PIGO mutations in intractable epilepsy and severe developmental delay with mild elevation of alkaline phosphatase levels



Aberrations in the glycosylphosphatidylinositol (GPI)–anchor biosynthesis pathway constitute a subclass of congenital disorders of glycosylation, and mutations in seven genes involved in this pathway have been identified. Among them, mutations in PIGV and PIGO, which are involved in the late stages of GPI-anchor synthesis, and PGAP2, which is involved in fatty-acid GPI-anchor remodeling, are all causative for hyperphosphatasia with mental retardation syndrome (HPMRS). Using whole exome sequencing, we identified novel compound heterozygous PIGO mutations (c.389C>A [p.Thr130Asn] and c.1288C>T [p.Gln430*]) in two siblings, one of them having epileptic encephalopathy. GPI-anchored proteins (CD16 and CD24) on blood granulocytes were slightly decreased compared with a control and his mother. Our patients lacked the characteristic features of HPMRS, such as facial dysmorphology (showing only a tented mouth) and hypoplasia of distal phalanges, and had only a mild elevation of serum alkaline phosphatase (ALP). Our findings therefore expand the clinical spectrum of GPI-anchor deficiencies involving PIGO mutations to include epileptic encephalopathy with mild elevation of ALP.

More than 100 mammalian cell-surface proteins are anchored to the plasma membrane by the addition of glycosylphosphatidylinositol (GPI) to their C-termini. More than 20 genes are involved in the GPI-anchor biosynthesis pathway[1, 2] of which 7 are mutated in GPI-anchor deficiencies, a subclass of congenital glycosylation disorders, in association with neurologic impairments.[3-7] Among them, mutations in PIGV, PIGO (both are involved in the last step of GPI-anchor synthesis), and PGAP2 (involved in fatty-acid GPI-anchor remodeling) have been identified in patients with hyperphosphatasia with mental retardation syndrome (HPMRS), also known as Mabry syndrome.[3-8]

PIGO encodes GPI ethanolamine phosphate transferase 3, which is also known as phosphatidylinositol-glycan biosynthesis class O. To date, only three HPMRS families with compound heterozygous mutations in PIGO have been reported. In this study, we performed whole exome sequencing of a Japanese family containing two affected siblings, one of them having epileptic encephalopathy, and identified novel PIGO mutations that expand the clinical spectrum of PIGO abnormalities to include epileptic encephalopathy.


DNA samples and subjects

All four family members (two affected siblings with epileptic encephalopathy and their parents) were analyzed. Clinical information, peripheral blood samples (individual II-1 and his parents), and the umbilical cord of individual II-2 were obtained after written informed consent was given. DNA was extracted using standard methods. Experimental protocols were approved by the institutional review board of Yokohama City University School of Medicine.

Whole exome sequencing (WES)

Genomic DNA was captured using the SureSelect Human All Exon v4 Kit (51 Mb; Agilent Technologies, Santa Clara, CA, U.S.A.) and sequenced on an Illumina HiSeq2000 (Illumina, San Diego, CA, U.S.A.) with 101 bp paired-end reads. Exome data processing, variant calling, and variant annotation were performed as previously described.[9] PIGO mutations detected by WES were confirmed by Sanger sequencing, and searched for in the variant database of our 408 in-house control exomes. For individual II-2, only those PIGO mutations identified in individual II-1 were checked by Sanger sequencing.

Flow cytometry

Surface expression of GPI-anchored proteins was examined as previously described.[8]


Clinical features

A summary of the clinical features of individuals II-1 and II-2 is shown in Table S1. Both siblings had intractable seizures and severe developmental delay, which were compatible with epileptic encephalopathy.

Case report 1

Individual II-1 is a 19-year-old male born to nonconsanguineous parents after a 38-week gestation with no asphyxia. His birth weight was 3,250 g (+0.5 standard deviation [SD]), height 52.0 cm (−1.4 SD), and head circumference 34.0 cm (−0.5 SD). Developmental milestones were delayed with no head control achieved at 6 months. At 1 year of age, he developed complex partial seizures with staring, crying, and irregular respiration leading to cyanosis. Brain magnetic resonance imaging (MRI) revealed no abnormalities (Fig. 1A,B). At 1 year and 11 months of age, he had intractable seizures refractory to valproate, zonisamide, and clonazepam. His body weight at this time was 10.54 kg (−0.8 SD), height 84.8 cm (−0.1 SD), and head circumference 45.3 cm (−1.9 SD). He was able to smile but unable to control his head or speak any meaningful words. He had a high arched palate and a tented mouth (Fig. 1E). His muscle tone was hypotonic, and deep tendon reflexes were normal with negative Babinski sign. Chorea was observed mainly in the upper extremities. He did not show brachytelephalangy or nail aplasia (Fig. 1F).

Figure 1.

T1-weighted brain MRI of individual II-1. Axial (A) and sagittal (B) images revealed no signal or structural abnormalities at 1 year of age. Axial (C) and sagittal (D) images at 6 years of age showing diffuse cerebral and cerebellar atrophy. Facial (E) and hand (F) photographs of individual II-1 at 19 years of age showing tented mouth (E) and no anomalous fingers (F).

Interictal electroencephalography (EEG), motor conduction velocities, visual evoked potential, short-latency somatosensory evoked potentials, and electroretinogram were normal. Auditory brain responses revealed only wave I. Serum alkaline phosphatase (ALP) levels were 436 U/L (normal range, 145–420),[10] and calcium and phosphate levels were normal. Metabolic analysis including lactate, pyruvate, very long fatty acids, and organic acid showed no abnormalities. His epileptic attacks sometimes led to generalized tonic–clonic seizures. Ictal EEG showed rhythmic fast waves, which appeared at the left side of the central sulcus, followed by diffuse irregular spikes and waves. Phenytoin and bromide treatment slightly decreased the seizure frequency. He was often admitted to the hospital (>40 times) with respiratory insufficiency following upper respiratory tract infection and/or prolonged convulsions, and initiated home oxygen therapy at 2 years of age.

Swallowing and hand movement gradually deteriorated, and spastic quadriplegia and hypertonus with rigidity of both upper and lower limbs appeared at 4 years of age. At 6 years of age, his condition gradually deteriorated, and a brain MRI at 6 years of age revealed diffuse cerebral and cerebellar atrophy (Fig. 1C,D). ALP was slightly elevated at around 10 years of age (900 U/L [normal range 130–560]), followed by a gradual decrease at around the age of 19 (300 U/L [normal range 65–260]). At this time he required mechanical ventilation. He had a very severe intellectual disability and partial seizures with dyspnea every day, despite administration of phenytoin, valproic acid, phenobarbital, bromide, clobazam, and nitrazepam. Pyridoxine has not been administered.

Case report 2

Individual II-2, the younger sister of individual II-1, was born without asphyxia. She did not show any facial dysmorphology or other congenital malformations. At 7 months of age, she developed generalized tonic–clonic seizures for which she was administered phenobarbital. At 1 year of age, she showed developmental delay with no head control. At this time, she was admitted to hospital due to epileptic convulsive status, and she died from multiorgan failure 3 days later. No autopsy was performed.

Identification of PIGO mutations and flow cytometry analysis

We filtered out variants registered in dbSNP135 data and our in-house 91 control exomes, and narrowed down 193 rare protein-altering and splice-site variants (Table S2). Among them, we identified compound heterozygous mutations in two genes: PIGO (GenBank accession number NM_032634.3) and SCUBE1 (NM_173050.3) (Table S3). No homozygous mutation was detected. Four mutations are rare, but one of two mutations in SCUBE1 is predicted as a polymorphism by web-prediction tools (Table S3). Therefore, PIGO mutations are the primary candidates. Two PIGO mutations were also found in his sister (individual II-2). A novel missense mutation c.389C>A (p.Thr130Asn) in exon 1 was inherited from their father and a novel nonsense mutation c.1288C>T (p.Gln430*) in exon 6 was inherited from their mother. Surface expressions of CD16 and CD24 on granulocytes from the individual II-1 were slightly, but clearly, decreased compared with a normal control and his mother, demonstrating GPI-anchor deficiencies in the patient (Fig. S1).


In this study, we report two siblings with severe epileptic seizures, developmental delay, and mild elevation of ALP caused by two novel compound heterozygous mutations in PIGO. In individuals II-1 and II-2 of the present study, the p.Thr130Asn mutation in PIGO is located in an alkaline phosphatase–like core domain, whereas the p.Gln430* mutation is expected to produce a truncated protein that lacks most transmembrane domains (Fig. 2B). To date, only three families with HPRMS are reported in association with compound heterozygous PIGO mutations: p.Leu957Phe and p.Thr788Hisfs*5 in the first family, p.Leu957Phe and c.3069+5G>A skipping exon 9 leading to c.2855_3069del (p.Val952Aspfs*24) in the second,[5] and c.355C>T (p.Arg119Trp) and c.2497_2498del (p.Ala834Cysfs*131) in the third.[8] These five mutations led to markedly decreased expression of CD16, CD24, and CD59 on granulocytes from the patient or failed to recover expression of GPI-anchored proteins in PIGO-deficient CHO cells, suggesting that expression of GPI-anchored proteins was severely impaired in the patients.[5, 8] On the other hand, individual II-1 with p.Thr130Asn and p.Gln430* mutations showed mildly decreased expression of CD16 and CD24 on the surface of blood granulocytes. This difference in the expression of GPI-anchored proteins might be associated with lacking characteristic features of HPMRS in individual II-1, such as facial dysmorphic features, hypoplasia of distal phalanges, and elevation of serum ALP. Of interest, both patients in our report and a patient reported by Kuki et al. possessed missense mutations commonly in an alkaline phosphatase–like core domain, and showed progressive cerebral and cerebellar atrophy, and more severe intractable epilepsy and developmental delay than the other two families with PIGO mutations reported by Krawitz et al.[5, 8] This fact raised a possibility that mutations in the alkaline phosphatase–like core domain can affect brain development and function more specifically regardless of expression of GPI-anchored proteins in blood granulocytes. Further accumulation of patients with PIGO mutations and functional analysis using neuronal cells are required for elucidating phenotype–genotype correlations in association with PIGO mutations.

Figure 2.

(A) Familial pedigree of individuals 1 (II-1) and 2 (II-2). (B) Distribution of PIGO mutations. Previously reported mutations are highlighted in red. (C) Individuals II-1 and II-2 carrying compound heterozygous mutations in PIGO. Their mother (I-1) carried c.1288C>T (p.Gln430*), and their father (I-2) carried c.389C>A (p.Thr130Asn).

Our data expand the clinical spectrum of GPI-anchor deficiencies to include epileptic encephalopathy. In addition, it has been recently reported that mutations in the SLC35A2 encoding UDP-galactose transporter cause a congenital disorder of glycosylation in three patients, and five of them showed seizures with hypsarrhythmia pattern on electroencephalography.[11, 12] Therefore, it is likely that abnormalities in glycosylation, including the GPI pathway, may be one of the underlying defects in epileptic encephalopathy.

In conclusion, we have described two siblings with epileptic encephalopathy that harbor novel compound heterozygous mutations in PIGO. Further genetic analysis of GPI-anchor synthesis pathway is needed for the understanding of epileptic encephalopathy.


We would like to thank the patients and their families for their participation in this study. We thank Aya Narita and Nobuko Watanabe for technical assistance. This work was supported by the Ministry of Health, Labour and Welfare of Japan; a Grant-in-Aid for Scientific Research (A), (B), and (C) from the Japan Society for the Promotion of Science (A: 24249019, B: 25293085 25293235, C: 23590363); the Takeda Science Foundation; the Japan Science and Technology Agency; the Strategic Research Program for Brain Sciences (11105137); and a Grant-in-Aid for Scientific Research on Innovative Areas (Transcription Cycle, Exploring molecular basis for brain diseases based on personal genomics) from the Ministry of Education, Culture, Sports, Science and Technology of Japan (12024421, 25129705).


We confirm that we have read the Journal's position on issues involved in ethical publication and affirm that this report is consistent with those guidelines. None of the authors has any conflicts of interest to disclose.


  • Image of creator

    Kazuyuki Nakamura is a pediatric neurologist, and researches for epilepsy and brain malformation.