR and 1,100 bp of the 3’ UTR, by resequencing using published and novel primers and conditions (12) (see Fig. 1 for layout and details of amplicons). We screened all amplicons in a panel of 24 HS patients and 22 patients with a history of FSs (from the cohort of 655). All details are available on request. We assessed the significance of genotypic and allelic contingency tables for the variants identified by using the χ 2 distribution, and for tables with insufficient cell counts, an exact probability test as implemented in the program RxC (from http://bioweb.usu. edu/mpmbio/rxc.asp). Should these variants be functional,

FIG. 1. Layout of human SOX2 gene showing amplicons screened and variation detected. ∗ Position of the variants detected.

SOX2 BRAIN MALFORMATION we estimated the minimal effect they would need to have for our cohort to have 75% power of detection, by using software available at http://statgen.iop.kcl.ac. uk/gpc/. We used as parameter values a disease prevalence of 0.1%, risk allele frequencies observed in our control population, and the relevant case-to-control ratio. We then increased the relative risk (under an additive model of inheritance) until 75% power of detection was reached.

537 RESULTS

MRI findings Novel mesial temporal abnormalities were noted in all subjects, whether they had seizures or not. The hippocampi were small and malformed. In comparison to normal subjects (Fig. 2A–D), hippocampi were abnormally orientated, foreshortened, and misshapen (Fig. 2E–G). The hippocampi were medially displaced and encroached

FIG. 2. Mesial temporal abnormalities in a patient with SOX2 haploinsufficiency. Coronal T1 -weighted (MRI acquisition parameters were TE, 4.2; TI, 450; TR, 15; NEX, 1; flip angle, 20; acquisition matrix, 256 × 128; field of view, 24 cm; producing 124 contiguous slices; voxel dimension, 0.9375 × 0.9375 × 1.5 mm) slices from posterior to anterior from a normal subject (A–D) showing normal position, shape, orientation, and size of the hippocampi (A–C, solid-headed arrows) and juxtahippocampal structures. Slices are at approximately the same rostral position as (E–H). Normal amygdala are also shown (D, open-headed arrows). Coronal T1 -weighted images from case 2 are shown in (E–H), illustrating developmental abnormality of the mesial temporal structures: it is apparent that the hippocampi themselves are malformed and malpositioned, and that mesial temporal structures 3–7 seen clearly in the normal subject (cf. Fig. 3) also are malformed and sometimes difficult to discern. From the most posterior image containing the hippocampi (E, solid-headed arrows), the abnormally medial position of the hippocampi is apparent, as the hippocampi intrude more medially into the ambient cistern; this also leads to the emergence of a fluid-filled space lateral to the hippocampi, an extension of the temporal horn of the lateral ventricle known erroneously in older texts as the “diverticulum of the subiculum.” In comparison with (A), it is possible to see in (E) that the lateral occipitotemporal sulcus also is deformed. In (F, G), successively more rostral images, the abnormal position, orientation, rotation, and size of the hippocampi remain apparent throughout their length (solid-headed arrows). In these more anterior images (F, G), precise delineation of some structures (for example, the subiculum, parahippocampal gyrus, and subiculum) becomes very difficult in comparison to the normal subject in (B, C). The amygdala (H, openheaded arrows) appear normal. Isolated bitemporal periventricular heterotopia is seen in (E–G) (arrowheads).

Epilepsia, Vol. 47, No. 3, 2006

538

S. SISODIYA ET AL. for controls) (14); hippocampal T2 measures were not elevated, making a diagnosis of HS unlikely. Callosal crosssectional area was reduced (3.67 cm2 , >2 SDs below ageand gender-matched control means) (14). The pattern of hippocampal abnormality is quite unlike that seen in HS. The amygdalae were not malformed. Additional brain malformations and neurologic characteristics were noted (Table 1). Two subjects (cases 2 and 4) also had isolated bitemporal periventricular heterotopia, without heterotopia elsewhere (Fig. 2E–G; Fig. 4).

FIG. 3. Higher-magnification images of normal and SOX2 individual MRI. A: The temporal regions from Fig. 2A are magnified, to show the body of the hippocampus (1), the subiculum (2), the ambient cistern (3), the parahippocampal gyrus (4), the collateral sulcus (5), the fusiform gyrus (6), and the lateral occipitotemporal sulcus (7). A magnified version of Fig. 2F is shown in B, highlighting the malformations and also showing how identification of comparable structures in the malformed brain is difficult.

on the ambient cistern. Instead of being tucked laterally into the temporal horns of the lateral ventricles, especially posteriorly, the medial displacement and incomplete rotation of the hippocampi generated a CSF-filled space sometimes referred to in older texts as the “diverticulum of the subiculum” (Fig. 2E). More anteriorly, the hippocampi were almost vertically aligned. The anatomy of adjacent structures was in turn so deformed in some cases as to make reliable definition of the subiculum, parahippocampal gyrus, and adjacent sulci difficult (Figs. 2–4). Brain volumetry, possible in one subject (case 2), showed that cerebral volume was 807 cc (within 2 standard deviations of the mean for age- and gender-matched controls) (14); hippocampal volumes were symmetric but very small (left, 0.90 cc; right, 0.94 cc; >6 SDs below the mean

Cognitive testing of case 2 Psychometry in case 2 showed verbal intellectual impairment (WAIS-R: VIQ = 69). No specific memory deficits were seen. Her immediate recall of a short story was impaired, but she showed good retention after a 30min delay (% recalled, 10th– 25th centile), better than would be predicted from her intellectual level. On a verbal learning task, her score was impaired over learning trials, but she showed clear benefit from the opportunity for rehearsal. On a recognition memory test involving distinctive colored scenes, she made no errors. Her language skills were relatively well developed. Her reading age was 8 years for single words, and her level of language comprehension for single words was 10 years and 2 months. She could name all the high-frequency items on an objectnaming test. Sox2 expression studies in developing mouse Immunohistochemistry (Fig. 5A) confirmed Sox2 expression in virtually all neuroblasts at midgestation (E11.5), but subsequently (E15.5) Sox2 expression becomes more restricted to ventricular zone cells, including the hippocampal anlage (Fig. 5). Assessing variation in the SOX2 gene in a large epilepsy cohort No mutations were found in the entire SOX2 coding and UTR region in any of the 24 HS patients or the 22 patients FIG. 4. Mesial temporal abnormalities are consistent. Similar mesial temporal abnormalities were apparent in other patients with SOX2 haploinsufficiency (cases 1, 3, and 4). Shown is case 4: the hippocampi (solid-headed arrows) were small and malformed. In the more anterior image from case 4 (B), the same range of abnormalities shown in Fig. 2 is apparent, with incomplete rotation of the hippocampi (solid-headed arrows) and deformation of the adjacent structures, leading, for example, to a more horizontal orientation of the collateral sulcus than normal, more apparent on the patient’s right-hand side (left side of image B), and deepening of the lateral occipitotemporal gyrus. Refer again to Fig. 3A for normal anatomy. In case 4, see bitemporal periventricular heterotopia (arrowheads).

Epilepsia, Vol. 47, No. 3, 2006

SOX2 BRAIN MALFORMATION

539

with a history of FSs. We detected only two nonsynonymous variants (single-nucleotide polymorphisms; SNPs): G976A, reported previously (12), seen with a minor allele frequency of 2% (cf 2.8% in ref. 12); the other SNP detected, rs11915160, was located in the 3’ UTR and seen at a minor allele frequency of 14.7% (no prior frequency data available). The rs11915160 variant did not associate with any of the following phenotypes: HS, FS, or any other epilepsy grouping (all epilepsy, TLE, cryptogenic epilepsy, symptomatic epilepsy, idiopathic generalized epilepsy): see Table 2. In addition, the G976A variant did not associate with TLE, HS, or FS (see Table 3). DISCUSSION

FIG. 5. Sox2 expression pattern in developing mouse brain. Sox2 is expressed in the developing mouse embryonic hippocampus. A–C: Immunohistochemistry was performed on coronal sections through the embryonic hippocampus to detect Sox2 protein expression (purple). Eosin (red), which stains cell cytoplasm, was used as a counterstain. Labelling is seen in the developing hippocampus, with a gradual restriction in the expression pattern with development (A–C). Note also the strong expression in the germinal epithelium (C). h, developing hippocampus; lv, lateral ventricle; IIIv, third ventricle; dt, dorsal thalamus; cp, choroid plexus; ge, ganglionic eminence; vlt, ventrolateral thalamus nucleus; har, habenular recess of third ventricle.

We have shown a novel pattern of cerebral malformation in patients with anophthalmia/microphthalmia caused by mutations in SOX2 leading to haploinsufficiency. This gene is known to be vital to eye development (12) and important in neural development (17,18), but its role in human brain development has not been reported previously. The brain abnormalities identified include two rare malformations: hippocampal malformation and temporal periventricular heterotopia without heterotopia around the occipital or trigonal regions of the lateral ventricles. Hippocampal malformation has rarely been reported; apparently sporadic cases have been identified on MRI (19) and unilaterally in a kindred with epilepsy (8). We have previously described bilateral hippocampal malformation at postmortem (20). In humans, no gene defect has yet been associated with hippocampal malformation, although hippocampal abnormalities have been described in association with other malformations and in animal models. The amygdalae were normal in our cases, suggesting very specific subregional patterning of mesial temporal structures. The hippocampi are associated with episodic memory formation, and bilateral abnormalities are usually associated with profound memory deficits (21). However, case 2 had no specific definable memory deficit, despite her very small hippocampal volumes and mild learning disability. As the observed abnormalities were developmental, relative preservation of memory function might be accounted for by brain plasticity. The abnormalities abutting the temporal horns of the lateral ventricles are isointense to gray matter on all sequences, establishing a diagnosis of neuronal heterotopia (22). Overall, bilateral periventricular heterotopia is not an unusual malformation in patients with refractory epilepsy and is occasionally found without a history of epilepsy. However, such malformation involves the occipital horns of the lateral ventricles, possibly with additional involvement of the trigones, frontal or temporal horns of the lateral ventricles (23), but has not previously been reported,

Epilepsia, Vol. 47, No. 3, 2006

540

S. SISODIYA ET AL. TABLE 2. Association between rs11915160 and clinical epilepsy phenotype Group

All patients

TLE

HS

FS

Symp

Cryp

Controls

Number CC CA AA MAF p values Genotype Allele Min RR

655 403 (0.7) 161 (0.28) 14 (0.02) 0.147

258 148 (0.71) 55 (0.27) 4 (0.02) 0.152

93 51 (0.65) 24 (0.31) 3 (0.04) 0.202

91 62 (0.72) 23 (0.27) 1 (0.01) 0.145

263 167 (0.72) 59 (0.26) 5 (0.02) 0.149

208 110 (0.66) 52 (0.31) 5 (0.03) 0.186

384 250 (0.73) 85 (0.25) 8 (0.02) 0.147

0.59 0.35 1.8

0.856 0.152 2.3

0.37 0.16 2.6

0.87 0.82 2.6

0.97 0.92 2.0

0.26 0.12 2.1

TLE, temporal lobe epilepsy; HS, hippocampal sclerosis; FS, history of febrile seizures; Symp, symptomatic epilepsy; Cryp, cryptogenic epilepsy; MAF, minor allele frequency; Min RR, minimal relative risk of homozygous risk genotype required to give 75% power of detection to this study, assuming an additive model of inheritance. Genotype counts are those for successful typings only.

to our knowledge, to be present solely around the temporal horns. In the majority of familial periventricular heterotopia cases, and a minority of sporadic cases, mutations in the filamin gene FLNA have been reported (24). We believe that an FLNA mutation is unlikely to be the cause of heterotopia because (a) both patients with heterotopia had a de novo mutation in another brain-expressed gene, SOX2; (b) the cerebral phenotypes seen are not described in FLNA mutation; (c) epilepsy onset was earlier than that usually seen in heterotopia caused by FLNA mutation; and (d) no relevant family history was present (23,24). Other genetic associations have been described for bilateral periventricular heterotopia (of more classic distribution), for example, with linkage to chromosome 5p (25), and a recently described autosomal recessive form due to mutations in ARFGEF2 (26). SOX2 is the third identified gene associated with periventricular heterotopia and is thus also implicated in human neuronal migration. This is the first direct evidence that SOX2 mutations can affect human cerebral development and neurologic function. No obvious phenotypic differences are observed between the complete gene deletion leading to haploinsufficiency in case 1 and the predicted premature protein terminations, in cases 2–4. This report reiterates the potential value of a candidate gene–MRI approach to the study of cerebral malformation associated with epilepsy (9,10). Such an approach is one of the few ways in which the genetic control of the fate map of the neural plate can be determined in humans. Heterozygous loss of Sox2 in the mouse does not give rise to any notable brain or eye phenotypes (27). Thus in this case, the mouse is seemingly a poor model for the human disease. However, the gene-expression pattern during normal mouse development (our results and refs. 28 and 29) is highly consistent with the observed phenotypic spectrum seen in the human haploinsufficiency cases. A more pertinent mouse model has recently been described (29). Compound heterozygotes with one copy of Sox2 knocked out and the second copy deleted for the previously

Epilepsia, Vol. 47, No. 3, 2006

defined neural cell–specific enhancer (18,27) and expressing Sox2 at reduced level show significant cerebral malformation (29). Parenchymal loss, ventricular enlargement, L-dopa–rescuable circling behavior, and epilepsy are observed. Neuronal ependymal cell abnormalities are seen, together with decreased adult neurogenic proliferation and reduction of glial fibrillary acidic protein (GFAP)/nestinpositive hippocampal cells (29). Given these findings, the further population genetic component of our study was undertaken to establish whether mutation or common variation in SOX2 might also underlie epilepsy (irrespective of syndrome), FS, or HS, in patients without severe eye anomalies (i.e., to test the candidacy of SOX2 variation in human epilepsy per se). The phenotypic overlap between patients with definite SOX2 mutation and patients with epilepsy is limited to an anatomic area of interest (the mesial temporal lobe), but it is worth reiterating that currently no genes for human hippocampal development or maldevelopment are known, so making SOX2 a better candidate than any others at this time. All the mutations associated with anophthalmia are protein-truncating mutations or missense changes in the TABLE 3. Association between G976A and various clinical phenotypes Group Number GG GA AA MAF p values Genotype Allele Min RR

TLE

HS

FS

Controls

258 162 (0.99) 3 (0.01) 0 0.009

93 68 (0.99) 1 (0.01) 0 0.007

91 62 (0.97) 4 (0.03) 0 0.031

192 128 (0.97) 7 (0.03) 0 0.026

0.118 0.121 5.2

0.274 0.279 6

0.746 0.756 6

TLE, temporal lobe epilepsy; HS, hippocampal sclerosis; FS, history of febrile seizures; MAF, minor allele frequency; Min RR, minimal relative risk of homozygous risk genotype required to give 75% power of detection to this study, assuming an additive model of inheritance. Genotype counts are those for successful typings only.

SOX2 BRAIN MALFORMATION DNA-binding domain and thus predicted to lead to loss of function. Milder missense changes might lead solely to the brain phenotype. However, we were unable to identify SOX2 mutation in any of the screened cases. It is therefore unlikely that in this cohort as a whole, subtle SOX2 mutation can be a frequent cause of either HS or FS, both much more common conditions than anophthalmia and microphthalmia. Although our methods did not allow deletion of the SOX2 gene to be identified, we believe it unlikely that any of the epilepsy cohort could have harbored a deletion, as this would have caused obvious eye malformation (12). In addition, we noted that nonmutational variation (SNPs) in the SOX2 gene was rare in our epilepsy cohort. Both G976A and rs11915160 are in the 3’ UTR and would therefore not have any effect on the amino acid sequence of the encoded SOX2 protein. Neither SNP was associated with the phenotype of HS or any other epilepsy subgrouping studied, including IGEs, examined for association because thalamic neuronal involvement is implicated in such epilepsies, and Sox2 expression is noted in mature thalamic neurons in adult mice (29). Our study provided 75% power for detection of association for a modest to high relative risk for these SNPs. The SOX genes tend to be small (1–5 kb) and well conserved across species. However, it is becoming increasingly apparent that functional variation affecting SOX gene expression may be, for the most part, in gene regulatory regions that stretch over tens of kilobases on either side of the gene. The expression of the SOX gene family is highly influenced by tissue-specific regulatory signals, as would be expected of complex transcriptional regulators. Recent studies have begun to examine enhancer sites in the region of SOX genes to unravel the complex control of expression of these genes (18,30). It is possible that functional variation residing in these regions could play a role in the development of epilepsy. Overall, common genetic variation within SOX2 is unlikely to contribute to the HS or FS phenotype, although we cannot exclude small effects. It is also very unlikely that subtle SOX2 mutations, causing an isolated brain phenotype, frequently underlie the relatively common conditions of HS or FS. We cannot exclude as a cause of FS or HS rare, private SOX2 mutations or mutations in its long-range control elements (31). The rare hippocampal malformation caused by SOX2 mutation associated with anophthalmia/microphthalmia is, however, unlikely to be the putative maldevelopmental template underlying most cases of HS that may have a developmental origin (6). We have not studied the possible candidacy of SOX2 variation in causing periventricular heterotopia. Acknowledgment: We thank the families for their help and support. We thank Prof David Goldstein for genetic advice and Dr. T. Spector for the control samples. N.K.R. is a Senior Surgical Scientist supported by the Academy of Medical Sciences/The Health Foundation. S.L.F. is supported by an MRC Cooperative

541

Group Grant. G.L.C. was supported by The Annals of Human Genetics Studentship in Human Population Genetics. The work was supported by MRC grant G69520.

REFERENCES 1.

2. 3. 4. 5.

6. 7. 8. 9. 10. 11.

12. 13. 14. 15. 16. 17. 18.

19.

20. 21. 22.

Li LM, Fish DR, Sisodiya SM, et al. High resolution magnetic resonance imaging in adults with partial or secondary generalised epilepsy attending a tertiary referral unit. J Neurol Neurosurg Psychiatry 1995;59:384–387. Mochida GH, Walsh CA. Genetic basis of developmental malformations of the cerebral cortex. Arch Neurol 2004;61:637–640. Berkovic SF, Jackson GD. The hippocampal sclerosis whodunit: enter the genes. Ann Neurol 2000;47:557–558. Tan NC, Mulley JC, Berkovic SF. Genetic association studies in epilepsy: “the truth is out there.” Epilepsia 2004;45:1429– 1442. Cavalleri GL, Depondt C, Burley M-W, et al. Failure to replicate any reported genetic associations with sporadic temporal lobe epilepsy: lessons for the study of complex traits. Brain 2005;128:1832–1840. Blumcke I, Thom M, Wiestler OD. Ammon’s horn sclerosis: a maldevelopmental disorder associated with temporal lobe epilepsy. Brain Pathol 2002;12:199–211. VanLandingham KE, Heinz ER, Cavazos JE, et al. Magnetic resonance imaging evidence of hippocampal injury after prolonged focal febrile convulsions. Ann Neurol 1998;43:413–426. Fernandez G, Effenberger O, Vinz B, et al. Hippocampal malformation as a cause of familial febrile convulsions and subsequent hippocampal sclerosis. Neurology 1998;50:909–917. Sisodiya SM, Free SL, Williamson KA, et al. PAX6 haploinsufficiency causes cerebral malformation and olfactory dysfunction in humans. Nat Genet 2001;28:214–216. Mitchell TN, Free SL, Williamson KA, et al. Polymicrogyria and absence of pineal gland due to PAX6 mutation. Ann Neurol 2003;53:658–663. Schmahl W, Knoedlseder M, Favor J, et al. Defects of neuronal migration and the pathogenesis of cortical malformations are associated with Small eye (Sey) in the mouse, a point mutation at the Pax-6-locus. Acta Neuropathol (Berl) 1993;86:126–135. Fantes J, Ragge NK, Lynch SA, et al. Mutations in SOX2 cause anophthalmia. Nat Genet 2003;33:461–463. Ragge NK, Lorenz B, Schneider A, et al. SOX2 anophthalmia syndrome. Am J Med Genet A 2005;135:1–7. Free SL, Mitchell TN, Williamson KA, et al. Quantitative MR image analysis in subjects with defects in the PAX6 gene. Neuroimage 2003;20:2281–2290. Andrew T, Hart DJ, Snieder H, et al. Are twins and singletons comparable? A study of disease-related and lifestyle characteristics in adult women. Twin Res 2001;4:464–477. Baulac S, Gourfinkel-An I, Nabbout R, et al. Fever, genes and epilepsy. Lancet Neurol 2004;3:421–430. Kishi M, Mizuseki K, Sasai N, et al. Requirement of Sox2-mediated signaling for differentiation of early Xenopus neuroectoderm. Development 2000;127:791–800. Zappone MV, Galli R, Catena R, et al. Sox2 regulatory sequences direct expression of a (beta)-geo transgene to telencephalic neural stem cells and precursors of the mouse embryo, revealing regionalization of gene expression in CNS stem cells. Development 2000;127:2367–2382. Baulac M, De Grissac N, Hasboun D, et al. Hippocampal developmental changes in patients with partial epilepsy: magnetic resonance imaging and clinical aspects. Ann Neurol 1998;44:223– 233. Thom M, Sisodiya SM, Lin WR, et al. Bilateral isolated hippocampal malformation in temporal lobe epilepsy. Neurology 2002;58:1683–1686. Spiers HJ, Maguire EA, Burgess N. Hippocampal amnesia. Neurocase 2001;7:357–382. Barkovich AJ, Kuziecky RI. Gray matter heterotopia. Neurology 2000;55:1603–1608.

Epilepsia, Vol. 47, No. 3, 2006

542 23. 24.

25. 26.

27.

S. SISODIYA ET AL. Poussaint TY, Fox JW, Dobyns WB, et al. Periventricular nodular heterotopia in patients with filamin-1 gene mutations: neuroimaging findings. Pediatr Radiol 2000;30:748–755. Sheen VL, Dixon PH, Fox JW, et al. Mutations in the Xlinked filamin 1 gene cause periventricular nodular heterotopia in males as well as in females. Hum Mol Genet 2001;10:1775– 1783. Sheen VL, Wheless JW, Bodell A, et al. Periventricular heterotopia associated with chromosome 5p anomalies. Neurology 2003;60:1033–1036. Sheen VL, Ganesh VS, Topcu M, et al. Mutations in ARFGEF2 implicate vesicle trafficking in neural progenitor proliferation and migration in the human cerebral cortex. Nat Genet 2004;36: 69–76. Avilion AA, Nicolis SK, Pevny LH, et al. Multipotent cell lineages

Epilepsia, Vol. 47, No. 3, 2006

28. 29. 30.

31.

in early mouse development depend on SOX2 function. Genes Dev 2003;17:126–140. Gray PA, Fu H, Luo P, et al. Mouse brain organization revealed through direct genome-scale TF expression analysis. Science 2004;306:2255–2257. Ferri AL, Cavallaro M, Braida D, et al. Sox2 deficiency causes neurodegeneration and impaired neurogenesis in the adult mouse brain. Development 2004;131:3805–3819. Uchikawa M, Takemoto T, Kamachi Y, et al. Efficient identification of regulatory sequences in the chicken genome by a powerful combination of embryo electroporation and genome comparison. Mech Dev 2004;121:1145–1158. Kleinjan DA, van Heyningen V. Long-range control of gene expression: emerging mechanisms and disruption in disease. Am J Hum Genet 2005;76:8–32