Identification of the Early-Onset Torsion Dystonia Gene

Source: Ozelius L, et al. The early onset torsion dystonia gene (DYT1) encodes an ATP binding protein. Nature Genetics 1997;17:40-47.

The hereditary dystonias include a heterogeneous group of disorders. At least six different gene loci have been described which lead to different dystonias:

1. Early-onset torsion dystonia (EOTD) on (chromosome) 9q34

2. DOPA-responsive dystonia on 14q21-22

3. Paroxysmal dystonia on 2q

4. A mixed phenotype on 8

5. Late onset focal dystonia on 18p

6. Lubag dystonia-parkinsonism on Xq13.1

A number of other biochemical conditions can produce pure dystonia, including tyrosine hydroxylase deficiency. These conditions reduce dopamine synthesis directly and GTP cyclohydrolase I insufficiency, which impairs synthesis of biopterin, a necessary co-factor for tyrosine hydroxylase. Additionally, a number of medical conditions can produce prominent dystonia in the context of a more generalized disorder (e.g., Wilson disease).

Of all of the hereditary causes of pure dystonia, EOTD is the most severe and the most common. Symptoms usually begin focally early in the second decade of life but generalize in approximately five years. EOTD displays autosomal dominant inheritance with 30-40% penetrance. The disorder is most common in Ashkenazi Jews (probably due to a "founder effect") but may be seen in non-Ashkenazi individuals as well. The EOTD locus on chromosome 9 has been found to encode two previously unknown proteins, termed torsinA and torsinB. Patients with EOTD display a GAG deletion in the torsinA gene. The net result of this mutation is the absence of one of two glutamic acid residues at the carboxy-terminus of torsinA. The identified mutation in torsinA is present in a heterozygous fashion in affected individuals, as might be expected in a dominantly-inherited disorder. Ozelius et al found this mutation in all of the 173 affected individuals from 64 Ashkenazi Jewish pedigrees as well as all of the 88 affected individuals from four non-Ashkenazi pedigrees. Not one of the 267 unaffected individuals from all pedigrees displayed the mutation. Thus, it appears that a common mutation is responsible for the EOTD phenotype within all ethnic groups (i.e., EOTD is genetically homogeneous).


The paper by Ozelius et al is the culmination of almost a decade of work and is a major landmark in dystonia research. This work may soon result in a commercially available test for EOTD, which is expected to be of very high predictive value—it will be able to tell the clinician whether or not a patient has EOTD with a high degree of certainty. At present, the function of torsinA or torsinB is unknown. Both proteins are highly homologous to one another, and both are predicted (based on structural considerations) to bind ATP. It is interesting that the GAG-deletion in torsinA produces disease in the heterozygous state. There are only a few known genetic disorders whereby deletion of a single amino acid produces disease in the heterozygous state. Among neurological disorders, there is a single precedent: deletion of a single arginine residue in the a1 subunit of the L-type voltage sensitive calcium channel results in dominantly inherited hypokalemic periodic paralysis. Identification of torsinA as playing a key role in a pathway related to dystonia may accelerate identification of gene products in other (non-9q34) hereditary dystonias, much as the discovery of dystrophin accelerated the discovery of associated proteins in the non-Duchenne muscular dystrophies. —rt