JLE

Epileptic Disorders

MENU

Detection of microchromosomal aberrations in refractory epilepsy: a pilot study Volume 12, issue 3, September 2010

Figures

See all figures

Auteur(s) : Jacinta M McMahon1, Ingrid E Scheffer1,2, Jillian K Nicholl3, Wendy Waters3, Helen Eyre3, Lyn Hinton3, Paul Nelson4, Sui Yu3,4, Leanne M Dibbens5,6, Samuel F Berkovic1, John C Mulley5,6,7

1Epilepsy Research Centre and Department of Medicine, University of Melbourne, Austin Health, Heidelberg, Victoria
2Department of Paediatrics, University of Melbourne, Royal Children's Hospital, Melbourne, Victoria
3Cytogenetics Unit, Genetics and Molecular Pathology, SA Pathology at Women's and Children's Hospital, North Adelaide
4Molecular Genetics Unit, Genetics and Molecular Pathology, SA Pathology at Women's and Children's Hospital, North Adelaide
5Epilepsy Research Program, Genetics and Molecular Pathology, SA Pathology at Women's and Children's Hospital, North Adelaide
6School of Paediatrics and Reproductive Medicine, The University of Adelaide, South Australia
7School of Molecular and Biomedical Sciences, The University of Adelaide, South Australia

Article reçu le 10 Decembre 2009, accepté le 22 Mai 2010

High resolution molecular approaches are now replacing light microscopy in cytogenetics. Small structural chromosomal aberrations are detected by genome-wide oligonucleotide-array comparative genome hybridization (CGH), which interrogates the entire human genome (Friedman et al., 2006). This approach simultaneously detects benign copy number variants (CNVs) comprising part of the normal differences between individuals, making the interpretation of low frequency and novel CNVs as benign or pathogenic challenging (Sharp et al., 2008). One recurrent 1.5 megabase (Mb) interstitial deletion at 15q13.3 accounts for one percent of cases with idiopathic generalised epilepsy (IGE) (Dibbens et al., 2009; Helbig et al., 2009). The same microdeletion has been detected in individuals with additional or alternative manifestations including intellectual disability (ID), autism spectrum disorders, schizophrenia (Mulley and Dibbens, 2009) and sometimes without any apparent disease phenotype (Dibbens et al., 2009; van Bon et al., 2009). Why the pathogenic CNVs can be so variable in their expressivity, and even nonpenetrant in some carriers, remains to be determined. Apart from targeted and anecdotal reports (de Kovel et al., 2010; Dibbens et al., 2009; Helbig et al., 2009; Heron et al., 2007; Marini et al., 2009; Mulley et al., 2006) there are no published genome-wide surveys in epilepsy cohorts to ascertain the overall rate of microchromosomal abnormalities associated with seizures, either treatable or refractory.

Extensive studies focussed on rearrangements of the subtelomere regions have identified pathogenic changes in patients who have both seizures and ID (Baker et al., 2002; Colleaux et al., 2001; Davies et al., 2003; Kleczkowska et al., 1993; Knight-Jones et al., 2000; Martin et al., 2002; Meinecke and Vogtel, 1987; Popp et al., 2002; Rio et al., 2002; Rossi et al., 2001; Slavotinek et al., 1999). For example, a study by Anderlid et al. (2002) found that four of their ten patients with ID and subtelomeric rearrangements also had seizures. Where described, predominantly generalised seizures have been identified in these patients, including myoclonic, absence, clonic (Knight-Jones et al., 2000) and generalised tonic-clonic seizures (Baker et al., 2002; Knight-Jones et al., 2000; Slavotinek et al., 1999), infantile spasms were the only other seizure type noted (Knight-Jones et al., 2000; Rossi et al., 2001).

Thirty per cent of idiopathic epilepsies have an inadequate response to antiepileptic drugs (Kwan and Brodie, 2000). Most cases with treatment failure have no known aetiology and have normal structural brain imaging. Refractory epilepsy can occur in the setting of normal intellect, but it can also be associated with varying degrees of ID. This association with ID may be due to a shared underlying aetiology, or in some cases an epileptic encephalopathy may cause the ID.

The 15q13.3 microdeletion is one example of a phenotype initially identified in individuals with ID that was subsequently extended to a pure IGE phenotype (Helbig et al., 2009). We know that the larger chromosomal abnormalities visible by light microscopy are often associated with seizures which in many cases are refractory (Singh et al., 2002). Here we posit that the smaller molecularly defined structural variations detectable by emerging technologies may also cause seizures. We examined cases with refractory seizures for the presence of microchromosomal abnormalities. Specific microchromosomal aberrations may lead to the identification of candidate genes for refractory epilepsy.

Patients and methods

Patients

We studied 20 subjects ascertained from the hospital and private epileptology practices of the authors and by referral to our epilepsy genetics research group. We selected cases with refractory epilepsy and no major structural lesions based on neuroimaging. The epilepsy syndrome for each patient was established, including data from EEG and MRI brain scans, where available. Intellectual status, whilst not a selection criterion, was determined by neuropsychological assessment, or when unavailable, clinical observation. The results of other investigations such as cytogenetics and fragile X molecular testing were also obtained. Cases of Dravet syndrome were excluded.

Patient 1 was included in the first phase of this study with subtelomere FISH and MLPA (see below), and was subsequently diagnosed with epilepsy and mental retardation limited to females (EFMR) with a known PCDH19 mutation (Dibbens et al., 2008). Since EFMR may be responsive to treatment (Scheffer et al., 2008) and her affected sister was not refractory, this subject was kept within the test cohort to determine if additional structural variation might account for the refractory nature of the seizures. Patient 15 was known to have a balanced translocation t(3;16)(p21.32;p11.1) which is also carried by his unaffected father and therefore thought to be unrelated to either the epilepsy or its refractory nature.

Subtelomere FISH

Fluorescence in situ hybridization (FISH) was carried out as previously described (Baker et al., 2002). This involved scanning all 41 of the subtelomeric regions from the human chromosome complement for deletions using a standard set of FISH probes.

Multiplex ligation-dependent probe amplification

Multiplex ligation-dependent probe amplification (MLPA) analysis (Schouten et al., 2002) expanded the sensitivity of the subtelomeric scan to detect duplications and smaller deletions, neither of which would be detectable by FISH. Genomic DNA of each patient was isolated using firstly a Qiagen Blood DNA mini kit and then the DNA was re-extracted using a Qiagen QIAquick PCR purification kit. MLPA testing was carried out using the SALSA PO69 and PO36B human telomere test kits (MRC-Holland, Amsterdam, The Netherlands). Both SALSA PO36B and PO69 kits contain one unique probe per subtelomeric region for all chromosomes except the short arms of acrocentric chromosomes. The two probe sets have no target sequence in common, but are generally within half a Mb of the telomere, have exact and known locations on the DNA sequence map and generally detect duplications and smaller deletions missed by multiprobe FISH (Northrop et al., 2005). In addition, SALSA P036B contains probes in the pericentromeric long-arm regions of the acrocentric chromosomes (referred as to 13*, 14*, 15*, 21*, 22*). MLPA analysis was carried out according to the manufacturer's instructions. PCR products were separated and quantified by capillary electrophoresis on an ABI 3100 Avant DNA analyzer (Applied Biosystem), using GeneMapper analysis software (version 3.7). Interpretation of output was done as described by (Mulley et al., 2006).

Array-CGH

Array comparative genomic hybridisation (array-CGH) was performed initially on a Nimblegen 135K 12 plex whole genome array (HG18 cat no 080310) with a same sex dye reversal. The hybridisation data was analysed using SignalMap software (v1.9 Roche NimbleGen; Madison, Wisconsin). Positive results were confirmed with a second array platform; SurePrint G3 human CGH 8X60K oligonucleotide microarray (Agilent Technologies, Santa Clara CA cat no.G4450A). These results were analysed using DNA Analytics (v 4.0 Agilent Technologies, Santa Clara CA).

The 10q21.2 microduplication detected by array CGH was further confirmed by dye swap and its gene content determined by Ensembl. The 15q13.3 microduplication was confirmed by quantitative PCR as described previously (Dibbens et al., 2009) and its gene content determined by Ensembl to verify that it was identical to the recurrent 15q13.3 microdeletion syndrome associated with a range of previously defined syndromes, including IGE (Mulley and Dibbens, 2009).

The Austin Health Human Research Ethics Committee approved this study and informed consent was obtained for all subjects.

Results

The subjects with refractory seizures had a variety of epilepsy syndromes. Average age of seizure onset was 5.07 years (range: six weeks-16 years). Thirteen of the 20 patients had generalised epilepsy, with generalised spike or polyspike and wave discharges based on EEG recording. Six patients had refractory focal epilepsy and one patient had both focal and generalised epilepsy syndromes. Their clinical details are summarised in table 1.

Neuropsychological testing had been conducted on 14 of the 20 patients. Intellectual disability was present in 10/20 patients, with the majority (8) falling in the mild ID range (table 1). Two had borderline intellect and one had normal development with later cognitive decline. MRI brain scans had been performed on 17 patients, with 14 reported as normal and three with non-specific findings.

None of the subjects had a documented family history of ID. Eight of the subjects had a family history of epilepsy, including patient 1 with EFMR. Standard cytogenetic analysis had been performed on 14 of the 20 patients. Patient 15 was found to have a balanced translocation t(3;16)(p21.32;p11.1) by standard cytogenetics. The other 13 subjects had normal chromosomes by standard cytogenetics. All 20 patients had standard molecular testing for fragile X syndrome and all were negative.

FISH and MLPA analyses of the 20 participants found no abnormalities in the subtelomere regions. High resolution oligonucleotide array CGH showed no detectable molecular abnormality at or near either of the translocation breakpoints 3p21.32 or 16p11.1 for the translocation in patient 15, whose status remained balanced at the molecular level. Patient 14 had the common recurrent IGE associated 15q13.3 microdeletion of approximately 1.5 Mb detected by both array platforms and confirmed by quantitative PCR. His mother was negative for the microdeletion. DNA from his deceased father was extracted from a paraffin embedded tumour tissue biopsy but was too degraded to obtain a result, thus it was not possible to determine whether the deletion was sporadic or familial. Patient 19 had a novel 10q21.2 microduplication of approximately 1 Mb, inherited from his unaffected mother. The 15q13.3 microdeletion and the 10q21.2 microduplication as detected by both the Nimblegen and Agilent oligonucleotide-array CGH platforms are shown in figure 1.
Table 1 Patients with refractory epilepsy.

Patient

Age (y)/gender

Age at seizure onset (y)

Seizure types

Number of AEDs

EEG

MRI

Intellect

Diagnosis

1

19 / F

2

CPS, SPS, H, SE

6

Right fronto-temporal ictal rhythm

Ventriculomegaly

Mild ID

Refractory focal epilepsy

2

22 / M

4

M, T

5

GSW, GPFA

Normal

Severe ID

Refractory SGE

3

22 / F

14

GTCS, M, Ab

5

GSW, PSW

ND

Normal

Refractory JME

4

16 / M

5.5

FS, GTCS, Ab

3

3Hz GSW

ND

Normal

Refractory generalised epilepsy

5

26 / F

10

GTCS, Ab, At

12

PSW

Normal

Normal

Refractory generalised epilepsy

6

4 / F

0.7

GTCS, CPS, SPS, Ab

3

Normal

Normal

Mild ID

Refractory focal epilepsy

7

41 / F

16

GTCS, M, Ab, NCS

7

2.5-3Hz GSW, PSW

Normal

Normal

Refractory generalised epilepsy

8

36 / F

4

GTCS, M, Ab, At, T, SE, NCS

8

2-3Hz GSW, GPFA with brief seizure

Normal

Normal with later decline

Refractory generalised epilepsy

9

20 / F

6

AEM, M

2

4-5Hz GSW, GPFA

ND

Learning difficulties

Absences with eyelid myoclonia

10

13 / F

2.5

FS, GTCS, CPS, M, H

7

Left temporal discharges

Normal

Borderline

Refractory focal epilepsy

11

20 / F

0.1

GTCS, CPS, SPS, T, SE

5

DS

Normal

Mild ID

Refractory focal epilepsy, ataxia

12

43 / F

10

GTCS, Ab, NCS

15

GSW, PSW

Normal

Normal

Refractory absence epilepsy

13

44 / M

5

GTCS, Ab

7

2-3Hz GSW, PSW

Normal

Mild ID

Refractory SGE

14

14 / M

3

Ab, single T

6

3Hz GSW

Normal

Borderline

Refractory CAE

15

19 / M

4

GTCS, Ab, M, SPS

12

GSW

Normal

Mild ID

Refractory generalised epilepsy

16

16 / F

1.5

Ab, CPS

6

GSW, PSW DS, MFID

Ventriculomegaly

Mild ID

Refractory TLE and CAE

17

14 / M

5

Ab, single T

3

GSW, PSW, MFID

Normal

Mild ID

Refractory CAE

18

19 / F

0.5

TCS, CPS, T

10

Slow spike-wave with GPFA

Normal

Moderate ID with decline

Refractory SGE

19

13 / M

5.5

SGTCS, CPS, SE

2

Irregular GSW

Hippocampal asymmetry

Normal

Refractory focal epilepsy

20

7 / M

2

TCS, Ab, At, M

4

Right temporo-occipital discharges

Normal

Mild ID

Refractory focal epilepsy

Discussion

Many of the epilepsies have been inferred or demonstrated to have a genetic basis (Helbig et al., 2008; Heron et al., 2007). Chromosomal imbalances have also been described in conditions involving seizures or EEG abnormalities at either the macro- (Singh et al., 2002) or micro- level (de Kovel et al., 2010; Helbig et al., 2009). Unbalanced translocations with deletions of the subtelomeres are known to lead to several severe syndromes involving ID and seizures, including Wolf-Hirschhorn and Miller-Dieker syndromes. Of particular note is ring 20, which is known to have a strong association with intractable epilepsy. Ring 20 formation may affect expression of neuronal genes within the p and/or q arm subtelomeric regions of chromosome 20.

In the previous studies that have identified subtelomeric changes in subjects with seizures, the nature of the seizures was predominantly generalised (Baker et al., 2002; Knight-Jones et al., 2000; Rossi et al., 2001; Slavotinek et al., 1999). Similarly, the majority of the refractory patients in this study have generalised epilepsies. Subtelomeric aberrations have been found in at least two patients with focal findings based on EEG recording (Knight-Jones et al., 2000). Unfortunately, most reports do not provide details of the seizures experienced by their subjects. We therefore included five patients with definite refractory focal epilepsy and three others with generalised and focal features based either on EEG or clinical features.

For our patients with refractory seizures we failed to find any changes in the subtelomeric regions at the resolution of either the FISH or MLPA molecular probes. Thus, we were not able to demonstrate that refractory seizures were commonly related to any specific subtelomeric microchromosomal rearrangement. The next question was whether interstitial sub-microscopic deletions or duplications anywhere in the remainder of the genome might trigger refractory seizures. The recently reported 15q13.3 microdeletion of approximately 1.5 Mb encompasses the genes MTMR15, MTMR10, TRPM1, KLF13, OTUD7A, CHRNA7 and ARHGAP11B and is associated with ID, autism, schizophrenia and IGE (Mulley and Dibbens, 2009). Based on published data from our group and others (de Kovel et al., 2010; Dibbens et al., 2009) and our own clinical experience, patients with the 15q13.3 microdeletion have seizures that are generally responsive to treatment. This microdeletion, whilst likely to contribute to the epilepsy phenotype seen in our patient, is not likely to be the cause of the refractory nature of his seizures.

The 10q21.2 microduplication spans 1.15 Mb and creates additional intact genomic copies of three genes: cell division control protein 2 (CDC2), a Rho GTPase (RHOBTB1) and an uncharacterised transmembrane protein 26 gene (TMEM26). It also duplicates the sequence encoding an antisense RNA (BC041470) and the 5’ end of the Ankyrin 3 gene (ANK3). Both CDC2 and TMEM26 are expressed in human brain (UCSC Genome Browser v215) making it possible that they contribute to the pathogenesis of seizures. Regulators of Rho proteins such as RHOBTB are known to regulate signal transduction and organization of the cytoskeleton. RHOBTB1 shows ubiquitous expression with high expression in foetal brain and regions of the adult brain including the hypothalamus, thalamus and prefrontal cortex; making it also feasible that duplication of this gene may contribute to the seizure phenotype. The Ankyrin 3 gene is also a good candidate epilepsy gene, showing high expression in foetal and adult brain as well as being involved in the clustering of sodium channels, which are well known to be involved in epilepsy, in axons. The microduplication overlaps the 5’ end of the normal copy of the ANK3 gene and therefore possibly disrupts the promoter region and thus expression of the normal gene copy. Thus, the mechanism by which the 10q21.2 microduplication contributes to the occurrence of refractory seizures in this patient is not yet understood.

There are no reports of this 10q21.2 microduplication in DECIPHER (Database of Chromosomal Imbalance and Phenotype in Humans) or in the Database of Genomic Variants (http://projects.tcag.ca/variation). Seven much smaller nonpathogenic CNVs (five deletions and two duplications) have been reported within the 10q21.2 region in the Database of Normal Variation (http://cnv.chop.edu). Whilst the size and novelty of the 10q21.2 duplication is highly suggestive of a role contributing to the seizures in this patient, incomplete penetrance would need to be invoked in order to explain the absence of symptoms in the mother with an affected child. Incomplete penetrance is frequently observed for other pathogenic CNVs in epilepsy and other disorders (Dibbens et al., 2009; Sharp et al., 2008; van Bon et al., 2009).

Eighteen of our 20 subjects did not have microchromosomal lesions. This pilot study suggests that this mechanism is not a frequent cause of epilepsy refractory to therapeutic treatments. The 15q13.3 deletion is known to act as a rare variant with high effect and thus requires additional genetic factors for expression. It is not normally associated with refractory seizures, thus, in our patient, other factors are likely responsible for refractoriness (de Kovel et al., 2010; Dibbens et al., 2009). The 10q21.2 microduplication is novel, but whether it can be excluded as causative for refractory seizures, since it was detected in the mother who did not have seizures, is difficult to assess. A range of other microdeletions are reported not to be associated with refractory seizures (de Kovel et al., 2010), consistent with the precedent established by the 15q13.3 deletion. Furthermore, no relationship has been confirmed between genetic variation and multidrug transporter proteins (Szoeke et al., 2006; Tan and Berkovic, 2006). Why one third of subjects with idiopathic epilepsy are refractory remains unsolved.

Acknowledgments

We are grateful to the patients and their families for participating in our research. We thank Beverly Johns and Marta Bayly for DNA extractions and carrying out the quantitative PCR to confirm the chromosome 15q13.3 microdeletion detected by array CGH.

Financial support.

This work was supported by SA Pathology, MS McLeod Foundation and the National Health and Medical Research Council of Australia.

Disclosure.

None of the authors has any conflict of interest to disclose.