Recessive mutations in ATP8A2 cause severe hypotonia, cognitive impairment, hyperkinetic movement disorders and progressive optic atrophy

Eleven patients from nine families were recruited to participate in this study. The Research Ethics Board of the Hospital for Sick Children approved this study and informed consent was obtained from all families according to the Declaration of Helsinki. The family of Patient 1 consented to a skin biopsy for functional studies to be performed.

Three unrelated patients (Patients 1, 2 and 5) had exome sequencing completed at GeneDx (Gaithersburg, MD). Genomic DNA was extracted from whole blood from affected children and their parents. Exome sequencing was performed on exon targets captured using either the Agilent SureSelect Human All Exon (50 Mb) V4 kit or Clinical Research Exome kit (Agilent Technologies, Santa Clara, CA). One microgram of DNA from peripheral blood was sonicated into 300 bp fragments, which were then repaired, ligated to adaptors, and purified for subsequent PCR amplification. Amplified products were then captured by biotinylated RNA library baits in solution following the manufacturer’s instructions. Bound DNA was isolated with streptavidin-coated beads and re-amplified. The final isolated products were sequenced using either the Illumina HiSeq 2500 or 4000 sequencing system with either 2 × 100-bp or 2 × 150-bp paired-end reads (Illumina, San Diego, CA). DNA sequence was mapped to the published human genome build UCSC hg19/GRCh37 reference sequence using BWA-Mem v0.7.8 [11]. Targeted coding exons and splice junctions of known protein-coding RefSeq genes were assessed for average depth of coverage with a minimum depth of 10X required for inclusion in downstream analysis. Local realignment around insertion-deletion sites was performed using the Genome Analysis Toolkit v2.3 [12]. Variant calls were generated simultaneously on all sequenced family members using either Samtools 0.1.18 or Samtools 0.1.18 along with GATK 2.3 HaplotypeCaller [11, 12]. All coding exons and surrounding intron/exon boundaries were analyzed. CNVs were called as previously described [13]. Automated filtering removed common sequence changes (defined as > 10% frequency present in the 1000 Genomes database). The targeted coding exons and splice junctions of the known protein-coding RefSeq genes were assessed for the average depth of coverage and data quality threshold values. Whole exome sequence data for all sequenced family members was analyzed using GeneDx’s XomeAnalyzer (a variant annotation, filtering, and viewing interface for WES data), which includes nucleotide and amino acid annotations, population frequencies (NHLBI Exome Variant Server and 1000 Genomes databases), in silico prediction tools and amino acid conservation scores (Mutation Taster, PhyloP, and CADD). Variants were filtered based on inheritance patterns, gene lists of interest, phenotype, and population frequencies, as appropriate. Resources including the Human Gene Mutation Database (HGMD), 1000 Genomes database, NHLBI Exome Variant Server, OMIM, PubMed, and ClinVar were used to evaluate genes and sequence changes of interest [14‐17]. Identified sequence changes of interest were confirmed in all family members by di-deoxy Sanger sequence analysis using an ABI3730 (Life Technologies, Carlsbad, CA) and standard protocols with a new DNA preparation. CNVs were confirmed in relevant family members by whole genome or exon-focused oligonucleotide array-based comparative genomic hybridization (Agilent Technologies, Santa Clara, CA), quantitative polymerase chain reaction, or multiplex ligation-dependent amplification. Patient 3 underwent exome sequencing at BGI, Denmark with subsequent annotation and interpretation at the Genome Diagnostics laboratory in Nijmegen, the Netherlands. Libraries were prepared using Agilent SureSelect Human All Exon enrichment kit version 5 (Agilent Technologies) and sequenced on an Illumina HiSeq4000 2x150bp. Reads were aligned to hg19 reference genome using BWA − 0.7.8-r455, variants were called with GenomeAnalysisTK 3.3–0-g37228af and annotated using an in-house pipeline. Exome sequencing for Patient 4 was carried out in the molecular laboratory of the Hôpital Pitié-Salpêtrière in Paris, France. Libraries were prepared from genomic DNA using Roche SeqCap MedExome kits and sequenced on an Illumina NextSeq 500 2x150bp high output (with 12 plexes). Reads were aligned to hg19 reference genome using BWA-mem, variants were called with GenomeAnalysisTK-2014.3-17-g0583018 and annotated using SNPEff-4. Custom scripts were utilized for variant filtration and prioritization. Exome sequencing for Patients 6 and 7 was carried out at Novogene using the Agilent SureSelect V6 enrichment kit with a paired-end (150 bp) protocol at a mean coverage of 50X. Reads were aligned to genome assembly hg19 with the Burrows-Wheeler Aligner (BWA) (Version 0.7.8-r455). Exome sequencing for Patients 8 and 9 was carried out at the Institute for Molecular Bioscience in Queensland, Australia. Libraries were prepared from genomic DNA using Nextera Rapid Capture Exome kits and sequenced on an Illumina HiSeq 2000 to a minimum average coverage of 95X. Reads were aligned to hg19 reference genome using BWA-mem, variants were called with GATK HaplotypeCaller v3.7 and annotated using SnpEff v4.3 m. Custom scripts were utilized for variant filtration and prioritization. Patients 10 and 11 were diagnosed at 12 de Octubre Hospital in Madrid, Spain and have been previously reported [9]. Detailed exome sequencing methods are outlined in Additional file 1: Table S1.

The clinical features of individuals with ATP8A2 mutations are described for 11 patients ranging from 2.5 years to 28 years (Table 1). Patients were of varied ethnicities with 8/11 (73%) demonstrating parental consanguinity. All patients exhibited severe hypotonia at or within 6 months after birth which persisted, with older children and adults being unable to achieve head control and/or sit independently (Table 2). Expressive and receptive language skills were impaired in all patients. Most children (9/11, 82%) did not develop any meaningful speech, with only two communicating with mono or di-syllabic words, or with the aid of pictograms (Table 2). Hyperkinetic movement disorders, specifically chorea or choreoathetosis were present in all patients (Additional file 2: Video S1; Additional file 3: Video S2). Dystonia and/or facial dyskinesia were noted in a smaller proportion of patients (Additional file 4: Video S3). Feeding difficulties were a significant feature of our cohort and were reported in 10/11 (91%) individuals. All 10 patients required modified feeds such as pureed or thickened feeds, and all required significant assistance with feeding. Three patients (Patients 1, 2 and 9) required gastrostomy tubes due to their risk of aspiration when drinking thin liquids. Weight was below the third percentile for 70% (7/10) of the patients who had a recent weight available. Occipitofrontal head circumference (OFC) was within the normal range for the majority of patients, and 5/11 (45%) had OFC at or below the second percentile for age. Optic atrophy was noted in 7/9 (78%) patients who underwent funduscopic examination. The two children who did not have optic atrophy did not have testing of visual evoked potentials. In one child (Patient 1) who underwent repeat funduscopic examinations, optic atrophy was noted to progress in the first few years of life. Initial ophthalmology examination at 10 months was entirely normal including funduscopic examination, and the patient had normal visual fixation and tracking. Repeat examination at 19 months revealed mild optic nerve pallor. At 26 months she was no longer able to visually fix or follow and optic nerve atrophy was seen. Visual evoked potentials at 3 years revealed absentcortical responses. At 4.5 years she had severe optic nerve atrophy (Fig. 1) with optical coherence tomography (OCT) showing selective thinning of the inner retinal layers. A smaller proportion of patients exhibited ophthalmoplegia (45%) or ptosis (36%) (Additional file 5: Video S4). Seizures were observed in two siblings from a single family, with the remaining 9/11 patients being seizure-free. Patient 8 did not require treatment with anti-epileptics and Patient 9 had seizures which were well-controlled with valproate and carbamazepine. Neither patient had EEG studies available for review. One patient was noted to have abnormal brainstem auditory evoked potentials (BAER) responses (Patient 4) consistent with sensorineural hearing loss. This patient also had abnormal somatosensory evoked potentials (SSEP) suggesting a myelinopathy involving the dorsal column of the spinal cord. One other patient (Patient 5) was noted to have abnormal hearing.

Nerve conduction studies and electromyography was performed on 8/11 patients with no evidence of a sensorimotor neuropathy observed (Table 3). Single fiber EMG was normal in one patient who had ptosis. Neuroimaging was performed in all but one patient. MRI of the brain revealed no intracranial abnormalities in 50% (5/10) patients (Table 3). Non-specific findings included mild cerebral or cortical atrophy (30%), mild delay in myelination (20%), thin corpus callosum (20%) or hypoplastic optic nerves (20%). Additional detailed clinical information is provided for Patients 1 and 3 as Additional file 6.

Biallelic ATP8A2 gene mutations were identified in all affected individuals, consistent with autosomal recessive inheritance (Table 4). Carrier parents were asymptomatic. Identified genetic variants included missense, splicing, intragenic deletions, and frameshift mutations, which were predicted to cause loss of function of ATP8A2, and were not present in the homozygous state in the Exome Aggregation Consortium (ExAC) public database or our internal exome databases of unaffected individuals. All variants occurred within highly conserved domains, and were predicted to be damaging or possibly damaging by multiple in silico models (SIFT, Polyphen, MetaSVM, MetaSVM, Mutation Taster, PhyloP, and CADD).

Initial studies were unable to detect any ATP8A2 in the control or fibroblasts from Patient 1, due to low ATP8A2 expression in this tissue. Subsequent reprogramming of the fibroblasts into induced pluripotent cells allowed derivation of tissues that expressed higher amounts of ATP8A2 to enable study. We found that early endoderm derivatives expressed high levels of ATP8A2 and allowed analyses of RNA expression and ATP8A2 protein expression. To ensure that the patient’s cells could differentiate into endoderm with equivalent efficiency as control cells, we assessed expression of two endoderm markers, FOXA2 and SOX17. Using expression of these markers, the patient’s cells more readily adopted the endoderm fate compared to control cells (Supplementary Figure S1). ATP8A2 mRNA expression was significantly lower in the patient compared to control cells and there was diminished ATP8A2 protein in the patient cells, confirming the deleterious effect of the mutations (Fig. 2). From these studies we conclude this patient had loss of function mutations in ATP8A2, which likely improved endoderm differentiation.