Abstract
Import of peroxisomal matrix proteins, crucial for peroxisome biogenesis, is mediated by the cytosolic receptors PEX5 and PEX7 that recognize proteins carrying peroxisomal targeting signals 1 or 2 (PTS1 or PTS2), respectively. Mutations in PEX5 or 12 other PEX genes cause peroxisome biogenesis disorders, collectively named the Zellweger spectrum disorders (ZSDs), whereas mutations in PEX7 cause rhizomelic chondrodysplasia punctata type 1 (RCDP1). Three additional RCDP types, RCDP2–3–4, are caused, respectively, by mutations in GNPAT, AGPS and FAR1, encoding enzymes involved in plasmalogen biosynthesis. Here we report a fifth type of RCDP (RCDP5) caused by a novel mutation in PEX5. In four patients with RCDP from two independent families, we identified a homozygous frame shift mutation c.722dupA (p.Val242Glyfs∗33) in PEX5 (GenBank: NM_001131023.1). PEX5 encodes two isoforms, PEX5L and PEX5S, and we show that the c.722dupA mutation, located in the PEX5L-specific exon 9, results in loss of PEX5L only. Both PEX5 isoforms recognize PTS1-tagged proteins, but PEX5L is also a co-receptor for PTS2-tagged proteins. Previous patients with PEX5 mutations had ZSD, mainly due to deficient import of PTS1-tagged proteins. Similarly to mutations in PEX7, loss of PEX5L results in deficient import of PTS2-tagged proteins only, thus causing RCDP instead of ZSD. We demonstrate that PEX5L expression restores the import of PTS2-tagged proteins in patient fibroblasts. Due to the biochemical overlap between RCDP1 and RCDP5, sequencing of PEX7 and exon 9 in PEX5 should be performed in patients with a selective defect in the import of PTS2-tagged proteins.
Introduction
Peroxisomes are subcellular organelles that perform several metabolic functions, e.g. fatty acid α- and β-oxidation and biosynthesis of plasmalogens, a class of membrane glycerophospoholipids (1). Peroxisomal import of matrix proteins, crucial for peroxisome biogenesis, is initiated when proteins carrying peroxisomal targeting signals 1 or 2 (PTS1 or PTS2) are recognized by their cytosolic receptors, encoded by PEX5 (peroxisome biogenesis factor 5) or PEX7 (peroxisome biogenesis factor 7) (2).
Mutations in PEX5 and 12 other PEX genes result in peroxisome biogenesis disorder (PBD) group A (1), the Zellweger spectrum disorders (ZSDs), including Zellweger syndrome (ZS; MIM 214100) and infantile Refsum disease (MIM 202370), while mutations in PEX7 cause PBD group B (1), rhizomelic chondrodysplasia punctata type 1 (RCDP1; MIM 215100) (3–5). Two additional RCDP types, RCDP2 (MIM 222765) and RCDP3 (MIM 600121), caused by mutations in GNPAT (glyceronephosphate O-acyltransferase) (6) and AGPS (alkylglycerone phosphate synthase) (7), are also disorders of peroxisome function, specifically plasmalogen biosynthesis (1). FAR1 encodes fatty acyl-CoA reductase 1, producing fatty alcohols used in plasmalogen biosynthesis (8). Mutations in FAR1 were recently shown to cause peroxisomal fatty acyl-CoA reductase 1 disorder (MIM 616154) (9), also referred to as RCDP4 (10).
The most severe form of the ZSDs is characterized by muscular hypotonia, poor feeding, distinctive facial dysmorphisms, global developmental delay and seizures, hepatic cysts and dysfunction, and usually death in the first year of life (11,12), whereas patients with the milder forms of ZSD may have developmental delays, hearing loss, vision impairment, liver dysfunction and episodes of hemorrhage, with prolonged survival possibly into adulthood (13). Similarly, the severe form of RCDP is characterized by rhizomelia, chondrodysplasia punctata, congenital cataracts, profound growth retardation and intellectual disability, seizures, and death within the first decade of life (14–16), while some individuals lack rhizomelia, show less profound growth and developmental delays, and have longer survival (14–16).
Here we describe four patients from two independent families carrying the same novel homozygous frame shift mutation c.722dupA (p.Val242Glyfs∗33) in PEX5 (GenBank: NM_001131023.1). We show that the c.722dupA, located in the PEX5 long (PEX5L) isoform-specific exon 9, affects the synthesis of PEX5L only, resulting in a peroxisomal dysfunction due to selective defect in the import of PTS2-tagged proteins, and causing RCDP instead of ZSD. Mutations in PEX5-affecting PEX5L only are a novel genetic cause of RCDP, suggested by us to be designated RCDP type 5 (RCDP5).
Results
Clinical description
Family A has three affected siblings of first cousin parents from Jammu and Kashmir in Pakistan (Fig. 1A). The oldest affected female (FA.II-1; Fig. 1A–D) was born at term with normal growth parameters. She had congenital cataracts and was operated at 5 months. After the age of 18–24 months, she showed growth delay and microcephaly. Her cognitive development, assessed at 7, 16 and 27 years of age, was consistent with severe intellectual disability (IQ 20–34, Supplementary Material, Table S1). Hearing assessed at age 7 was normal. At the age of 17 years, she developed seizures, and EEG revealed epileptiform activity. Radiological examination at age 27 showed mild shortening of the humeri (3.6 cm < the 10th centile) (17), short femoral necks, narrow ilii and coxa vara, in addition to slight thoracic lordosis and scoliosis and short pedicles. At the last clinical examination at 27 years, her height was 10 cm below the third centile, weight for height was at the 10th centile, and her occipitofrontal circumference (OFC) was at the 2.5th centile. She was ambulatory with a broad-based gait. There were signs of peripheral neuropathy with areflexia, generalized muscle atrophy in the extremities, contractures of the large joints, swan neck deformity of the hands, pes cavus and pes equinus. Neurophysiological examination suggested pronounced, demyelinating motor and sensory neuropathy with nerve conduction velocities between 12 and 20 m/s (normal >50 m/s). She had mild palmoplantar keratodermia, but no ichthyosis. Cardiac ultrasound was normal, but a 24 h EKG investigation registered episodes of sinus tachycardia. Abdominal ultrasound did not reveal organ pathology. She wore aphakic spectacles (right eye +18 and left eye +16). Examination of the fundi did not indicate retinitis pigmentosa. She had a pleasant personality, and performed simple tasks at a work centre.
Pedigrees and pictures of affected individuals. (A) Pedigree of family A (FA), (B–D) FA.II-1 at 27 years of age, (B) note bushy eyebrows and thick spectacle prisms after removal of cataracts, (C) swan neck deformity of the hands, (D) and distal muscle wasting with pes equinus deformity, compatible with chronic demyelinating neuropathy. (E–H) FA.II-6 at 9.5 years of age (E, F and H) and 10.5 years of age (G), (E) note similar bushy eyebrows and thick spectacle prisms after removal of cataracts as in his elder sister, (F) similar signs of neuropathy with swan neck deformity of the hands, (G) and muscle wasting in the lower extremities, but without foot deformity, (H) and shortening of the extremities and multiple joint contractures (elbow, hip and knee), thoracic kyphosis and scoliosis. (I) Pedigree of family B (FB), (J) chromatogram of the Sanger sequence showing the heterozygous c.722dupA mutation in PEX5 in FA.I-1 and FA.I-2 (top), and the homozygous c.722dupA mutation in FA-II-1, FA.II-3, FA.II-6 and FB.II-1 (bottom). fs, frame shift.
The second affected female (FA.II-3; Fig. 1A) was born with normal growth parameters. Congenital cataracts were operated at 4 months, followed by several operations due to secondary cataracts. Her vision remained severely impaired. After the age of 18 months, she showed growth delay and microcephaly, with height and OFC below the third centiles. She had severe bronchial asthma from the first year of life. A gastric-feeding tube was placed at 4 years. At the same age, she had a vocabulary of three single words and could not sit unaided. Her psychomotor development was regarded as profoundly delayed. Hearing was not formally assessed, but she reacted to sounds and voices. At 8 years peripheral neuropathy was suspected due to muscular atrophy, contractures and hyporeflexia. Walking was limited to short distances with support. She had vertical and horizontal nystagmus. Radiological examination at 5 years revealed punctate calcifications around the hips, mild shortening of the humeri (<10th centile) (17), irregular metaphyses, small and irregular femoral capital epiphyses, narrow ilii and coxa vara, and at 8 years slight s-shaped scoliosis and compression fractures. At age 8, she developed epilepsy, and two EEG examinations showed epileptiform activity. From 9 years, she had recurrent respiratory infections, and she died in association with non-convulsive status epilepticus and pneumonia at age 13.
The youngest child, an affected male (FA.II-6; Fig. 1A, E–H), was born at term with normal birth parameters. He had congenital cataracts, operated at 2 months. After 2 years, he showed growth delay and microcephaly. Similar to FA.II-3 he developed bronchial asthma, complicated by frequent airway infections. At 3 years, he had a vocabulary of ∼20 words and walked a few steps independently. His cognitive level, assessed at 6.5 and 10.5 years of age, was consistent with severe intellectual disability (IQ 20–34; Supplementary Material, Table S1). He mostly spoke single worlds, but could utter some short sentences. Cerebral MRI examination at 7 years was normal. Radiological examinations showed normal length of long bones at age 8 months, but mild shortening of the humeri (<10th centile) (17) and metaphyseal cupping at 6 years. Radiographs and additional description from skeletal surveys are provided in Fig. 2A–D. At the last clinical visit at 10.5 years his height was 19 cm below the 3rd centile, weight for length was at the 50th centile and his OFC was 1.5 cm < the 2.5th centile. He had independent ambulation. Peripheral neuropathy was suspected due to hyporeflexia, muscular weakness and muscular atrophy. Neurophysiological examination was incomplete due to lack of cooperation, but indicated delay of F-responses. He recently developed seizures, and an EEG showed epileptiform activity. Dermatological examination at age 10 years was normal. Abdominal ultrasound at the same age did not reveal organ pathology. He wore aphakic spectacles (right eye +16 and left eye +14.5). A recent electroretinography examination did not indicate retinitis pigmentosa. He had a positive and pleasant behaviour and was attending a special education class, where he had friends and cooperated well.
Radiographs of patient FA.II-6. (A) Radiograph of the knees at 8 months showing extensive punctate calcification (indicated by arrow). (B) Pelvis at 6 years showing narrow ilii, small and fragmented capital femoral epiphyses (indicated by arrow) with subtle calcifications. (C and D) Frontal and lateral view of the spine at 11 years showing slight s-shaped scoliosis, osteopenia and anisospondyly.
Family B has a single affected female (FB.II-1; Fig. 1I) born to first cousin parents from Punjab in Pakistan. She was born after an uncomplicated pregnancy with normal weight and OFC. She had congenital cataracts, which were operated at 6 months of age with replacement of intraocular lenses. She showed postnatal growth delay and microcephaly. At 3 years her developmental age corresponded to 12–18 months, and she walked independently from 4 years. Cerebral MRI examination at age 3 was normal. At the last clinical examination at 6.5 years, her height and OFC were both below the 0.4th centile, and weight for height was at the 25th centile. She was ambulant with no contractures, but had raised tone at the ankles. Reflexes were normal and she did not have abnormalities of foot or hand posture to indicate a peripheral neuropathy. At clinical assessment, her impressive and expressive language skills were judged as corresponding to ∼12 months, and her gross and fine motor skills to ∼18 months. Radiological examination of the spine and left radius and ulna at the age of 7 years was normal. She required help to meet all her care needs.
Identification of a novel variant in the PEX5L-specific exon 9 in PEX5
Whole-exome sequencing of DNA from individuals FA.I-1, FA.II-1 and FA.II-6 was performed (sequencing statistics are provided in Supplementary Material, Table S2). The genetic variants were filtered using FILTUS (http://folk.uio.no/magnusv/filtus.html) assuming recessive inheritance due to multiple affected siblings. Variants were removed when they occurred with a frequency >0.01 in the 1000 Genomes Project (http://browser.1000genomes.org), the NHLBI Exome Sequencing Project Exome Variant Server (http://evs.gs.washington.edu/EVS), or in our in-house database of 268 exomes, or were predicted to have low impact according to snpEff (18), SIFT (19) or PolyPhen-2 (20). This resulted in homozygous variants in five candidate genes (Supplementary Material, Table S3). Among these, the homozygous c.722dupA (p.Val242Glyfs∗33) in PEX5 (GenBank: NM_001131023.1) was considered to be likely disease-causing because of the established role of PEX5 in peroxisome biogenesis. PEX5 encodes a long (PEX5L) and a short (PEX5S) isoform, distinguished by the 111 bp exon 9 (coding exon 7) present in PEX5L only (21,22). The novel c.722dupA variant in PEX5 is an insertion of one nucleotide after genomic position chr12 : 7354871, hg19. It is located in exon 9, thus resulting in a frame shift affecting PEX5L only, starting with p.Val242Gly and introducing a stop codon 32 codons downstream. The variant was not reported in the Exome Aggregation Consortium Browser (http://exac.broadinstitute.org/) containing >60 000 exomes (accessed 28 April 2015). Sanger sequencing confirmed that the affected siblings FA.II-1, FA.II-3 and FA.II-6 were homozygous for the c.722dupA variant in PEX5, and both parents were heterozygous (Fig. 1A and J; Supplementary Material, Table S4). The c.722dupA (p.Val242Glyfs∗33) variant in PEX5 was submitted to the ClinVar database (http://www.ncbi.nlm.nih.gov/clinvar), accession number SCV000222895.
FB.II-1 was independently referred for metabolic analyses due to suspicion of a peroxisomal disorder. The results of the studies of peroxisomal parameters in cultured fibroblasts indicated a PTS2 protein import defect and thus PEX7 sequencing was performed, followed by PEX5 sequencing. This revealed no pathogenic mutations in PEX7, but identified the same homozygous c.722dupA (p.Val242Glyfs∗33) variant in PEX5 (Fig. 1I and J).
The novel variant in PEX5 results in selective loss of PEX5L
We assessed the predicted loss of PEX5L in fibroblasts from the patients. Quantitative real-time PCR (qRT-PCR) showed that the mean PEX5L transcript level in fibroblasts from FA.II-3 and FA.II-6 was 25% of the level in controls (Fig. 3A), indicating that it was subjected to nonsense-mediated decay. Furthermore, immunoblot analysis using an antibody which recognizes both PEX5 isoforms detected two bands, corresponding to PEX5L and PEX5S in control fibroblasts, whereas only the single band corresponding to PEX5S was detected in fibroblasts from FA.II-1, FA.II-3 and FB.II-1 (Fig. 3B). No bands corresponding to the two PEX5 isoforms were present in the PEX5 deficient human fibroblast cell line, PEX5.11 (Fig. 3B). This confirmed that the novel c.722dupA variant in PEX5 leads to a selective loss of the PEX5L isoform.
Loss of PEX5L in patient fibroblasts. (A) Relative quantification of the PEX5L transcript level in fibroblasts from patients FA.II-3 and FA.II-6 using two independent primer sets amplifying PEX5L only (PEX5L-1 and PEX5L-2) showed a significant reduction in the mean level compared with five controls (P < 0.005). Data from a representative experiment are presented as the mean PEX5L level in patients (n = 2) and controls (n = 5). *P < 0.005. Error bars represent the 95% confidence intervals. (B) Immunoblot analysis using an antibody binding to both PEX5L and PEX5S isoforms showing the band corresponding to PEX5L in HEK293T cells transfected with PEX5L (PEX5L) and the band corresponding to PEX5S in HEK293T cells transfected with PEX5S (PEX5S). Both isoforms are detected in fibroblasts from controls (C1 and C2), while only the band corresponding to PEX5S was detected in the three patients FA.II-3, FA.II-1 and FB.II-1. The PEX5 deficient fibroblast cell line PEX5.11 lacked bands for both PEX5 isoforms. Two different exposure times are shown. Equal loading was confirmed by Ponceau-S staining.
Loss of PEX5L results in a selective peroxisomal import defect of PTS2-tagged proteins
In order to characterize the peroxisomal dysfunction resulting from the selective loss of PEX5L, biochemical parameters relevant for import of PTS1- and PTS2-tagged proteins were measured in blood from the three living patients (FA.II-1, FA.II-6 and FB.II-1), and in fibroblasts from all four patients (Tables 1 and 2). Most peroxisomal matrix proteins are targeted to the organelle by a PTS1 (2), including several enzymes involved in peroxisomal fatty acid β-oxidation (23). Accumulation of very long chain fatty acids (VLCFAs) is the biochemical hallmark in ZSDs (24,25), whereas levels are normal in RCDP (26). PTS1-mediated peroxisomal protein import was evaluated by measuring the amount of VLCFAs in plasma, and the peroxisomal β-oxidation capacity in fibroblasts using C26 : 0 and pristanic acid as substrates. The results were within reference ranges (Table 1), except for a mildly elevated C26/C22 ratio in individual FB.II-1 in plasma (Table 1), but not in fibroblasts (0.05, ref. 0.03–0.07), confirming normal import of the PTS1-tagged enzymes involved in peroxisomal β-oxidation. PTS1-targeted protein import was further evaluated by immunoblot analysis using an antibody against the first enzyme of the peroxisomal β-oxidation system, acyl-CoA oxidase (ACOX1). This revealed normal intra-peroxisomal processing of ACOX1 in fibroblasts from all four patients similar to the control fibroblasts, indicating normal import of ACOX1 in these cell lines (Fig. 4A). In contrast, a ZSD human fibroblast cell line only showed unprocessed ACOX1 (Fig. 4A). Furthermore, immunofluorescence microscopy analysis performed using an antibody against PTS1-tagged catalase, a commonly used marker for PTS1 protein import (27). The results showed peroxisomal staining in control cells (Fig. 4B) as well as in fibroblasts from all patients (Table 1, Fig. 4C), indicating normal peroxisomal import of PTS1-tagged catalase.
Normal peroxisomal import of PTS1-tagged proteins in patients with the PEX5 c.722dupA
| VLCFAs (PLA) (µmol/l) | β-Oxidation (FBRO) (pmol)/(h × mg) protein | IB ACOX1 (FBRO) peroxisomal processing | IF catalase (FBRO) peroxisomal staining | |||
|---|---|---|---|---|---|---|
| C26 | C26 : 0/C22 : 0 | C26 : 0 | Pristanic acid | |||
| FA.II-1 | 0.92 | 0.02 | 983 | 1010 | + | + |
| FA.II-3 | ND | ND | 830 | 1627 | + | + |
| FA.II-6 | 0.57 | 0.01 | 815 | 1098 | + | + |
| FB.II-1 | 1.3, 1.45a | 0.037, 0.045a | ND | ND | + | + |
| Referenceb | 0.45–1.32 | 0.00–0.02 | 800–2040 | 790–1690 | + | + |
| ZS | 1.8–8.1 | 0.069–0.453 | 50–350 | 0–30 | − | − |
| VLCFAs (PLA) (µmol/l) | β-Oxidation (FBRO) (pmol)/(h × mg) protein | IB ACOX1 (FBRO) peroxisomal processing | IF catalase (FBRO) peroxisomal staining | |||
|---|---|---|---|---|---|---|
| C26 | C26 : 0/C22 : 0 | C26 : 0 | Pristanic acid | |||
| FA.II-1 | 0.92 | 0.02 | 983 | 1010 | + | + |
| FA.II-3 | ND | ND | 830 | 1627 | + | + |
| FA.II-6 | 0.57 | 0.01 | 815 | 1098 | + | + |
| FB.II-1 | 1.3, 1.45a | 0.037, 0.045a | ND | ND | + | + |
| Referenceb | 0.45–1.32 | 0.00–0.02 | 800–2040 | 790–1690 | + | + |
| ZS | 1.8–8.1 | 0.069–0.453 | 50–350 | 0–30 | − | − |
Reference values are from Laboratory Genetic Metabolic Diseases, Amsterdam, The Netherlands.
aReference values from Laboratory Medicine, Sheffield Children's Hospital, Sheffield, UK: C26 : 0, 0.33–1.5 µmol/l; C26/C22, 0.005–0.030. The two measurements were performed 1 month apart at age 5.5 years.
bThe values for RCDP1-3 are the same as for the reference. FBRO, fibroblasts; IB, immunoblot; IF, immunofluorescence; ND, not done; peroxisomal processing, cleavage of the 72 kDa precursor to the 50 kDa protein, PLA, plasma; VLCFAs, very long chain fatty acids; ZS, Zellweger syndrome.
Normal peroxisomal import of PTS1-tagged proteins in patients with the PEX5 c.722dupA
| VLCFAs (PLA) (µmol/l) | β-Oxidation (FBRO) (pmol)/(h × mg) protein | IB ACOX1 (FBRO) peroxisomal processing | IF catalase (FBRO) peroxisomal staining | |||
|---|---|---|---|---|---|---|
| C26 | C26 : 0/C22 : 0 | C26 : 0 | Pristanic acid | |||
| FA.II-1 | 0.92 | 0.02 | 983 | 1010 | + | + |
| FA.II-3 | ND | ND | 830 | 1627 | + | + |
| FA.II-6 | 0.57 | 0.01 | 815 | 1098 | + | + |
| FB.II-1 | 1.3, 1.45a | 0.037, 0.045a | ND | ND | + | + |
| Referenceb | 0.45–1.32 | 0.00–0.02 | 800–2040 | 790–1690 | + | + |
| ZS | 1.8–8.1 | 0.069–0.453 | 50–350 | 0–30 | − | − |
| VLCFAs (PLA) (µmol/l) | β-Oxidation (FBRO) (pmol)/(h × mg) protein | IB ACOX1 (FBRO) peroxisomal processing | IF catalase (FBRO) peroxisomal staining | |||
|---|---|---|---|---|---|---|
| C26 | C26 : 0/C22 : 0 | C26 : 0 | Pristanic acid | |||
| FA.II-1 | 0.92 | 0.02 | 983 | 1010 | + | + |
| FA.II-3 | ND | ND | 830 | 1627 | + | + |
| FA.II-6 | 0.57 | 0.01 | 815 | 1098 | + | + |
| FB.II-1 | 1.3, 1.45a | 0.037, 0.045a | ND | ND | + | + |
| Referenceb | 0.45–1.32 | 0.00–0.02 | 800–2040 | 790–1690 | + | + |
| ZS | 1.8–8.1 | 0.069–0.453 | 50–350 | 0–30 | − | − |
Reference values are from Laboratory Genetic Metabolic Diseases, Amsterdam, The Netherlands.
aReference values from Laboratory Medicine, Sheffield Children's Hospital, Sheffield, UK: C26 : 0, 0.33–1.5 µmol/l; C26/C22, 0.005–0.030. The two measurements were performed 1 month apart at age 5.5 years.
bThe values for RCDP1-3 are the same as for the reference. FBRO, fibroblasts; IB, immunoblot; IF, immunofluorescence; ND, not done; peroxisomal processing, cleavage of the 72 kDa precursor to the 50 kDa protein, PLA, plasma; VLCFAs, very long chain fatty acids; ZS, Zellweger syndrome.
Deficient peroxisomal import of PTS2-tagged proteins in patients with the PEX5 c.722dupA
| Plasmalogens (ERY or FBRO) % | GNPAT activity (FBRO) nmol/ (2 h × mg protein) | PLA (µmol/l) | α-Oxidation (FBRO) pmol/(h × mg protein) phytanic acid | ACAA1 (FBRO) | |||
|---|---|---|---|---|---|---|---|
| C16 DMA/C16 : 0 | C18 : 0 DMA/C18 : 0 | IB peroxisomal processing | IF peroxisomal staining | ||||
| FA.II-1 | 2.2 | 7.2 | 3.8 | 427 | 7 | − | − |
| FA.II-3 | ND | ND | 4.3 | ND | 11 | − | − |
| FA.II-6 | 2.5 | 6.7 | 4.3 | 343 | 8 | − | − |
| FB.II-1 | 3.8a | 2.7a | 4.6, 4.8 | 360, 397b | ND | − | − |
| Reference | 6.8–11.9 | 10.6–24.9 | 5.4–10.6 | 0.5–9.9 | 28–95 | + | + |
| RCDP1 | <6.8 | <10.6 | 1.0–4.2 | >9.9c | ≤22 | − | − |
| RCDP2 | <6.8 | <10.6 | <0.1 | 0.5–9.9 | 28–95 | + | + |
| RCDP3 | <6.8 | <10.6 | 1.0–4.2 | 0.5–9.9 | 28–95 | + | + |
| ZS | <6.8 | <10.6 | 0.1–0.9 | 1.6–115.7 | 0–10 | − | − |
| Plasmalogens (ERY or FBRO) % | GNPAT activity (FBRO) nmol/ (2 h × mg protein) | PLA (µmol/l) | α-Oxidation (FBRO) pmol/(h × mg protein) phytanic acid | ACAA1 (FBRO) | |||
|---|---|---|---|---|---|---|---|
| C16 DMA/C16 : 0 | C18 : 0 DMA/C18 : 0 | IB peroxisomal processing | IF peroxisomal staining | ||||
| FA.II-1 | 2.2 | 7.2 | 3.8 | 427 | 7 | − | − |
| FA.II-3 | ND | ND | 4.3 | ND | 11 | − | − |
| FA.II-6 | 2.5 | 6.7 | 4.3 | 343 | 8 | − | − |
| FB.II-1 | 3.8a | 2.7a | 4.6, 4.8 | 360, 397b | ND | − | − |
| Reference | 6.8–11.9 | 10.6–24.9 | 5.4–10.6 | 0.5–9.9 | 28–95 | + | + |
| RCDP1 | <6.8 | <10.6 | 1.0–4.2 | >9.9c | ≤22 | − | − |
| RCDP2 | <6.8 | <10.6 | <0.1 | 0.5–9.9 | 28–95 | + | + |
| RCDP3 | <6.8 | <10.6 | 1.0–4.2 | 0.5–9.9 | 28–95 | + | + |
| ZS | <6.8 | <10.6 | 0.1–0.9 | 1.6–115.7 | 0–10 | − | − |
Reference values are from Laboratory Genetic Metabolic Diseases, Amsterdam, The Netherlands.
aMeasurement in fibroblasts using two controls showing values for C16 : 0 dimethylacetal (DMA)/C16 : 0 of 12.0 and 12.9% and for C18 : 0 DMA/C18 : 0 of 8.9 and 9.9%.
bReference values from Laboratory Medicine, Sheffield Children's Hospital, Sheffield, UK; PLA phytanic acid, <19.3 µmol/l. The two measurements were performed 1 month apart at age 5.5 years.
cCan be normal, increases with age. ERY, erythrocytes; FBRO, fibroblasts; IB, immunoblot; IF, immunofluorescence; ND, not done; PLA, plasma; peroxisomal processing, peroxisomal cleavage of 44 kDa ACAA1 to 41 kDa; ZS, Zellweger syndrome.
Deficient peroxisomal import of PTS2-tagged proteins in patients with the PEX5 c.722dupA
| Plasmalogens (ERY or FBRO) % | GNPAT activity (FBRO) nmol/ (2 h × mg protein) | PLA (µmol/l) | α-Oxidation (FBRO) pmol/(h × mg protein) phytanic acid | ACAA1 (FBRO) | |||
|---|---|---|---|---|---|---|---|
| C16 DMA/C16 : 0 | C18 : 0 DMA/C18 : 0 | IB peroxisomal processing | IF peroxisomal staining | ||||
| FA.II-1 | 2.2 | 7.2 | 3.8 | 427 | 7 | − | − |
| FA.II-3 | ND | ND | 4.3 | ND | 11 | − | − |
| FA.II-6 | 2.5 | 6.7 | 4.3 | 343 | 8 | − | − |
| FB.II-1 | 3.8a | 2.7a | 4.6, 4.8 | 360, 397b | ND | − | − |
| Reference | 6.8–11.9 | 10.6–24.9 | 5.4–10.6 | 0.5–9.9 | 28–95 | + | + |
| RCDP1 | <6.8 | <10.6 | 1.0–4.2 | >9.9c | ≤22 | − | − |
| RCDP2 | <6.8 | <10.6 | <0.1 | 0.5–9.9 | 28–95 | + | + |
| RCDP3 | <6.8 | <10.6 | 1.0–4.2 | 0.5–9.9 | 28–95 | + | + |
| ZS | <6.8 | <10.6 | 0.1–0.9 | 1.6–115.7 | 0–10 | − | − |
| Plasmalogens (ERY or FBRO) % | GNPAT activity (FBRO) nmol/ (2 h × mg protein) | PLA (µmol/l) | α-Oxidation (FBRO) pmol/(h × mg protein) phytanic acid | ACAA1 (FBRO) | |||
|---|---|---|---|---|---|---|---|
| C16 DMA/C16 : 0 | C18 : 0 DMA/C18 : 0 | IB peroxisomal processing | IF peroxisomal staining | ||||
| FA.II-1 | 2.2 | 7.2 | 3.8 | 427 | 7 | − | − |
| FA.II-3 | ND | ND | 4.3 | ND | 11 | − | − |
| FA.II-6 | 2.5 | 6.7 | 4.3 | 343 | 8 | − | − |
| FB.II-1 | 3.8a | 2.7a | 4.6, 4.8 | 360, 397b | ND | − | − |
| Reference | 6.8–11.9 | 10.6–24.9 | 5.4–10.6 | 0.5–9.9 | 28–95 | + | + |
| RCDP1 | <6.8 | <10.6 | 1.0–4.2 | >9.9c | ≤22 | − | − |
| RCDP2 | <6.8 | <10.6 | <0.1 | 0.5–9.9 | 28–95 | + | + |
| RCDP3 | <6.8 | <10.6 | 1.0–4.2 | 0.5–9.9 | 28–95 | + | + |
| ZS | <6.8 | <10.6 | 0.1–0.9 | 1.6–115.7 | 0–10 | − | − |
Reference values are from Laboratory Genetic Metabolic Diseases, Amsterdam, The Netherlands.
aMeasurement in fibroblasts using two controls showing values for C16 : 0 dimethylacetal (DMA)/C16 : 0 of 12.0 and 12.9% and for C18 : 0 DMA/C18 : 0 of 8.9 and 9.9%.
bReference values from Laboratory Medicine, Sheffield Children's Hospital, Sheffield, UK; PLA phytanic acid, <19.3 µmol/l. The two measurements were performed 1 month apart at age 5.5 years.
cCan be normal, increases with age. ERY, erythrocytes; FBRO, fibroblasts; IB, immunoblot; IF, immunofluorescence; ND, not done; PLA, plasma; peroxisomal processing, peroxisomal cleavage of 44 kDa ACAA1 to 41 kDa; ZS, Zellweger syndrome.
Selective defect in import of PTS2-tagged proteins in patient fibroblasts. (A) Immunoblot analysis using an antibody against PTS1-tagged ACOX1 in fibroblasts showed bands corresponding both to the unprocessed 72 kDa ACOX1 and the processed 50 kDa ACOX1 in control (C1), and in all four patients (FA.II-6, FA.II-1, FA.II-3 and FB.II-1). Only unprocessed 72 kDa ACOX1 was detected in a peroxisome deficient ZSD cell line (ZSD). These results indicated normal import of ACOX1, which is cleaved from the 72 kDa precursor to the 50 kDa ACOX1 protein after peroxisomal entry. (B) Immunofluorescence microscopy analysis using an antibody against PTS1-tagged catalase in fibroblasts showed peroxisomal staining in control (B) as well as in all four patients, shown is cells from FA.II-3 (C). (D) Immunoblot analysis using an antibody against ACAA1 in fibroblasts showed a band corresponding to the processed 41 kDa ACAA1 in the control (C1), while only a band corresponding to the unprocessed 44 kDa ACAA1 was detected in the four patients (FB.II-1, FA.II-3, FA.II-1 and FA.II-6), as well as in cells from a peroxisome deficient ZSD cell line (ZSD), indicating abnormal peroxisomal processing of ACAA1 due to deficient import of PTS2-tagged ACAA1 in these cell lines. (E) Immunofluorescence microscopy analysis using the same antibody against ACAA1 showed peroxisomal staining in control cells (E), whereas the staining was cytosolic in cells from the four patients, indicating deficient peroxisomal import of PTS2-tagged ACAA1, shown are cells from FA.II-3 (F).
Three PTS2-tagged peroxisomal matrix proteins are known: AGPS, involved in plasmalogen biosynthesis (7), phytanoyl-CoA 2-hydroxylase (PHYH) (28), involved in phytanic acid α-oxidation, and acetyl-CoA acyltransferase 1 (ACAA1) (29), which has a redundant function in fatty acid β-oxidation (30). All RCDP types are characterized by low levels of plasmalogens (31–33), while patients with RCDP1 in addition show accumulation of phytanic acid and reduced intra-peroxisomal processing of the precursor form of ACCA1 (26,34). Peroxisomal import of PTS2-tagged proteins was evaluated by measuring the levels of plasmalogens in erythrocytes or fibroblasts, and the activity of the PTS1-targeted GNPAT in fibroblasts. Reduced activity of GNPAT is typically observed in cells with deficient import of the PTS2-targeted AGPS (35,36). The plasmalogen levels were mildly reduced in all four patients, as was the GNPAT activity measured in FA.II-1, FA.II-3 and FA.II-6 (Table 2), indicating deficient import of AGPS. Import of PTS2-tagged PHYH was assessed by measuring the level of phytanic acid in plasma from the patients, and the phytanic acid α-oxidation in fibroblasts. The levels of phytanic acid were highly elevated (Table 2), and phytanic acid α-oxidation markedly reduced (Table 2), indicating deficient import of PHYH. Immunoblot analysis using an antibody against PTS2-tagged ACAA1, showed peroxisomal processing in the control cells, while only the unprocessed ACAA1 was detected in all four patients, as well as in a peroxisome deficient ZSD cell line (Table 2, Fig. 4D), indicating deficient import of ACAA1 in these cell lines. Finally, immunofluorescence microscopy analysis using the same ACAA1 antibody showed peroxisomal staining in control cells (Fig. 4E), while the patient cells showed cytoplasmic staining (Table 2, Fig. 4F), further confirming the deficient peroxisomal import of ACAA1.
Together, the results confirm that the selective loss of PEX5L in the patients results in a peroxisomal dysfunction due to deficient import of PTS2-tagged proteins only.
Expression of PEX5L in patient fibroblasts restores the peroxisomal import of PTS2-tagged proteins
In order to confirm that the isolated defect in import of PTS2-tagged proteins was indeed caused by loss of PEX5L, we ectopically expressed either PEX5L or PEX5S in fibroblasts from FA.II-3 and the PEX5 deficient ZSD human fibroblast cell line PEX5.11, aiming to restore the peroxisomal import of PTS2-tagged proteins. The cells were co-transfected either with PTS2-tagged green fluorescent protein (PTS2-GFP) and PEX5S, or with PTS2-GFP and PEX5L. Both the ZSD cell line and the patient fibroblasts transfected with PEX5S showed a cytoplasmic fluorescence pattern (Fig. 5A and B), indicating that PEX5S could not restore peroxisomal import of PTS2-tagged proteins in either cell line. In contrast, both cell lines transfected with PEX5L showed a peroxisomal fluorescence pattern consistent with restored import of PTS2-tagged proteins (Fig. 5C and D). These results confirm that loss of PEX5L is responsible for the peroxisomal dysfunction due to the isolated defect in import of PTS2-tagged proteins in our patients.
PEX5L expression restored import of PTS2-tagged proteins in patient fibroblasts. (A and B) Fibroblasts from FA.II-3 and a PEX5 deficient cell line, PEX5.11 were co-transfected with PEX5S and PTS2-GFP (C and D) or PEX5L and GFP-PTS2. A cytoplasmic fluorescence pattern compatible with defective import of PTS2-tagged proteins was observed in both cell lines transfected with PEX5S (A and B), whereas both cell lines transfected with PEX5L showed a peroxisomal fluorescence pattern compatible with restored import of PTS2-tagged proteins.
Discussion
Previously reported patients with mutations in PEX5 had ZSD (21,37–39), primarily due to deficient import of PTS1-tagged proteins, but often accompanied by impaired import of PTS2-tagged proteins (37). The patients with the novel c.722dupA (p.Val242Glyfs∗33) mutation in PEX5 lacked several features typically associated with ZSD, such as hepatic dysfunction and sensorineural hearing loss (11–13). In contrast, some of their key features such as congenital cataracts and chondrodysplasia punctata are usually associated with RCDP (14–16). Altogether, the phenotype in our patients, including congenital cataracts, postnatal growth delay, microcephaly and severe intellectual disability in all four patients, mild shortening of humeri in three patients, and chondrodysplasia punctata in two patients, was consistent with RCDP (14–16).
Both PEX5L and PEX5S recognize PTS1-tagged proteins (40), but PEX5L is also a co-receptor critical for PEX7-mediated import of PTS2-tagged proteins (21,41,42). This suggested that the novel c.722dupA mutation in PEX5 demonstrated to result in a selective loss of PEX5L, might cause RCDP due to an isolated defect in import of PTS2-tagged proteins. Indeed, the normal import of PTS1-tagged proteins in the patients excluded that they have ZSD, and their isolated defect in the peroxisomal import of PTS2-tagged proteins is similar to the peroxisomal defect in RCDP1, thus explaining their RCDP. Ectopic expression of PEX5L restored the peroxisomal import of PTS2-tagged proteins in cultured patient fibroblasts, thus we prove that the homozygous c.722dupA mutation in PEX5 causes selective loss of PEX5L, resulting in an isolated defect in the peroxisomal import of PTS2-tagged proteins, and thereby give rise to RCDP5.
Plasmalogen deficiency is regarded as the primary cause of RCDP (31–33). Plasmalogens are components of cellular and subcellular membranes, as well as specialized membranes, such as myelin, and secreted membranes, such as synaptic vesicles (31). The plasmalogen level is relatively low in the brain of newborns, but increases during development to reach its peak around 30 years of age (31,43). Plasmalogens provide structural support to membranes, functions as a reservoir for lipid second messengers, and protect against oxidative damage (31,44). However, the pathomechanism of plasmalogen deficiency in RCDP is unknown (31). Our patients with RCDP5 had a less pronounced reduction in tissue plasmalogen level compared with patients with severe RCDP, and their clinical phenotype is analogous to other RCDP patients with milder biochemical defects, who often have less pronounced skeletal abnormalities, later onset of epilepsy, and less profound growth delay and intellectual disability (14–16,36). Similar to other RCDP patients with partial plasmalogen deficiency, patients with RCDP5 reached more developmental milestones, such as self-feeding, independent ambulation and limited impressive and expressive language skills, than most RCDP patients with severely reduced plasmalogen levels (14–16,36), but it is still apparent that even partial plasmalogen deficiency has significant consequences. Bilateral cataract does not seem to correlate with the severity of the plasmalogen deficiency, as it is a consistent finding in RCDP patients and knock out mouse models with both severe and partial plasmalogen deficiencies (14–16,36,45–47). PEX5L is regarded as critical for PEX7-mediated peroxisomal import of PTS2-tagged proteins (21,41,42), but loss of PEX5L did not result in the most severe clinical and biochemical RCDP phenotypes in the RCDP5 patient. This suggests that PEX7 sustains low level import of the PTS2-targeted AGPS in the absence of the co-receptor PEX5L.
Similarly to patients with RCDP1 (15,16), the patients in this study had accumulation of phytanic acid in plasma, due to deficient import of PTS2-tagged PHYH. Accumulation of phytanic acid due to mutations in PHYH, results in adult Refsum disease (ARD; MIM 266500) (28), characterized by onset from adolescence to adulthood with anosmia, early-onset retinitis pigmentosa and variable combinations of neuropathy, deafness, ataxia and ichthyosis (48–51). ARD patients are also at risk of developing cardiac arrhythmias (48,49). Clinical and neurophysiological studies in the three affected siblings in family A indicated that they had demyelinating peripheral neuropathy, which to the best of our knowledge has not been reported in RCDP, but is an important sign of ARD (48,49). However, additional signs of ARD were not recognized in our patients (anosmia was not assessed). Phytanic acid accumulates with age and in a diet-dependent manner in patients with deficient phytanic acid α-oxidation (52). Due to the considerably elevated levels of phytanic acid, a diet restricted in phytanic acid was recently initiated in the patients, aiming to prevent toxic effects of accumulation.
To conclude, we report a novel c.722dupA mutation in PEX5, the first to exclusively affect PEX5L expression, and that unlike previous mutations in PEX5, results in RCDP instead of ZSD. The peroxisomal dysfunction in the patients was due to an isolated defect in import of PTS2-tagged proteins, previously described from mutations in PEX7 only. Mutations in PEX5 resulting in a selective loss of PEX5L represents a fifth type of RCDP. Due to the clinical and biochemical overlap between RCDP1 and RCDP5, sequencing of PEX7 and the PEX5L-specific exon 9 in PEX5 should be considered in patients with an isolated defect in peroxisomal import of PTS2-tagged proteins
Materials and Methods
Consent
This study was approved by the Norwegian South-Eastern Regional Ethics Committee. Informed written consent was obtained from the parents in both families investigated.
Whole-exome sequencing
Genomic DNA from FA.I-1, FA.II-1 and FA.II-6 was used to capture exomes with the Illumina TruSeq Exome Enrichment kit and sequenced on an Illumina HiSeq2000 with 100 bp paired-end reads. Alignment against the GRCh37 human reference genome was performed with a Burrows-Wheeler Aligner (v.0.5.9) (53) PCR-duplicates marking and removal with Picard (v.1.104), and indel realignment, base quality recalibration, and joint variant calling with the Genome Analysis Toolkit (GATK, v.2.5) (54). Finally, functional annotation was performed with snpEff (18) and Variant Effect Predictor (55) using Ensembl 71. Variants were analyzed with FILTUS (v.99-9) (http://folk.uio.no/magnusv/filtus.html), assuming recessive inheritance due to parental consanguinity. Variants with a high probability of being technical artifacts, as computed by the GATKs ‘variant quality score recalibration’ procedure, were removed in the filtering process.
Sanger sequencing
Sanger sequencing was performed with Big DyeTM Terminator v.3.1 Cycle Sequencing Kit (Applied Biosystems) on an ABI 3730 sequencer (Applied Biosystems), using primers for PEX5 (37) and PEX7 (56) in individual FB.II-1 as described, and primers for validation and segregation analysis for the novel variant in PEX5 in family A provided in Supplementary Material, Table S4.
qRT-PCR analysis
RNA from fibroblasts was converted to cDNA using the High Capacity cDNA Reverse Transcription Kit (Applied Biosystems). qRT-PCR amplifications were performed using the SYBR Green JumpStart Taq ReadyMix (Sigma-Aldrich) on a QuantStudio™ 12 K Flex Real-Time PCR System (Applied Biosystems). Samples were run in triplicates, and amplification levels were calculated according to the 2-ΔΔCt method (57). Two PEX5L-specific primer sets and two primer sets amplifying PMM1 and GADPH transcripts as endogenous references were used and provided in Supplementary Material, Table S5. Dissociation curve analysis revealed a single product for each primer pair. Negative template controls and genomic DNA controls (cDNA synthesis reactions without RT enzyme) did not show amplification. Statistical significance was evaluated using a two-tailed Student's t-test.
Cell culture
Skin biopsies were obtained and cultured according to standard procedures. Fibroblasts were transfected as described (37). A cell line deficient of both PEX5 isoforms, PEX5.11 (37), was used as control in several experiments.
Biochemical analysis
Concentrations of VLCFAs and phytanic acid in plasma (58), and peroxisomal phytanic acid α-oxidation (59), peroxisomal β-oxidation using C26 : 0 and pristanic acid as substrates (60), and the activity of GNPAT (61) were measured in cultured fibroblasts as described.
Immunoblot analysis
Immunoblot analysis was performed on protein extracts from cultured fibroblasts using an antibody against ACOX1 (in house generation), ACAA1 (Atlas antibodies, article number HPA007244) and PEX5 (Proteintech Europe, 12545-1-AP) as described (62).
Immunofluorescence microscopy analysis
Immunofluorescence microscopy analysis was performed on cultured fibroblasts using an antibody against catalase (in house generation) and ACAA1 (Atlas antibodies, HPA007244) as described (27).
Funding
This study was supported by the Norwegian Southeastern Regional Health Authorities (2011071), and ‘Legatet til Henrik Homans Minde’ to T.B., and ‘Anders Jahres fond til vitenskapens fremme’ to D.M.
Acknowledgements
We are grateful to the families who participated in this study. We thank Madeleine Fannemel, Kristin Ørstavik, Eli Ormerod and Heidi-Karin Lundlie from Oslo University Hospital, Oslo, for genetic counselling, electrophysiological studies, taking skin biopsies and collecting skeletal images, respectively. We thank Petra Mooyer and Conny Bakker-Dekker from the Laboratory Genetic Metabolic Diseases, Academic Medical Center, Amsterdam, for the study of peroxisomal parameters in fibroblasts. The sequencing service was provided by the Norwegian High-Throughput Sequencing Centre, a national technology platform supported by the ‘Functional Genomics’ and ‘Infrastructure’ programs of the Research Council of Norway and the Southeastern Regional Health Authorities.
References
Author notes
These authors contributed equally to this work.
