Human Molecular Genetics Advance Access originally published online on March 16, 2005
Human Molecular Genetics 2005 14(8):1077-1086; doi:10.1093/hmg/ddi120
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Neonatal epileptic encephalopathy caused by mutations in the PNPO gene encoding pyridox(am)ine 5'-phosphate oxidase
1Institute of Child Health, University College London with Great Ormond Street Hospital for Children NHS Trust, London WC1N 1EH, UK, 2Guy's and St Thomas's Hospital NHS Trust, London SE1 9RT, UK, 3Neurometabolic Unit, National Hospital, Queen Square, London WC1N 3BG, UK, 4Department of Histopathlogy, Barts and The London School of Medicine and Dentistry, Queen Mary, University of London, London E1 1BB, UK, 5Division of Metabolic and Endocrine Diseases, Department of General Pediatrics, University Children's Hospital, D-69120 Heidelberg, Germany and 6Institute of Human Genetics, Ruprecht-Karls University, 69120 Heidelberg, Germany
* To whom correspondence should be addressed. Tel: +44 2079052628; Fax: +44 2074046191; Email: p.clayton{at}ich.ucl.ac.uk
Received January 19, 2005; Revised February 15, 2005; Accepted March 3, 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
In the mouse, neurotransmitter metabolism can be regulated by modulation of the synthesis of pyridoxal 5'-phosphate and failure to maintain pyridoxal phosphate (PLP) levels results in epilepsy. This study of five patients with neonatal epileptic encephalopathy suggests that the same is true in man. Cerebrospinal fluid and urine analyses indicated reduced activity of aromatic L-amino acid decarboxylase and other PLP-dependent enzymes. Seizures ceased with the administration of PLP, having been resistant to treatment with pyridoxine, suggesting a defect of pyridox(am)ine 5'-phosphate oxidase (PNPO). Sequencing of the PNPO gene identified homozygous missense, splice site and stop codon mutations. Expression studies in Chinese hamster ovary cells showed that the splice site (IVS3-1g>a) and stop codon (X262Q) mutations were null activity mutations and that the missense mutation (R229W) markedly reduced pyridox(am)ine phosphate oxidase activity. Maintenance of optimal PLP levels in the brain may be important in many neurological disorders in which neurotransmitter metabolism is disturbed (either as a primary or as a secondary phenomenon).
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
The main feature of neonatal epileptic encephalopathy (NEE) is the onset within hours of birth of a severe seizure disorder that does not respond to anticonvulsant drugs and can be fatal. Some patients with NEE have shown biochemical changes in the cerebrospinal fluid (CSF) and urine that indicate reduced activity of aromatic L-amino acid decarboxylase (AADC) (1
,2). These include (i) build up of metabolites of L-3,4-dihydroxyphenylalanine (L-DOPA) [including a markedly elevated CSF concentration of 3-methoxytyrosine and increased urinary excretion of vanillactic acid (VLA)] and (ii) decreased CSF concentrations of the dopamine metabolite, homovanillic acid (HVA), and the serotonin metabolite, 5-hydroxyindoleacetic acid (5-HIAA) (3–5). However, the reported patients with NEE had additional biochemical changes in the CSF (raised concentrations of glycine and threonine) (1,2) that were most easily explained by cellular deficiency of pyridoxal phosphate (PLP), the co-factor for AADC, threonine dehydratase and the glycine cleavage enzyme. Treatment with PLP stopped the seizures where pyridoxine treatment had failed (2). This suggested a defect in an enzyme required for the synthesis of intracellular PLP from dietary pyridoxine (but not from pyridoxal/PLP—pyridox(am)ine 5'-phosphate oxidase (PNPO) deficiency.
The human PNPO gene (OMIM 6032870) is situated on chromosome 17q21.2 and extends 
7.5 kb (6). It contains seven exons and produces an mRNA transcript of 2.4 kb that encodes a protein of 261 amino acids. We have sequenced PNPO in five patients with biochemistry suggestive of PNPO deficiency and have identified homozygous missense, splice site and stop codon mutations all of which markedly reduced PNPO activity when expressed in Chinese hamster ovary (CHO) cells.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Phenotype
Patient details are summarized in Tables 1 and 2. Additional features are given below.
|
|
Family G.
The parents were first cousins of Turkish origin and they lost at least four (and possibly six) children from NEE (1
). G1: the first child was born at 32 weeks gestation and died at 21 days. This infant was thought to have non-ketotic hyperglycinaemia (NKH) on the basis of seizures plus elevated plasma and CSF glycine, but threonine concentrations were also elevated. G2, G3, G4: triplets, conceived by in vitro fertilization (IVF) and born at 24 weeks; two died after 1 day and one survived longer. This child had elevated plasma glycine and threonine concentrations, severe feeding problems and intestinal obstruction; surgical exploration showed malrotation and volvulus. She also had recurrent anaemia requiring transfusion. She had treatment-resistant epilepsy from 2 weeks and also developed progressive microcephaly. She died at 6 months of her epilepsy plus symptoms of ileus and severe failure to thrive (1.56 kg). G5, G6: twins boys conceived by IVF and born at 29 weeks following premature labour. Both were in poor condition at birth, required intubation and ventilation and had a (compensated) metabolic acidosis on day 1 (e.g. G5: pH 7.36, pCO2, 3.2 kPa, base excess (BE)–10 mmol/l). From birth, they had severe convulsions, myoclonus, rotatory eye movements and sudden clonic contractions. The electroencephalogram (EEG) showed a burst suppression (BS) pattern and biochemical investigations revealed episodes of hypoglycaemia and also intermittent acidosis. The lactate concentrations in blood, CSF and urine were moderately elevated (e.g. G5: CSF lactate 3.2 mmol/l), and alanine concentrations were normal. The severe epilepsy was resistant to conventional anticonvulsant therapy. Treatment with pyridoxine led to a partial improvement of clonic contractions and lip-smacking automatisms. However, severe seizure activity continued and the infants died at the ages of 17 and 19 days, respectively. G7: the parents requested prenatal diagnosis for the next pregnancy. DNA analysis from the chorionic villus indicated that the foetus was affected (homozygous for the R229W mutation) and they requested the termination of pregnancy.
Family J.
The parents were of east African Asian origin and were second cousins (2). They had difficulty in conceiving and had two unsuccessful attempts at IVF and one miscarriage. Their two live-born children were both conceived naturally: J1: a daughter was born at 33 weeks gestation by emergency caesarean section for fetal distress. She had Apgar scores of 6 (1 min) and 7 (5 min) and a metabolic acidosis on day 1. Onset of seizures was at 2 h; a computed tomography (CT) scan of the brain was interpreted as showing bilateral ischaemic changes in the basal ganglia. The seizures were never fully controlled and she had additional problems with anaemia and with abdominal distension and vomiting and she died aged 5 weeks. J2: a son was born at 35 weeks gestation by emergency caesarean section for fetal distress (abnormal cardiotocogram (CTG) trace and meconium-stained liquor). Birth weight was 2.12 kg and Apgar scores 7 (1 min) and 8 (5 min). He developed signs of respiratory distress and required ventilation for 12 h; blood gas analysis revealed a profound metabolic acidosis (pH 7.09, BE–19 mmol/l) and he was also anaemic (packed cell volume 28%). He improved with sodium bicarbonate therapy and transfusion but then started to develop abnormal posturing, limb movements and mouth movements. Despite loading with phenobarbitone, he developed further fits and apnoea requiring re-intubation. Examination of the abdomen revealed a liver edge palpable 3 cm from the costal margin and abdominal ultrasound confirmed hepatomegaly. A cranial CT showed an area of cerebral oedema in the left parieto-occipital region and two EEGs showed BS and discontinuous patterns. The fits were now tonic–clonic with mouthing and eye movements. For 2 days, the fits appeared to be controlled by phenobarbitone. However, between the 7th and 12th day of life, he developed further seizures and eventually status epilepticus. The latter did not respond to pyridoxine (50 mg twice daily) or to a midazolam infusion, and he required thiopentone infusion and ventilation. By now, the EEG showed a severe BS pattern with bursts consisting entirely of sharpened epileptiform activity. During the initial period of acidosis, blood lactate was mildly elevated (e.g. blood lactate 2.84 mM, lactate/pyruvate ratio 44) but subsequent determinations and CSF lactate analysis were normal.
J2 was given 50 mg of PLP (Solgar) via his nasogastric tube. One hour later the seizures had stopped, but he had become extremely hypotonic and apnoeic with no response to pain. An EEG showed prolonged isoelectric periods with only transient responses to auditory stimulation. (These phenomena are similar to those seen during treatment of a neonate with pyridoxine-dependent epilepsy.) Treatment with PLP was continued at a dose of 10 mg/kg 6-hourly. The infant remained akinetic and unresponsive for 4 days (during which he had a period of hypotension requiring colloid and dopamine infusion). Thereafter, there was a steady improvement in both clinical status and EEG. Two weeks after starting PLP therapy and while still on vigabatrin as anticonvulsant, the infant had two further seizures and the dose of PLP was increased to 15 mg/kg 6-hourly. Repeat analyses of CSF were undertaken when J2 was receiving PLP therapy: the 3-methoxytyrosine concentration had fallen to within the normal range, the HVA had risen to within the normal range and the 5-HIAA had risen to just below the normal range. The CSF glycine concentration had fallen to normal, and concentrations of threonine, taurine and histidine remained elevated.
J2 was the only patient in this series to receive PLP treatment and only he survived beyond the neonatal period. Unfortunately, he showed persistent central hypotonia and by the second year of life severe painful dystonic spasms as well as some seizures. He had marked acquired microcephaly and moderate to severe developmental delay. Treatment with L-DOPA and carbidopa was helpful in controlling the dystonic spasms.
Family K.
The parents were of Asian origin and consanguineous. K1 was the first child, born at 32 weeks gestation, weighing 2.08 kg and with Apgar scores of 10 (1 and 5 min). She developed intractable seizures from 30 min; these were refractory to phenytoin, phenobarbitone, clonazepam, lignocaine, paraldehyde and piracetam. Pyridoxine, 100 mg twice daily, also failed to improve the seizures. The EEG was consistent with severe myoclonic encephalopathy. The infant was hypotonic but there was no organomegaly or dysmorphism. On the third day of life, a metabolic acidosis was noted (pH 7.05, pCO2 5.2 kPa, BE–20.1 mmol/l, HCO3 13.3 mmol/l). She was sufficiently sick at this time to require intubation and ventilation. She was extubated on the 10th day but again developed metabolic acidosis (pH 7.27, pCO2 3.9, PO2 4.6, BE–12 mmol/l, HCO3 13.3 mmol/l, lactate 1.4 mmol/l). Weight gain remained poor in spite of total parenteral nutrition and the infant died at 23 days. Blood lactate measurements had varied between 1.4 and 7.1 mmol/l, with the higher values being recorded when the infant was most unwell. Postmortem examination showed milk aspiration and bronchopneumonia. The thymus was atrophied. The liver was congested and showed extramedullary haemopoiesis. The kidneys showed proximal tubular necrosis. The brain showed widespread changes including laminar cortical and selective hippocampal damage. The degree of astrocytosis suggested that brain damage was probably prenatal in onset. Electron microscopy revealed unusual ‘fuzzy’ mitochondria in the myocardial tissue. K2, a sister of K1, was born following the spontaneous onset of premature labour at 29+4 weeks gestation with a weight of 1.23 kg and Apgar scores of 1 at 9 min and 5 at 10 min. She developed abnormal movements and a high-pitched cry within 30 min of birth. Continuous positive airway pressure (CPAP) was required for frequent apnoeas and full ventilation (with minimal pressures and oxygen) was required from day 3. The seizures remained resistant to phenobarbitone, phenytoin and clonazepam. Intravenous pyridoxine, 200 mg daily, was commenced without benefit. Reasonable seizure control was eventually achieved with a high-dose clonazepam infusion. The EEG showed a severe BS pattern, which continued even when no seizures were evident clinically. Pyridoxine treatment, although not effective in seizure control, was associated with a fall in 3-methoxytyrosine (from 2860 to 900 nmol/l) and a rise in 5-HIAA (from 151 to 299 nmol/l) but no rise in CSF HVA or reduction in CSF threonine. K2 became progressively less responsive, showing no spontaneous movements or spontaneous respiratory effort, and she died at the age of 15 days. Blood lactate levels had varied between 1.8 (stable) and 7.0 mM (unwell). K3: when the third pregnancy was diagnosed, the parents were offered prenatal diagnosis on the basis that both their affected children had been homozygous for mutations in the PNPO gene (X262Q) and this was probably responsible for their fatal epileptic encephalopathy. The foetus was shown to be homozygous normal and the pregnancy was continued. K3 was born at term and is a female infant now 13 months old who has not developed seizures and shown normal growth and development.
Mutation analysis
The affected infants in family G (1) were homozygous for a missense mutation, R229W, affecting a highly conserved arginine residue (Figs 1A, F and 2A). Restriction enzyme studies showed that the parents were heterozygous for the R229W mutation (Fig. 2A). This was not detected in 80 chromosomes from control subjects of Turkish origin (data not shown).
|
|
In family J (2
), the propositus was homozygous for the following variations (i) a missense mutation/polymorphism, E50K, in a non-conserved region of the gene (Figs 1B, F and 2B) and (ii) a splice site mutation IVS3-1g>a (Figs 1C and 2C). The mRNA in the patient's fibroblasts lacked exon 4 (54 bases) (Fig. 1E) confirming that the splice site mutation causes aberrant splicing. A restriction enzyme test showed that the parents were heterozygous for both mutations (Fig. 2B and C). These were not present in 114 control chromosomes of subjects of Asian origin (data not shown). In family K, the propositus (K1) was homozygous for a stop codon mutation (X262Q) (Figs 1D and 2D). Restriction enzyme studies showed that the affected sibling (K2) was also homozygous for the stop codon mutation and that the parents were heterozygotes (Fig. 2D). Prenatal diagnosis of a further sibling (K3) indicated that neither allele carried the stop codon mutation. The foetus was correctly predicted to be unaffected. This mutation was not detected in 160 chromosomes from control subjects of Asian origin (data not shown).
Expression in mammalian cells
In vitro expression studies in CHO cells showed that transfection with wild-type PNPO produced readily measurable activity, whereas transfection with DNA containing the R229W mutation resulted in a reduced level of PNPO enzyme activity of 30% (21, 40%; n=2) of that of the wild-type (Table 3). Transfection with DNA containing the X262Q stop codon mutation produced undetectable activity of PNPO. Similarly, transfection of CHO cells with a construct containing the cDNA of the fibroblasts of J2 engineered to mimic the effects of the splice site mutation (i.e. deletion of exon 4) plus E50K, produced no PNPO activity (Table 3). However, when the DNA contained E50K only, PNPO activity in the transfected cells was similar to that in cells transfected with wild-type DNA (128, 143%; n=2) (Table 3). It was concluded that the splice site mutation was the pathogenic mutation in J2.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Dietary vitamin B6 enters the blood stream as pyridoxine, pyridoxamine and pyridoxal. The conversion of pyridoxine and pyridoxamine to the active co-factor, PLP, requires the activity of a kinase and then of PNPO; synthesis of active co-factor from dietary pyridoxal or PLP requires the kinase only (7
). PNPO probably also plays a role in a salvage pathway recycling the co-factor from degraded enzymes (8). The PNPO enzyme present in the brain appears to be the same as that in the liver and other non-neural cells (8). In 2003, Musayev et al. (8) cloned the cDNA encoding human PNPO, expressed it in Escherichia coli and studied the properties of the recombinant enzyme. They showed that the three-dimensional folding of the human PNPO is very similar to those of the E. coli and yeast enzymes. Furthermore, the binding sites for the substrate and for the flavine mononucleotide (FMN) co-factor are highly conserved between the human and E. coli enzymes. Recently, Kang et al. (6) reported on the genomic organization and tissue distribution of human PNPO. They pointed out that the PNPO gene had the characteristics of a housekeeping gene: absence of TATA-like sequences, presence of Sp1-binding sites and more importantly, the presence of CpG islands in the regulatory region.
It has long been known that dietary deficiency of vitamin B6 can produce (i) biochemical changes suggestive of PLP deficiency (including raised plasma concentration and urinary excretion of threonine and glycine) (9) and (ii) seizures in young infants (10). Recent work in the mouse suggests that circadian control of enzyme activities in the brain and peripheral tissues may be achieved through regulation of tissue concentrations of PLP. Pyridoxal kinase is activated by three circadian PAR bZip transcription factors which are in turn activated by BMAL1 and CLOCK (the members of the positive limb of the circadian oscillator) and repressed by CRY and PER proteins (the members of the negative limb). When all three PAR bZip transcription factors are knocked out, mice show deranged neurotransmitter metabolism and severe epilepsy (11). In the liver, gluconeogenesis is abolished by the deletion of Bmal1 and depressed by the mutation of Clock (12). The study of inborn errors affecting PLP synthesis will provide a unique opportunity for studying the role of PLP as a regulator of enzyme activities in man.
All five patients with biochemical results suggestive of PNPO deficiency had mutations in the PNPO gene which could be expected to result in reduced enzyme activity, and all of these mutations were shown by expression studies to reduce or abolish enzyme activity. Affected infants in family G had a mutation of Arg229, the last amino acid in the sequence RLHDR (residues 225–229 in the human gene) which is completely conserved in all species examined. Structural studies of recombinant human PNPO have shown that Arg229 residue is one of many amino acids that help to bind FMN tightly to the active site of PNPO (8). Substitution of Arg with Trp (R229W) would be predicted to disrupt this interaction and that of other amino acids in the vicinity. The immediately adjacent Asp228 is involved in one of two pairs of intersubunit salt–bridge interactions, interacting with Arg181. This is conserved in the yeast, E. coli and human PNPO structures (8). In addition, in E. coli invariant residues Arg197 and His199 (which are the equivalent of Arg225 and His227 in human PNPO) are found in the flexible turn located between the strands S6 and S7 and directly interact with PLP at the active site (13). These two residues act as a clamp on PLP sandwiching the pyridine moiety onto the isoalloxazine ring of FMN while catalysis takes place (13). The expression studies in CHO cells confirmed the importance of Arg229 for the activity of PNPO with the substitution of Arg with Trp (R229W) markedly reducing the activity of PNPO (Table 3).
Affected infants in family J had two PNPO sequence variations (IVS3-1g>a and E50K). Expression studies with the construct engineered to contain both the splice site mutation and the E50K confirmed the mutant PNPO was not active (Table 3). We considered that E50K was more likely to be a polymorphism than a pathogenic mutation. It is not conserved across species and characterization of the roles of the N- and C-terminal regions of human PNPO has shown that deletion of the N-terminal 56 residues affects neither the binding of coenzyme nor catalytic activity (6). Expression studies of E50K alone confirmed that the substitution of Glu with Lys at this position did not have a deleterious effect on enzyme activity and hence is a polymorphism. Indeed, it appears that the mutant E50K PNPO is slightly more active than the wild-type PNPO (Table 3). Therefore, it can be concluded that the pathogenic mutation in this patient is the splice site mutation, IVS3-1g>a, which was predicted to result in the loss of exon 4. Studies of the mRNA in cultured skin fibroblasts confirmed the deletion of exon 4. If translated, this would produce a protein lacking amino acids 122–139. The complete loss of ß-sheet 4 and parts of 
-helix 2 and ß-sheet 5 (8) could be expected to have a major effect on the tertiary structure of PNPO. The deletion would also remove Arg138 and Gln139 which are involved in binding FMN (8).
The abolition of the stop codon (X262Q) would be predicted to cause a C-terminal 28 amino acid extension (Fig. 1F), which may disrupt the dimeric structure and therefore the activity of the enzyme. In addition, translational readthrough may disrupt determinants associated with the 3'-UTR, which are required for mRNA stability. This phenomenon is seen in chain termination mutations of the 2 globin gene. Seven single-nucleotide variants of the natural termination codon of the 2 globin gene have been reported which result in thalassemia (14). Mutation of the stop codon of the chain of haemoglobin enables mRNA translation to continue to the next in-phase termination codon located within the polyadenylation signal and extends the chain by 31 amino acids. Studies have shown that the mRNA of the most extensively studied variant of this group (chain termination mutants), HbCONSTANT SPRING, is unstable because of disruption of a putative RNA/protein complex associated with the 2 globin 3'-UTR which is required for mRNA stability in erythroid cells (15). Expression studies confirmed that X262Q PNPO was a null activity mutant (Table 3).
Having confirmed that all five patients had mutations in the PNPO gene that reduce the activity of the enzyme, we can now review the phenotype of PNPO deficiency (Tables 1 and 2). All the patients were born prematurely and all but one had low Apgar scores and/or required intubation. Early acidosis was also common. Thus, PNPO-deficiency must enter the differential diagnosis of hypoxic-ischaemic encephalopathy (HIE) in a prematurely born infant. Further work is needed to determine whether a simple non-invasive test such as urinary organic acids (vanillactate) can be used to screen neonates diagnosed as having HIE for PNPO deficiency. Seizures commenced on the first day of life and were associated with an EEG showing a BS pattern. PNPO deficiency must, therefore, figure in the differential diagnosis of Ohtahara syndrome and NKH. Fortunately, it would appear that the CSF glycine/plasma glycine ratio in PNPO deficiency is not as high as it is in NKH and, of course, the other amino acid, and neurotransmitter amine metabolite abnormalities are not seen in NKH. An EEG showing BS is usually considered an indicator of severe cortical dysfunction—the EEG is flat apart from the intermittent seizure activity. Our results indicate that BS cannot be considered a marker of irreversible cortical damage. As it is frequently seen in NKH as well as in PNPO deficiency, it may sometimes be a marker of an elevated glycine concentration in the brain: in PNPO deficiency, this is reversible.
The biochemical abnormalities in the CSF and urine were as for AADC deficiency with the additional features of raised glycine (in all five), threonine (4/5), taurine (4/5), histidine (5/5) and low arginine (3/5). These could all be explained by reduced activities of specific PLP-dependent enzymes. Glycine, threonine and histidine are catabolized by PLP-dependent enzymes. One possible explanation for an elevated taurine concentration can be derived from the consideration of the pathway for taurine synthesis. Taurine is synthesised from cysteine via cysteine sulphinate. Cysteine sulphinate has two possible metabolic fates; it can be converted to pyruvate and sulphite by a PLP-dependent aspartate aminotransferase or converted via a PLP-dependent decarboxylase to hypotaurine (and hence to taurine). The aminotransferase reaction involves the conversion of enzyme-bound PLP to pyridoxamine-phosphate (at least in E. coli) (16). Thus, this enzyme (like threonine dehydratase) might be particularly susceptible to deficiency of the enzyme required to convert pyridoxamine-phosphate back to PLP. Thus, PNPO deficiency might favour the conversion of cysteine sulphinate to taurine. Another possibility is that raised CSF taurine is merely a reflection of cell damage as taurine is present at high concentrations inside cells. The three patients with low CSF arginine also had low plasma arginine suggesting that this was the cause.
Concentrations of amino acids in plasma were highly variable, depending on dietary protein/intravenous amino acid intake. Most of the samples were taken when protein intake was low and showed normal/elevated glycine and threonine and low arginine. The low arginine in plasma and secondarily in CSF was probably due to reduced synthesis from proline and glutamate via 
1-pyrroline 5-carboxylate (P5C) and ornithine due to reduced activity of ornithine 
-aminotransferase (OAT). In the neonate (in contrast to the adult), OAT activity is orientated towards ornithine and arginine synthesis (17).
An elevated blood lactate was documented at one time or another in all five patients. The significance of this is difficult to interpret in such sick infants; levels were not sufficiently high to be the main cause of marked acidosis when it occurred.
Hypoglycaemia may have been in part attributable to the disordered glucose homeostasis that occurs in sick premature infants (18). We had insufficient data on fuels and hormones to determine whether impaired gluconeogenesis or glycogenolysis (secondary to the PLP deficiency) might be a contributory factor.
Measurement of PLP and pyridoxal in frozen CSF samples from three of the infants showed all had low concentrations of both these compounds. These results need to be interpreted with caution as samples were taken when the patients were having very frequent seizures and a poor oral intake of vitamin B6. In order to answer the question of whether PNPO is required to recycle PLP from pyridoxamine-phosphate, we will need methods that can measure all the B6 vitamers. PLP given by nasogastric tube to J2 was dramatically effective in stopping the seizures and improving the appearances of the EEG. When ingested in physiological doses, PLP is hydrolysed by intestinal phosphatase(s) prior to absorption to pyridoxal. It is likely that the same occurs with pharmacological doses of PLP and that treatment with pyridoxal would be equally effective. Measurements made when J2 was receiving 30 mg/kg/day of PLP indicated that the CSF PLP had normalized but only at the expense of a 35-fold increase in the CSF level of pyridoxal.
The family histories of our patients suggest that parents heterozygous for PNPO deficiency have reduced rates of conception and perhaps also an incidence of homozygous affected pregnancies higher than the expected one in four. There are three observations that may be of relevance here. First, it is known that high doses of pyridoxine are toxic to the testis and spermatozoa (although the mechanism is not understood) (19). Secondly, both testis and spermatozoa express a unique transcript of the pyridoxal kinase gene (20) (the translation product of which can probably convert pyridoxine to pyridoxine-5'-phosphate, one of the substrates for PNPO). Thirdly, rats given pyridoxine show increased rates of implantation and an increased number of live births but at high doses, a reduction in the body weight of the pups (21).
In summary, mutations in the PNPO gene resulting in reduced PNPO activity are responsible for ‘NEE mimicking aromatic amino acid decarboxylase deficiency’ and for ‘pyridoxine-resistant, PLP-sensitive seizures’. The biochemical phenotype of PNPO deficiency is complex but the observed abnormalities in the CSF and urine can largely be explained by defects in specific PLP-dependent enzymes. These effects are important, not only for the management of this disorder but also for our understanding of (i) how PLP homoeostasis prevents epilepsy and (ii) how the biochemical activity of the brain and liver can be regulated by variation in PLP levels (e.g. in their circadian rhythms). Finally, our findings have implications for the treatment of epilepsy. PLP has been used successfully in Japan in the treatment of a range of epilepsies from infantile spasms to status epilepticus in an adult (22–24); neurologists in the west need to look anew at its role.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Patients
All studies were approved by the Ethics Committee of Great Ormond Street Hospital/Institute of Child Health. Patient details are summarized in Tables 1 and 2 and in Results.
Mutation analysis of the PNPO gene
Genomic DNA was extracted from venous blood by a modified version of the ammonium acetate salting out method (25,26). In the case of K1, genomic DNA was extracted from samples of occipital lobe, liver and muscle obtained at postmortem and stored frozen. The 7 exons and intron/exon boundaries of the PNPO gene were amplified by PCR using intronic primers (Supplementary Material, Table S1). A typical PCR reaction using 100 ng of genomic DNA contained 25 pmol of each primer, 1xNH4 reaction buffer (Bioline Ltd, London, UK), 0.2 mmol/l dNTPs and 0.5 µl (2.5 units) BioPro DNA polymerase (Bioline Ltd) (added after a ‘hot start’). Details of the annealing temperature and MgCl2 concentrations used are provided in Supplementary Material, Table S1. Cycling conditions were typically 96°C for 10 min, followed by 35 cycles of 30 s at 96°C, 30 s at 53–68°C, 30–90 s at 72°C and a final extension at 72°C for 10 min. PCRx Enhancer System (Invitrogen Ltd, Paisley, UK) was used for amplification of exon 2. Mutations were detected directly by sequencing the amplified regions using the BigDye Terminator version 3.1 Cycle Sequencing Kit (Applied Biosystems, Warrington, UK) and the MegaBase capillary DNA sequencer (Amersham Biosciences UK Ltd, Chalfont St Giles, Bucks, UK). All new sequence changes were confirmed by digestion with a restriction enzyme.
cDNA sequencing
Total RNA was extracted from fibroblasts using TRIZOL® reagent (Invitrogen Ltd) according to the manufacturer's instructions. An aliquot of 2.5 µg of total RNA in a final volume of 20 µl were treated with 2 units of DNase I (Invitrogen Ltd) for 15 min at room temperature. DNase I was inactivated by the addition of 5 mM EDTA (pH 8.0) and heating at 65°C for 10 min. One microliter of oligo(dT)20 (50 µM) (MWG-Biotech AG, Milton Keynes, UK) and 1 µl of random nonamer primers (50 µM) (MWG-Biotech AG) were annealed to the RNA at 70°C for 10 min. The reaction was placed immediately on ice for 1 min prior to first strand cDNA synthesis in a total volume of 45 µl containing 1xFirst strand buffer (Invitrogen Ltd), 10 mM DTT (Invitrogen Ltd), 0.5 mM dNTPs (Bioline Ltd, London, UK), 20 units RNase inhibitor (Invitrogen Ltd) and 200 units SuperScriptTM III RNase H– reverse transcriptase (Invitrogen Ltd). The reaction was carried out at 42°C for 50 min and 15 min at 70°C. An aliquot of 2 µl of RT reaction was used to amplify the PNPO cDNA using primers shown in Supplementary Material, Table S1. After initial denaturation at 94°C for 5 min, PCR consisted of 35 cycles of 30 s at 94°C (denaturation), 30 s at 72°C (annealing) and 30–90 s at 72°C (extension). A final extension time of 10 min at 72°C was used. The DNA products were sequenced as described previously.
Cloning and generation of mutations by site-directed mutagenesis
To facilitate expression vector construction, an EcoRI recognition site was introduced at both ends of the PNPO cDNA by PCR with primers shown in Supplementary Material, Table S1. The cDNA was amplified using ProofStart DNA polymerase (Qiagen Ltd, West Sussex, UK) and the primers described in Supplementary Material, Table S1. PCR was carried out for 50 cycles of 30 s at 94°C (denaturation), 30 s at 65°C (annealing) and 75 s at 72°C. A final extension time of 10 min at 72°C was used. The wild-type PNPO cDNA and the PNPO cDNA of HJ were cloned into the EcoRI-restriction site of the plasmid vector pSP72 (Promega, Southampton, UK). Site-directed mutagenesis was carried out using the QuikChange® XL Site-Directed Mutagenesis Kit (Stratagene, La Jolla, CA, USA), according to the manufacturer's instructions. Following mutagenesis, the PNPO gene from one clone was completely sequenced to confirm that no other errors had been introduced into the cDNA by the PCR-based procedure. Then, an EcoRI digested fragment was subcloned into the mammalian expression vector pIRES2–enhanced green fluorescent protein (EGFP) (BD BioSciences Clontech, Palo Alto, CA, USA). Correct orientation of the gene was confirmed by digestion with ApaI.
Expression studies
CHO cells were cultured in Dulbecco's modifed Eagle's medium (DMEM) (Invitrogen Ltd) supplemented with 10% fetal bovine serum (Sigma-Aldrich, Gillingham, Dorset, UK), 1% (v/v) penicillin/streptomycin (Sigma-Aldrich) and 2 mmol/l of L-glutamine (Sigma-Aldrich) at 37°C in a 5% (v/v) CO2/air atmosphere. Cells were transfected with wild-type or mutant cDNA constructs using LipofectamineTM transfection agent (Invitrogen Ltd). Experiments were carried out in 175 cm2 flasks and 3.52x106 cells were seeded prior to transfection. Each construct was transfected using 14 µg of DNA and 88 µl Lipofectamine. The transfection complexes were removed 6 h after transfection and replaced with supplemented DMEM as described previously. Twenty-four hours after transfection, the cells were visualized using an inverted fluorescent microscope, and the efficiency of transfection was estimated by the number of cells expressing EGFP. Cells were washed twice with phosphate-buffered saline (PBS) and harvested in 8 ml of PBS. Cells were centrifuged at 778g for 10 min and were resuspended in 100 µl cold water. The lysates used for the protein and enzyme assays were isolated from the cells by freezing (–150°C) and thawing (37°C) three times. Lysates were centrifuged at 16 000g for 5 min and the resulting supernatant represented the crude cellular lysate. PNPO enzyme activity was measured with pyridoxamine 5'-phosphate as substrate using the method described by Kang et al. (6), but with the following modifications: enzyme activity was measured by following the formation of PLP at 414 nm at 25°C in 0.1 M Tris–HCl (pH 8.4) containing 0.1 mM PMP in a total volume of 1 ml.
Measurement of pyridoxal and PLP in CSF
The concentration of pyridoxal and PLP in CSF was determined using a Chromsystems HPLC kit (München, Germany).
|
SUPPLEMENTARY MATERIAL |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
Supplementary Material is available at HMG Online.
|
ACKNOWLEDGEMENTS |
|---|
This work was funded by a grant from the Birth Defects Foundation (UK) and by the Horst Bickel Prize (sponsored by SHS Gesellschaft für klinische Ernähhrung mbH, Heilbronn, Germany). Research at the Institute of Child Health and Great Ormond Street Hospital for Children NHS Trust benefits from R & D funding received from the NHS Executive. We are grateful to Kerra Pearce for her skilful operation of the MegaBase sequencer within the London IDEAS Genetics Knowledge Park.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONRESULTSDISCUSSIONMATERIALS AND METHODSSUPPLEMENTARY MATERIALREFERENCES |
|---|
- Bräutigam, C., Hyland, K., Wevers, R., Sharma, R., Wagner, L., Stock, G.J., Heitmann, F. and Hoffmann, G.F. (2002) Clinical and laboratory findings in twins with neonatal epileptic encephalopathy mimicking aromatic L-amino acid decarboxylase deficiency. Neuropediatrics, 33, 113–117.[CrossRef][Web of Science][Medline]
- Clayton, P.T., Surtees, R.A., DeVile, C., Hyland, K. and Heales, S.J. (2003) Neonatal epileptic encephalopathy. Lancet, 361, 1614.[CrossRef][Web of Science][Medline]
- Hyland, K., and Clayton, P.T. (1990) Aromatic amino acid decarboxylase deficiency in twins. J. Inherit. Metab. Dis., 13, 301–304.[CrossRef][Web of Science][Medline]
-
Hyland, K., Surtees, R.A.H., Rodeck, C. and Clayton, P.T. (1992) Aromatic L-amino acid decarboxylase deficiency: clinical features, diagnosis, and treatment of a new inborn error of neurotransmitter amine synthesis. Neurology, 42, 1980–1988.
[Abstract/Free Full Text] -
Pons, R., Ford, B., Chiriboga, C.A., Clayton, P.T., Hinton, V., Hyland, K., Sharma, R. and De Vivo, D.C. (2004) Aromatic L-amino acid decarboxylase deficiency: clinical features, treatment, and prognosis. Neurology, 62, 1058–1065.
[Abstract/Free Full Text] - Kang, J.H., Hong, M.L., Kim, D.W., Park, J., Kang, T.C., Won, M.H., Baek, N.I., Moon, B.J., Choi, S.Y. and Kwon, O.S. (2004) Genomic organization, tissue distribution and deletion mutation of human pyridoxine 5'-phosphate oxidase. Eur. J. Biochem., 271, 2452–2461.[Web of Science][Medline]
-
McCormick, D.B. and Chen, H. (1999) Update on interconversions of vitamin B-6 with its coenzyme. J. Nutr., 129, 325–327.
[Abstract/Free Full Text] - Musayev, F.N., Di Salvo, M.L., Ko, T.P., Schirch, V. and Safo, M.K. (2003) Structure and properties of recombinant human pyridoxine 5'-phosphate oxidase. Protein Sci., 12, 1455–1463.[CrossRef][Web of Science][Medline]
-
Park, Y.K. and Linkswiler, H. (1970) Effect of vitamin B6 depletion in adult man on the excretion of cystathionine and other methionine metabolites. J. Nutr., 100,110–116.
[Abstract/Free Full Text] -
Coursin, D.B. (1954) Convulsive seizures in infants with a pyridoxine deficient diet. JAMA, 154, 406–408.
[Abstract/Free Full Text] -
Gachon, F., Fonjallaz, P., Damiola, F., Gos, P., Kodama, T., Zakany, J., Duboule, D., Petit, B., Tafti, M. and Schibler, U. (2004) The loss of circadian PAR bZip transcription factors results in epilepsy. Genes Dev., 18, 1397–1412.
[Abstract/Free Full Text] - Rudic, R.D., McNamara, P., Curtis, A.M., Boston, R.C., Panda, S., Hogenesch, J.B. and Fitzgerald, G.A. (2004) BMAL1 and CLOCK, two essential components of the circadian clock, are involved in glucose homeostasis. PLoS Biol., 2, 1893–1899.
- di Salvo, M.L., Safo, M.K., Musayev, F.N., Bossa, F. and Schirch, V. (2003) Structure and mechanism of Escherichia coli pyridoxine 5'-phosphate oxidase. Biochim. Biophys. Acta, 1647, 76–82.[Medline]
- Weatherall, D.J., Clegg, J.B., Higgs, D.R., and Wood, W.G. (2001) The hemoglobinopathies. In Scriver, C.R., Beaudet, A.L., Sly, W.S. and Valle, D. (eds), The Metabolic and Molecular Bases of Inherited Disease, 8th edn. McGraw Hill, New York, pp. 4571–4636.
-
Weiss, I.M. and Liebhaber, S.A. (1994) Erythroid cell-specific determinants of alpha-globin mRNA stability. Mol. Cell. Biol., 14, 8123–8132.
[Abstract/Free Full Text] - Furumo, N.C. and Kirsch, J.F. (1995) Accumulation of the quinonoid intermediate in the reaction catalyzed by aspartate aminotransferase with cysteine sulfinic acid. Arch. Biochem. Biophys., 319, 49–54.[CrossRef][Medline]
- Champion, M.P., Bird, S., Fensom, T. and Dalton, R.N. (2002) Ornithine aminotransferase deficiency (gyrate atrophy) presenting with hyperammonaemic encephalopathy. J. Inher. Metab. Dis., 25, 29.
- Cowett, R.M. and Farrag, H.M. (2004) Selected principles of perinatal–neonatal glucose metabolism. Semin. Neonatol., 9, 37–47.[CrossRef][Medline]
- Plassmann, S. and Urwyler, H. (2001) Improved risk assessment by screening sperm parameters. Toxicol. Lett., 119, 157–171.[CrossRef][Web of Science][Medline]
- Fang, X., Zhou, Z.M., Lu, L., Yin, L.L., Li, J.M., Zhen, Y., Wang, H. and Sha, J.H. (2004) Expression of a novel pyridoxal kinase mRNA splice variant, PKH-T, in human testis. Asian J. Androl., 6, 83–91.[Web of Science][Medline]
- Marathe, M.R. and Thomas, G.P. (1987) Effects of pyridoxine on implantation and pregnancy in Wistar rat. Toxicol. Lett., 39, 159–164.[CrossRef][Web of Science][Medline]
- Ohtsuka, Y., Matsuda, M., Ogino, T., Kobayashi, K. and Ohtahara, S. (1987) Treatment of the west syndrome with high-dose pyridoxal phosphate. Brain Dev., 9, 418–421.[Web of Science][Medline]
- Takuma, Y. (1998) ACTH therapy for infantile spasms: a combination therapy with high-dose pyridoxal phosphate and low-dose ACTH. Epilepsia, 39, 42–45.
-
Nakagawa, E., Tanaka, T., Ohno, M., Yamano, T. and Shimada, M. (1997) Efficacy of pyridoxal phosphate in treating an adult with intractable status epilepticus. Neurology, 48, 1468–1469.
[Free Full Text] -
Miller, M.A. (1988) A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res., 16, 1215.
[Free Full Text] -
Davies, J.P., Winchester, B.G. and Malcolm, S. (1993) Sequence variations in the first exon of -galactosidase. Am. J. Med. Genet., 30, 658–663.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  F. N. Musayev, M. L. Di Salvo, M. A. Saavedra, R. Contestabile, M. S. Ghatge, A. Haynes, V. Schirch, and M. K. Safo Molecular Basis of Reduced Pyridoxine 5'-Phosphate Oxidase Catalytic Activity in Neonatal Epileptic Encephalopathy Disorder J. Biol. Chem., November 6, 2009; 284(45): 30949 - 30956. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Bagci, J Zschocke, G F Hoffmann, T Bast, J Klepper, A Muller, A Heep, P Bartmann, and A R Franz Pyridoxal phosphate-dependent neonatal epileptic encephalopathy BMJ Case Reports, May 10, 2009; 2009(may10_1): bcr1120081247 - bcr1120081247. [Abstract] [Full Text]
|
|
|
|
 |
|
 K. Hyland Clinical Utility of Monoamine Neurotransmitter Metabolite Analysis in Cerebrospinal Fluid Clin. Chem., April 1, 2008; 54(4): 633 - 641. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S Bagci, J Zschocke, G F Hoffmann, T Bast, J Klepper, A Muller, A Heep, P Bartmann, and A R Franz Pyridoxal phosphate-dependent neonatal epileptic encephalopathy Arch. Dis. Child. Fetal Neonatal Ed., March 1, 2008; 93(2): F151 - F152. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. Surtees and N. Wolf Treatable neonatal epilepsy Arch. Dis. Child., August 1, 2007; 92(8): 659 - 661. [Full Text] [PDF]
|
|
|
|
 |
|
 K. Hyland Inherited Disorders Affecting Dopamine and Serotonin: Critical Neurotransmitters Derived from Aromatic Amino Acids J. Nutr., June 1, 2007; 137(6): 1568S - 1572S. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. L. Pearl, J. L. Taylor, S. Trzcinski, and A. Sokohl The Pediatric Neurotransmitter Disorders J Child Neurol, May 1, 2007; 22(5): 606 - 616. [Abstract] [PDF]
|
|
|
|
|
|
S. Bach, M. Knockaert, J. Reinhardt, O. Lozach, S. Schmitt, B. Baratte, M. Koken, S. P. Coburn, L. Tang, T. Jiang, et al. Roscovitine Targets, Protein Kinases and Pyridoxal Kinase J. Biol. Chem., September 2, 2005; 280(35): 31208 - 31219. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||