Abstract
ALDH7A1 and PNPO deficiencies are rare inborn errors of vitamin B6 metabolism causing perinatal seizure disorders. The phenotypic variability, however, is broad. To assess the frequency of these deficiencies in unexplained infantile epilepsy, we screened 113 patients for mutations in both genes. We identified 1 patient with an epilepsy phenotype resembling Dravet syndrome and likely pathogenic mutations in ALDH7A1. Presenting features were highly atypical of pyridoxine-dependent epilepsy, including febrile seizures, response to anticonvulsive drugs, and periods of seizure freedom without pyridoxine treatment. “Hidden” vitamin B6 deficiencies might be rare but treatable causes of unexplained epilepsy extending beyond the classical phenotypes.
Pyridoxine-dependent epilepsy and pyridoxamine-phosphate oxidase deficiency are rare autosomal recessive enzyme defects in the vitamin B6 metabolism due to pathogenic mutations in the ALDH7A1 and PNPO genes. The classical picture of both disorders is characterized by severe, persistent neonatal seizures that are refractory to treatment with anticonvulsant drugs but can be controlled by administration of pyridoxine or pyridoxal-5′-phosphate. 1,2
For pyridoxine-dependent epilepsy, various atypical clinical presentations have been described, including epilepsies starting as late as the third year of life and seizures partially responsive to antiepileptic drugs. 3,4 Also, patients have been reported with no instant or obvious response to pyridoxine, 4 comorbidities such as structural brain abnormalities and a history of birth asphyxia/hypoxic-ischemic encephalopathy. 5,6 The phenotype of pyridoxamine-phosphate oxidase deficiency is more severe and only few cases have been described, including 1 atypical case with a milder phenotype. 7
In order to assess the frequency of hidden vitamin B6 deficiencies in patients with unexplained seizures starting in the first 18 months of life, we screened a cohort of 113 epilepsy patients for mutations in the ALDH7A1 and PNPO genes. In this cohort, we identified 1 patient with a phenotype reminiscent of Dravet syndrome carrying heterozygous mutations in ALDH7A1.
Patients and Methods
Patients
Venous blood samples were obtained from a cohort of 113 patients with unexplained seizure disorders starting during the neonatal or infantile period (0-18 months) recruited on a prospective basis through 2 major pediatric epilepsy centers (Dianalund, Denmark, and Raisdorf/Kiel, Germany) during the years 2000-2011. The cohort also included patients with atypical clinical features such as birth asphyxia or hypoxic-ischemic encephalopathy, cerebral hemorrhage, lack of response to pyridoxine treatment, and variable age of onset in infancy as these symptoms have previously been described with pyridoxine-dependent epilepsy.
Molecular Analysis
DNA was isolated from blood following standard procedures. All 18 exons of ALDH7A1 and all 7 exons of PNPO including exon-intron boundaries were amplified by polymerase chain reaction (PCR) as previously described. 8 –10 Purified PCR products were directly sequenced using an ABI PRISM 3730 Genetic Analyzer (Applied Biosystems, Foster City, CA). Resulting sequence data were subsequently analyzed with Mutation Surveyor V3.24 (Softgenetics, State College, PA) comparing sequences with the reference sequences obtained from the National Center for Biotechnology Information Genome database (ALDH7A1: NM_001182.2; PNPO: NM_018129.2).
Identified variants including splice site mutations were analyzed for potential pathogenic effects using Alamut Version 2.0 (Interactive Biosoftware, Mont-Saint-Aignan, France) in combination with the freely available prediction programs Align-GVGD (http://www.agvgd.iarc.fr), PolyPhen Version 2 (http://www.genetics.bwh.harvard.edu/pph2/), and SIFT Alignment (http://www.sift.jcvi.org/). Potentially, pathogenic mutations were compared with the Human Gene Mutation Database (HGMD® Professional 2011.4, Biobase, Wolfenbüttel, Germany).
Results
Clinical Features of Patient Cohort
A total of 113 unrelated patients (58 male, 55 female) with severe unexplained epilepsy with a median onset of 5.1 months (ranging from intrauterine seizures to 18 months) were included (Table 1). Pyridoxine and/or pyridoxal-5′-phosphate treatment was documented in 7 of 113 patients without clear treatment response. In 8 patients with the clinical picture of Dravet syndrome, mutational analysis of SCN1A was negative.
Epilepsy syndromes in patients screened for mutations in ALDH7A1 and pyridoxamine-phosphate oxidase deficiency.
aMRI: unspecific abnormalities, that is, presumably not causal for the epilepsy, included delayed myelination, hypoplasia of corpus callosum, gray and/or white matter atrophy, cerebellar atrophy, periventricular leukomalacia, etc.
bMRI: lesional abnormalities, that is, possibly causal for the epilepsy, included intracranial hemorrhage, hippocampus atrophy/sclerosis, defects of neuronal migration, dysplasia, etc.
cEpilepsies with epileptic spasms as predominant seizure type but not fulfilling diagnostic criteria of West syndrome.
dSIGEI = severe idiopathic generalized epilepsy of infancy. SCN1A testing (Sanger sequencing and multiplex ligation-dependent probe amplification [MLPA]) was performed in all patients, 1 patient was identified to be mutation positive by exome sequencing during the course of this study.
eMyoclonic astatic epilepsy (n = 2), idiopathic generalized epilepsy (n = 3), generalized epilepsy with febrile seizures plus (n = 1).
Molecular Analysis
Sequence analysis of ALDH7A1 identified 1 patient with a known, common pathogenic mutation c.1195G>C; p.(Glu399Gln) 9 and a novel variant c.1468A>G; p.(Arg490Gly) (Figure 1). The novel missense mutation c.1468A>G shows a Grantham score of 12510 and was predicted to be likely deleterious by SIFT and Polyphen2. In conjunction with the overall phenotype of the patient (see below), the mutations were considered to be likely pathogenic, even though formal testing was not possible as parental DNA was not available and as metabolic testing could not be performed. No pathogenic mutations in PNPO could be detected. Identified single nucleotide polymorphisms and unknown variants for both genes are shown in Supplementary Table 1.
Electropherograms of the identified patient with pyridoxine-dependent epilepsy showing wild-type ALDH7A1 allele (above) with (A) c.1195G>C mutation in exon 14 and (B) c.1468A>G mutation in exon 17.
Case Summary
The boy was the second child of unrelated parents of Northern European decent. Pregnancy was complicated by maternal gestational diabetes, the patient was born at term with normal birth parameters. A single generalized tonic-clonic seizure occurred on day 5, with no further seizures in the neonatal period and early normal development. Prolonged generalized tonic-clonic seizure occurred at 3 and 6 months during febrile infections and further generalized tonic-clonic seizure were observed with fever approximately once per month. During febrile infections, the boy often appeared jittery and twitchy. Intermittently, generalized tonic-clonic seizure were more pronounced on the left. At age 2, alternating myoclonic seizures were observed. Later on, various seizure types occurred, including multifocal myoclonic seizures, atonic seizures, dyscognitive seizures, and rare tonic seizures as well as generalized tonic-clonic seizure with fever. Periods with seizures alternated with long seizure-free periods, only interrupted by seizures during febrile infections, especially with vomiting.
After achieving initial developmental milestones adequately, the patient had cognitive and speech delay beginning in the third year of life, and notable speech retardation as well as muscular hypotonia more prominent on the left was noted at the age of 3½ years. In adolescence, testing showed an Intelligence Quotient of 85; the patient was able to ride a bicycle, enjoyed swimming, and was capable of reading, writing, and simple calculations.
The patient was treated with primidone, phenytoin, valproate, ethosuximide, sulthiame, and carbamazepine. Seizure-free periods were seen on primidone and ethosuximide as well as on primidone, carbamazepine, and valproate.
A first electroencephalographic (EEG) recording during the neonatal period was unremarkable. EEG recordings later showed diffuse slowing and dysrhythmic background activity and abnormal theta rhythms. Intermittently, EEGs were reported to be normal.
Cranial ultrasonography in the neonatal period and repeated brain computed tomographic scans between 7 months and 19 years were unremarkable. The patient was negative for mutations in SCN1A. A deletion in AUTS2 gene was detected prior to ALDH7A1 testing, 11 considered to be unrelated to the epilepsy phenotype.
Routine metabolic screening at age of 3½ years showed unremarkable amino acid levels in serum and urine except for slightly elevated methylhistidine in urine (0.837 μmol/mL, reference limit: 0.7 μmol/mL).
The patient never received treatment with pyridoxine and died at age 31 from epileptic status prior to genetic diagnosis.
Discussion
As recent studies have described a broader phenotype spectrum in vitamin B6–dependent epilepsies as previously anticipated, 6 we were motivated to screen a cohort of 113 patients with early-onset epilepsies for mutations in ALDH7A1 and PNPO. In summary, we identified 1 patient with likely pathogenic mutations in ALDH7A1, consistent with the diagnosis of pyridoxine-dependent epilepsy. In contrast to ALDH7A1, we could not identify pathogenic mutations in PNPO. The phenotype of pyridoxamine-phosphate oxidase deficiency is usually more severe and might therefore be less frequent or absent in atypical cases.
The phenotype of the described patient extends the phenotypic spectrum of pyridoxine-dependent epilepsy. Although the patient had a single neonatal seizure on postnatal day 5, he presented with a Dravet-like phenotype including alternating prolonged seizures during febrile infections, theta rhythms, and generalized discharges on EEG as well as plateauing of cognitive development. In adolescence, the patient had long seizure-free periods of up to 4 years without pyridoxine treatment, which is counterintuitive at first glance. However, partial or initial response to treatment with antiepileptic drugs in patients with pyridoxine-dependent epilepsy has been reported previously. 3,4 An exacerbation during febrile illnesses has frequently been described in pyridoxine-dependent epilepsy, 3,7 which might mimic the Dravet-like phenotype.
The patient died at the age of 31 years during status epilepticus, prior to genetic diagnosis and metabolic testing. Therefore, we were unable to validate the pathogenicity of the mutations in ALDH7A1 through metabolic tests. However, until now, most studies only screened patients with abnormal metabolic testing for mutations in ALDH7A1 and PNPO 7 ; hence, we do not know the range of metabolic measures in mutation carriers and cannot comment on the validity of genetic testing versus metabolic screening.
In most previous studies on pyridoxine-dependent epilepsy and pyridoxamine-phosphate oxidase deficiency, genetic testing and phenotypic descriptions were performed in patients with a clinical suspicion of vitamin B6–dependent epilepsies. Therefore, these studies are less suited to detect cases with an atypical presentation. We performed a genetic screen for vitamin B6–dependent epilepsies in patients from specialized pediatric epilepsy centers, focusing on severe and often unexplained epilepsies starting in infancy. Our data suggest that pyridoxine-dependent epilepsy could be found in up to 1% of patients with unexplained severe epilepsy, a similar frequency previously found in the United Kingdom Infantile Spasms Study. 12
The deletion in AUTS2 gene was considered to contribute to the cognitive and speech delay, but being unrelated to the epilepsy phenotype. Molecular function of AUTS2 is unknown, but several patients with intellectual disability and autism spectrum disorder have been identified with disruptions of AUTS2, suggesting a role for AUTS2 in normal cognitive development. 13 –15 Epilepsy or seizures were described only in very few patients with complex genomic rearrangements or duplications of AUTS2. The largest study to date described phenotypic features of 17 patients with exon-disrupting deletions; none of these patients had epilepsy. 13
Pyridoxine-dependent epilepsy can be identified in patients with severe epilepsies including Dravet-like phenotypes. Fever sensitivity is a typical, but not exclusive, feature of Dravet syndrome and might occur in other epilepsy syndromes. Pyridoxine-dependent epilepsy could result in atypical phenotypes mimicking well-defined epilepsy syndromes, including partial response to anticonvulsive drugs and seizure-free periods without pyridoxine treatment. Therefore, patients with severe, unexplained epilepsy could be screened for mutations in ALDH7A1, either through genetic or metabolic analysis. This is particularly relevant given the treatment options in pyridoxine-dependent epilepsy and the possibility for genetic counseling.
Footnotes
Acknowledgment
We would like to thank all participating patients and their families.
Author Contributions
SvS, DL, US, and IH designed the study. AB, SvS, RSM, RB, HM, JAJ, LLK, and HH were responsible for patient recruitment and phenotyping. AB, NMVD, and MJAvK performed the sequence analysis. AB, SvS, MJAvK, and IH prepared the manuscript. MJAvK and IH were the mentors, who contributed equally to this work. All authors contributed to the final version of the manuscript.
Authors’ Note
Patient recruitment and phenotyping was done in Dianalund, Denmark, and Raisdorf/Kiel, Germany; DNA extraction was done by local research laboratories in Dianalund and Kiel; sequencing analysis was performed at Department of Medical Genetics, University Medical Center Utrecht, The Netherlands. The study was presented as a platform presentation at the national epilepsy conference in Germany (German Epilepsy Society [DGfE] meeting, Stuttgart, 2012) and as a poster at the European Congress on Epilepsy (London, 2012).
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by the medical faculty of the Christian-Albrechts-University of Kiel, Germany, the German Research Foundation (Grant HE 5415/3-1 to IH) and the Eurocores program EuroEPINOMICS of the European Science Foundation (http://www.euroepinomics.org).
Ethical Approval
The study was approved by the ethics committee of Christian-Albrechts-University, Kiel. Approval for patient recruitment for genetic research studies is also available at the Danish Epilepsy Center, Dianalund, Denmark. Informed consent was obtained from all patients or their legal guardians in case of minors.