Background: The three-nucleotide deletion, GAG (within the gene TOR1A), is the only proven cause of childhood-onset dystonia (DYT1). A potentially pathogenic role of additional sequence changes within TOR1A has not been conclusively shown.
Methods: DNA sequencing of exon 5 of TOR1A in a patient with DYT1.
Results: Detection of sequence change c.863G>A in exon 5 of TOR1A in the patient. The G>A transition results in an exchange of an arginine for glutamine (p.Arg288Gln) in subdomain α5 of TOR1A. Several findings point to a potentially pathogenic role of the sequence change in the patient: The base change is absent in 1000 control chromosomes; an Arg at position 288 of TOR1A has been conserved throughout vertebrate evolution, indicating an important role of Arg288 in TOR1A function; functional studies demonstrate enlarged perinuclear space in HEK293 cells overexpressing TOR1A with the p.Arg288Gln mutation. The same morphological changes are observed in cells overexpressing the common GAG TOR1A mutation but not in cells overexpressing wild-type TOR1A.
Conclusions: The sequence change described here may be a novel pathogenic mutation of TOR1A in DYT1.
Statistics from Altmetric.com
Dystonia 1 (early-onset dystonia or Oppenheim dystonia; DYT1) is a severe autosomal-dominant childhood-onset dystonia. Most frequently, dystonia starts in the limbs and generalises within a few years of onset. However, clinical presentation can vary remarkably with respect to both age and site of onset and disease progression. Dystonia does not generalise in all instances and may remain focal or segmental in some. Furthermore, penetrance is reduced. Only about 30% of mutation carriers are clinically affected.1
The disease gene (TOR1A) on the distal long arm of chromosome 9 (9q34)2 was identified in 1997.3 To date, only one proven pathogenic mutation has been detected in TOR1A. This mutation is a deletion of one of two adjacent GAG trinucleotides within exon 5 (904_906delGAG/907_909delGAG) and results in a loss of glutamine in TOR1A (Δ302/303). The Δ302/303 deletion has independently occurred several times in different families and populations. It is most frequent in the Ashkenazi Jewish population in which it was introduced by a common founder.4 5 Given a prevalence of early-onset dystonia of 1/16,000–1/20,000 in this population and taking into account the reduced penetrance, the allele frequency is about 1/5000–1/6000. In the non-Jewish population, the disease gene is estimated to occur at a frequency of about 1/60,000.6
TOR1A encodes the AAA+ chaperone protein torsin A (TOR1A), which is composed of 332 amino acids. TOR1A is a luminal protein associated with the endoplasmic reticulum (ER) and to some extent with the nuclear envelope (NE). The Δ302/303 mutation causes a redistribution of TOR1A from the ER to the NE and causes alterations of the NE ultrastructure.7 It also recruits wild-type TOR1A to the NE.7 Finally, Δ302/303 TOR1A induces inclusions in the NE.
Here, we describe a previously not recognised base change in exon 5 of TOR1A in a patient with dystonia and provide evidence for a causal role of this mutation in the patient’s phenotype.
MATERIALS AND METHODS
DNA was extracted from peripheral lymphocytes according to standard procedures. The six exons including intron–exon boundaries of GCH1 were sequenced as previously described.8 Exon 5 of TOR1A was amplified and sequenced using primers Dyt1_ex5_forward 5′-GATGGGAGGGTAGGTAAGTG-3′ and Dyt1_ex5_reverse 5′-GGAAACAATGCCAGGGAGAA-3′ applying a standard sequence protocol.
HEK 293 cells were transiently transfected with mammalian expression constructs encoding human wild-type (hWT) TOR1A and human mutant (hΔGAG) TOR1A, as described previously.9 A G>A mutation was inserted into the hWT TOR1A cDNA at position 863 of the published sequence NM_000113 by site-directed mutagenesis applying the Quickchange procedure (Statagene, La Jolla, CA) with primers Tor_IVM_fw (5′-GTGGAAATGCAGTCCCAAGGCTATGAAATTGATG-3′) and Tor_IVM_rev (5′-CATCAATTTCATAGCCTTGGGACTGCATTTCCAC-3′). This resulted in the exchange of an arginine by glutamine at position 288 of the protein. All constructs were checked by sequencing. Transfection of HEK 293 cells was carried out using the Lipofectamine Plus method (Gibco BRL) according to the manufacturer’s instructions using 10 μg DNA for each construct.
Cells grown to 80% confluence were harvested and fixed in 2.5% glutaraldehyde (Paesel-Lorei, Frankfurt, Germany) in 0.1 M cacodylate-buffer, fixed in 1% OsO4 in 0.1 M cacodylate buffer thereafter, and dehydrated in 70% ethanol. The 70% ethanol step was saturated with uranyl acetate for contrast enhancement. Dehydration was completed in propylene oxide. The pellets were then embedded in Araldite (Serva, Heidelberg, Germany). Ultra-thin sections were produced on a FCR Reichert Ultracut ultramicrotome (Leica, Bensheim, Germany), mounted on pioloform-coated copper grids and contrasted with lead citrate. Sections were analysed and documented with an EM 10A electron microscope (Carl Zeiss, Oberkochen, Germany).
Western blot analysis
In order to compare the expression levels achieved by transient transfection, we analysed the transfected HEK 293 cells by Western blotting. Cells were harvested by trypsination and lysates were prepared by resuspending PBS-washed cells in lysis buffer containing 50 mM Tris pH 8.0, 50 mM NaCl, 1% NP-40 and protease inhibitors. Equal amounts of protein (30 μg) were resolved on 10% polyacrylamide gels, transferred to nitrocellulose (Schleicher & Schuell Bioscience, Dassel, Germany), and Western blot analysis was performed using anti-V5 antibody (Invitrogen GmbH, Karlsruhe, Germany) (diluted 1:1000). The amount of loading was controlled for by β-actin immunoreactivity. Densitometric analysis (by volume) was performed according to standard procedures.
The female index patient is presently 18 years old. She suffers from generalised dystonia, including dysphagia and severe dysarthria, and is wheelchair-bound. She has severe contractures of multiple joints, including ankles, knees, hips and elbows, and is unable to lift her arms by more than 5 cm. The muscle tone is elevated, tendon reflexes increased and the Babinski sign present bilaterally. She also has a marked facial palsy and reduced tongue mobility. At night, she experiences painful episodes owing to dystonic movements. Cranial MRI revealed mild cerebellar atrophy. However, she has no cerebellar signs, no ataxia or parkinsonism. Mental performance is not affected and she is doing well at school. Laboratory tests revealed normal HVA and 5-HIAA levels (132 and 82 nmol/l, respectively, compared with the normal range of 115–488 and 66–141 nmol/l). She does not respond to L-Dopa, which was administered at a maximum dose of 300 mg per day. Other laboratory parameters, including molecular genetic analysis of PANK2 and GCH1, as well as lysosomal enzymes, organic acids, amino acids, carbohydrate deficient transferrine and lactate in plasma and cerebrospinal fluid, were normal. The patient was born to a 29-year-old healthy mother after an uneventful pregnancy. As a newborn, she was passive and spontaneous movement was reduced. She was able to sit at 9 months, crawled at 10 months and walked with support by 18 months of age. During her second year of life, stalling of gross motor development became obvious and she never learned to walk independently. At 4 years of age, an achillectomy was performed because she was able to move on tiptoes only. She developed generalised dystonia that started bilaterally in the lower limbs and became dependent on a wheelchair at age 5 years. Fine motor abilities also declined. At age 13 years, she could not eat or drink without assistance. Her mother, now 47 years old, was also examined by a neurologist. She is healthy and does not display any abnormal neurological signs or symptoms. DYT1 mutation analysis was requested by the patient and her parents.
Sequencing of exons 1–6 of the GCH1 gene in the index patient did not reveal any abnormality. We then sequenced exon 5 of TOR1A that contains the GAG deletion (Δ302/303), which is characteristic of early-onset dystonia. Although this deletion was not detected, we found a G>A transition at position 863 (G863A). This base change is predicted to result in the substitution of an arginine by a glutamine at amino-acid position 288 (Arg288Gln). We also sequenced exon 5 in the patient’s parents and found the same heterozygous G>A transition in the mother.
In order to test whether this base change is a polymorphism, we sequenced exon 5 of TOR1A in 500 healthy German controls. All controls (ie, 1000 chromosomes) carried a G at position 863. There is no evidence of the transition observed in the patient being a polymorphism.
Possible effects of the Arg288Gln mutation were studied in cells that had been transfected with Arg288Gln TOR1A, and as controls with hWT TOR1A and with hΔGAG TOR1A. The three constructs consistently resulted in different expression levels in independent trials. The highest expression was observed for the hWT TOR1A construct and the lowest levels for the hΔGAG TOR1A construct, as shown by densitometric ratios of torsinA/β-actin (fig 1).
Furthermore, the ultrastructure of the NE of HEK293 cells transfected with all three constructs was studied by electron microscopy. The NE of cells transfected with the hWT TOR1A construct displayed the typical uniform space between the outer and inner nuclear membrane (fig 2A) similar to untransfected HEK cells (fig 2B). In contrast, a focally enlarged perinuclear space was found in nuclei of HEK293 cells overexpressing Arg288Gln TOR1A. The electron-lucent spaces were filled with unidentifiable remnants of membranes (fig 2C). The same results were obtained in HEK 293 cells overexpressing the hΔGAG TOR1A construct (fig 2D).
Finally, we investigated whether arginine at the relevant position in TOR1A has been conserved during evolution. As shown in figure 3, an R (arginine) is found in all vertebrates. This finding points to an important role of Arg288 in the normal function of TOR1A.
We detected a specific amino-acid change at position 288 (Arg288Gln) of TOR1A in a patient with severe early-onset dystonia. Several findings point to a potentially functional role of this mutation in the phenotype of the patient:
1) The base change is absent in 1000 control chromosomes and thus appears to be specific for the patient.
2) Functional studies revealed focally enlarged perinuclear space in nuclei of HEK 293 cells transfected with Arg288Gln TOR1A (fig 2C), but not in those overexpressing hWT TOR1A (fig 2A). The same enlargement of the perinuclear space was observed in HEK 293 cells overexpressing the hΔGAG TOR1A construct (fig 2D) (see also10 11).
3) An arginine at position 288 of TOR1A has been highly conserved during evolution (fig 3). This indicates an important function of this amino acid within TOR1A.
Deformations of the NE, such as enlargement of the perinuclear space, are considered characteristic of torsinA-related pathology and are found both in cell systems (overexpression of GAG deletion in cells) and in transgenic mouse models of dystonia1 (GAG deletion in transgenic mice).9 12 Therefore, the observation of changes in the NE induced by the present mutation in a cell system lends support to a pathogenic role of this mutation in dystonia.1 There is also no indication that the findings described here are artefacts, owing to different expression levels of the three constructs. Although even high levels of hWT TOR1A did not result in cellular pathology, obvious changes were found in cells expressing low levels of Arg288Gln TOR1A. Furthermore, the constructs themselves, which were tagged at the carboxy-terminus, cannot have caused cellular pathology as these constructs have been shown not to differ from untagged ones.9 It needs to be stressed, however, that deformations of the NE have so far not been reported in post-mortem tissue of patients.
The disturbed function of Arg288Gln TOR1A can be deduced from the normal structure/function of TOR1A and the pathogenic effects of the GAG deletion (Δ302/303). TOR1A is an AAA+ ATP binding chaperone protein. There are extensive homologies with the C-terminal domains of AAA+ proteins ClpA and ClpB. Together, they make up the ClpAB-C/torsin family of AAA+ proteins.13 The C-terminal region of TOR1A extends from amino-acid residues 272–332 and is composed of the α-helical subdomains α5, α7 and the sensor 2 domain C-terminal to subdomain α6.14 TOR1A appears to occur as a homohexamer that is formed by interaction of the various subdomains of TOR1A with each other.
The Arg288Gln mutation described here lies within subdomain α5. This subdomain is adjacent to subdomain α6 that is affected by the GAG deletion. Previous findings suggest that Δ302/303 interferes with ATP binding and oligomerisation of TOR1A.14 15 Similarly, the Arg288Gln mutation described here might interfere with oligomerisation of TOR1A via subdomain α5.
Similar to the common GAG deletion, the mutation described here also appears to be inherited at reduced penetrance, as the same mutation was detected in the asymptomatic mother.
To date, three additional potential mutations other than the GAG deletion have been described. A patient with early-onset dystonia and myoclonus was found to carry a 18-bp deletion within exon 5 (966_983del, Phe323_Tyr328del).16 One of the patient’s brothers also carried this deletion but had myoclonus only. The patient’s mother, also a carrier of the deletion, displayed possible dystonia and no myoclonus. Additional studies revealed a SCGE gene mutation in the same patients.17 SCGE encodes -sarcoglycan and mutations in this gene are common in autosomal-dominant myoclonus dystonia (dystonia 11).18 Therefore, the 18-bp deletion might not have caused disease in the patients of this family. However, it cannot be ruled out that it contributed to dystonia in the patients. A second mutation (934_937delAGAG) was detected in a healthy blood donor with no neurological manifestations.19 Unfortunately, the proband remained anonymous and could not be further evaluated neurologically. Furthermore, a single amino-acid exchange at position 216 of the protein has been described. At this position, a histidine (H) is found in 12% and an aspartic acid (D) in 88% of the normal population.20 This polymorphism might modify susceptibility to disease, as the H allele weakens manifestation of the GAG deletion in cellular systems.14 Consistent with this observation, the H allele appears to be increased in non-manifesting carriers of the GAG deletion and decreased in GAG deletion carriers manifesting dystonia.19
In conclusion, the Arg288Gln mutation described here might be a novel pathogenic mutation in TOR1A that causes early-onset primary dystonia (DYT1).
We thank Dorothee Ringleb-Wieden for excellent technical assistance and Professor Dr Klaus Altland for performing densitometric analysis.
Competing interests: None.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.