Autosomal dominant cerebellar phenotypes

The dominantly inherited ataxias comprise a series of clinical phenotypes that include ataxia, dysarthria, dysmetria, and intention tremor resulting from the involvement of the cerebellum and its afferent and efferent pathways. These inherited ataxias occur commonly throughout the world and represent an area of intense neurologic research. Although the clinical manifestations and neuropathologic findings of cerebellar disease are predominant, there may be characteristic changes in the basal ganglia, brainstem, spinal cord, optic nerves, retina, and peripheral nerve. In large families with dominantly inherited disease, there are many gradations, from pure cerebellar manifestations to mixed cerebellar and brainstem disorders, cerebellar and basal ganglia syndromes, and spinal cord or peripheral nerve disease. Rarely, dementia may be present as well. The clinical picture may be consistent within a family with dominantly inherited ataxia, but on occasion there is a characteristic syndrome in the majority of affected family members and an entirely different phenotype in one or several members. For these reasons, I suggested in 1990 ^[1] that for the autosomal dominant phenotypes, the genotype would be required to settle the issue of disease assignment and classification.

Neurologists have written extensively as to the appropriateness of various classifications and have debated, often passionately at meetings, the uniqueness and placement of newly evaluated families with inherited cerebellar disease. Categories of disease abound, therein satisfying the needs of selective investigators to be either unifiers or splitters, often based on quite arbitrary factors ^[1]. Scholarly, yet divergent, classifications of dominantly inherited syndromes have been formulated by Greenfield, ^[2] Konigsmark and Weiner, ^[3] Harding, ^[4] and the World Federation of Neurology Study Group on Cerebellar Ataxias (unpublished, 1993). Harding ^[4] separates the autosomal dominant cerebellar ataxias (ADCAs) into type I and type II, in which ataxia is combined to a varying degree with pyramidal and extrapyramidal signs and ophthalmoplegia. Type II is distinct in having the consistent additional feature of retinal degeneration.

My colleagues and I have evaluated more than 100 families with another dominantly inherited type of cerebellar degeneration, Machado-Joseph disease (MJD), in the United States, Canada, Portugal, India, Japan, and Brazil, and we believe we are witnessing the expression of a single gene resulting in a multiplicity of phenotypes. It is believed to have been distributed worldwide in the 16th century by Portuguese navigators and brought by Portuguese immigration to the northeastern and western coast regions of the United States, where it is most commonly found. Type 1 disease occurs early in individuals with pyramidal and extrapyramidal signs; type 2 has an intermediate age of onset (20 to 50 years) and presents with cerebellar, pyramidal, and extrapyramidal manifestations. Type 3 has a later onset (after 50 years) and is associated with a progressive pancerebellar syndrome and peripheral neuropathy. Type 4 disease refers to dominantly inherited parkinsonism with ataxia, distal atrophy, and sensory loss. These clinical studies indicated to us that a presumed single genotype can result in several distinct phenotypes because of varying host genetic modifiers and normal genetic heterogeneity, but genotypic verification was necessary ^[1,5-7].

The problem posed by dominantly inherited diseases of the nervous system, such as the inherited cerebellar degenerations, is the lack of a known primary metabolic clue, or primary storage product, to indicate a potential molecular basis of disease. In these cerebellar disorders, there is a patterned neuronal degeneration with variable secondary gliosis. Even the findings of specific protein abnormalities on two-dimensional gels, or of specific changes in messenger RNA on Northern blots obtained from brain samples of patients having a dominantly inherited cerebellar disease, are not sufficient evidence to warrant the conclusion that these products are due to primary gene mutations rather than to being the result of disease ^[8]. Therefore, new research strategies must be employed to determine the molecular basis of autosomal dominant disorders and the molecular explanation for phenotypic variation within, and between, families. Recently, linkage analysis has proved to be a powerful tool with which to map the chromosomal location of a series of important, dominantly inherited neurologic diseases for which no biochemical insight was known, and these new findings have brought great clarity to the sorting out and classification of the various clinical phenotypes Table 1 ^[1,9].

Zoghbi et al ^[10,11] found linkage of one American family with autosomal dominant olivopontocerebellar atrophy (OPCA) to the HLA loci on the short arm of chromosome 6 (6p22-p23). A maximum lod score of 5.83 was found at a recombination fraction of 0.12. Similarly, Rich et al ^[12] reported linkage to the HLA-A locus in the Schut-Swier OPCA family, and the disease locus was about 15 cMo telomeric to the HLA-A locus on the short arm of chromosome 6. These studies followed the pioneer observations of classic linkage to HLA in a large OPCA family by Jackson et al in 1977 ^[13]. The families that map to 6p can now be referred to as ADCA type I, subtype spinocerebellar atrophy type 1 (SCA1). However, other families with OPCA phenotypes identical to those cited have undergone linkage analysis without any linkage to HLA ^[14]. Dominant OPCA is genetically heterogeneous. A common OPCA phenotype may be due to separate genotypes and, as mentioned from our extensive MJD clinical data, there are good grounds to postulate that a single genotype can produce many different phenotypes.

In 1990, Orozco et al ^[15] added additional clinical phenotypic complexity by describing 263 patients from the Holguin province of Cuba who have ADCA. These patients have a common ancestry, and the population may be the largest homogeneous group of patients yet described. Age of onset ranged from 2 to 65 years, and there was considerable clinical variability within the various families. The authors stress the high likelihood that the patients from the Holguin region are descendants of common ancestors and thus their cerebellar degeneration results from a "founder effect". The presumed genetic homogeneity has resulted in significant clinical variability between, and within, families to a degree that is atypical for SCA1 type of OPCA. The neuropathologic findings are compatible with OPCA but not with MJD, as the inferior olives were affected in all examined cases. Clinical findings sometimes encountered in OPCA (including parkinsonian rigidity, optic disk pallor, spasticity, and retinal degeneration) were minimally present in these families, but slow saccadic eye movements were present in many patients ^[1,7,15]. This interesting symptom complex (referred to as ADCA type I subtype SCA2) could be a unique Cuban cerebellar degenerative disease. Where it fit into the spectrum of OPCA and MJD was speculative on a clinical and neuropathologic basis. Determination of the genotype, first by linkage analysis and then by gene cloning and sequencing, would settle the issue.

In late 1993, in a rapid series of brilliant clinical-molecular correlation studies, the genotypic specificities of the ADCA type I (SCAs 1 and 2), MJD, and other SCA phenotypes, including the syndrome of dentato-rubro-pallidoluysian atrophy (DRPLA), were defined. Orr et al ^[16] confirmed that SCA1 maps to 6p22-p23 and found a highly polymorphic CAG repeat in this region. This CAG repeat was found to be unstable and expanded in individuals with SCA1. Patients were found with 43 or more CAG repeats and control subjects with 36 or fewer repeats, and there was a direct correlation between the size of the (CAG)_n repeat expansion and the earlier age of onset of SCA1. Larger alleles occurred in juvenile cases, and anticipation was present in subsequent generations. Chung et al ^[17] found results similar to those of Orr et al ^[16] and noted an increase in repeat number, with paternal transmission of disease indicating imprinting was occurring. Gispert et al ^[18] assigned the positional map of SCA2 to 12q23-q24.1, following linkage analyses for the Cuban pedigrees. Thus similar clinical phenotypes of SCA1 and SCA2 mapped separately, to 6p and 12q.

Takiyama et al ^[19] assigned the gene for MJD to the long arm of chromosome 14 (14q24.3-q32) by genetic linkage to microsatellite loci D14S55 and D14S48, with a multipoint lod score Zm = 9.719. They found it to be a most common autosomal dominant spinocerebellar degeneration in Japan, and its presence is distributed throughout the country. At the Third International Workshop on Machado-Joseph Disease held in Furnas, Sao Miguel, Azores, Portugal (April 1994), there were reports that North American and Brazilian MJD families also map to 14q, suggesting a single gene locus for this disorder in North and South America and in Japan ^[20-22].

In 1994, Koide et al ^[23] and Nagafuchi et al ^[24] reported that the syndrome of dominantly inherited DRPLA, characterized by progressive ataxia, choreoathetosis, dystonia, seizures, myoclonus, and dementia, mapped to 12p12-ter, and they identified an unstable CAG-repeat expansion in all 22 DRPLA patients examined. Larger expansions were found in earlier-onset patients with progressive myoclonus epilepsy. Control subjects had up to 26 CAG repeats and patients had 49 or more repeats. Anticipation and paternal imprinting were present, with larger expansions in children noted when the disease was inherited from their fathers.

Stevanin et al ^[25] suggested another SCA locus (SCA3, next in order in the literature) in a family with ADCA type I ataxia in which linkage to SCA1 (6p) and SCA2 (12q) has been excluded. Thus an additional genotype needs to be defined. In 1994, Gardner et al, ^[26] using the microsatellite marker D16S422, reported an additional family (SCA4, next in order in the literature) that localizes to 16q24-ter (lod score = 4.08 at theta = 0). This family had prominent sensory axonal neuropathy, cerebellar and pyramidal tract signs, and normal eye movements. Also, ADCA type II having dominantly inherited ataxia and retinal degeneration has been analyzed, and the two known loci for ADCA type I (SCAs 1 and 2) were excluded, as were candidate loci, retinitis pigmentosa 1 locus (RP1), and the genes for rhodopsine and peripherin-rds, responsible for dominantly inherited retinitis pigmentosa ^[27]. Thus ADCA type II is genetically unique although phenotypically similar to ADCA type I (SCAs 1 and 2) families.

Based on these new findings, it is important to emphasize that future classifications of the dominant ataxias should be based on genotype of the family and not its clinical phenotype. This point was brought home at the recent MJD workshop by the report of a SCA1-expanded CAG repeat in an Australian family reported by Nicholson et al, ^[28] not with the typical OPCA-like appearance but rather with MJD features. Lazzarini et al, ^[29] at the same meeting, described a large French family that mapped to the MJD 14q24.3 locus that had three phenotypes in three generations, including the MJD majority phenotype and also a SCA1- (OPCA) type patient and a patient having the spinopontine phenotype. Thus, a genotype causing dominantly inherited ataxia produced a majority phenotype, 6p with SCA1 or 14q with MJD; but it may also express a minor phenotype, 6p with MJD or 14q with a SCA1 phenotype, in some persons in the same family.

An increasing number of new families are being identified on a worldwide basis as having dominantly inherited ataxia associated with unstable triplet repeats. In this issue of Neurology, Genis et al ^[30] describe a large family (M-ADCA1) from Spain having dominantly inherited ataxia type I (SCA1), in which increased CAG triplet repeats were present in affected individuals (41 to 59 repeats) compared with normal control subjects (6 to 39 repeats). Several affected persons with increased repeats had minor signs and symptoms that heralded the full expression of disease of SCA1 by years. Of considerable interest was their finding two individuals in this family who had expanded repeats and were asymptomatic. These at-risk persons do have the mutation and are considered to be preclinical in their stage of the disease process. Determination of the CAG-repeat status of at-risk persons has become a highly significant fact to determine more precisely their potential outcome. There was also an inverse correlation between the number of CAG repeats and the age of onset of the disease in this large family, thus showing anticipation. A more aggressive course of disease was encountered in the children of affected fathers.

Most recently, two additional reports of interest have appeared. Gouw et al ^[31] reported a four-generation family of 42 individuals, with 12 clinically affected with a dominantly inherited ataxia with an associated cone dystrophy. These patients had early loss of color discrimination, with retinal and macular signs followed by a gradual progression of cerebellar function and occurrence of pyramidal signs. Linkage analysis with polymorphic markers D6S89 (for SCA1 on 6p) and D12S79 (for SCA2 on 12q) gave negative results, thus excluding these loci and indicating that an additional new locus exists for this form of dominantly inherited ataxia with retinal degeneration, similar to previous reports of this syndrome ^[27]. Belal et al ^[32] described a Tunisian family with a dominantly inherited ataxia that mapped to the SCA2 locus on 12q that had a phenotype very similar to the original Cuban family except for extrapyramidal signs found in 23% (4/17) of the Tunisian patients but minimally expressed in the Cubans. Clearly SCA2 has a much broader geographic distribution than originally thought, and it is anticipated that additional kindreds will be found in new ethnic groups.

It is highly gratifying to see clarity finally emerge based on these new positional mapping data. CAG unstable repeat expansions in SCA1 and DRPLA join a growing list of inherited neurologic disorders with unstable repeats (Huntington's disease, CAG; myotonic dystrophy, CTG; fragile X syndrome, CGG; Kennedy's syndrome, CAG) ^[33]. MJD has just been added to this list with the report of Kawaguchi et al ^[34] indicating the presence of CAG unstable repeat expansions in MJD patients. The next challenge will be to utilize this information to diagnose directly by mutational analysis persons at risk for developing SCA, MJD, or DRPLA. Finally, identification of the gene products, the ataxins, responsible for each of the dominant ataxias is the next immediate goal that, it is hoped, will lead to therapy to correct a postulated gain of function for mutant gene expression.

In fact, preliminary characterization of ataxin-1 (SCA1) has just been reported by Banfi et al, ^[35] and it is clear that the transcriptional and translational regulation of ataxin-1, the SCA1-encoded protein, is quite complex. The SCA1 transcript is 10,660 bases and is transcribed from both the wild type and SCA1 alleles ^[35]. It is of note that the CAG repeat, which codes for a polyglutamine tract, lies within the coding region. The gene is 450 kb in length and has nine exons, with the first seven exons located in a 5' untranslated region and the last two exons containing the coding region. Further, the first four noncoding exons are alternatively spliced in different tissues. Thus, rapid progress is indeed occurring to characterize the structure of at least one ataxin gene. It will be of considerable interest to learn how variable CAG-repeat expansions, coding for polyglutamine tract lengths, can lead eventually to pathogenic events and produce SCA1 with clinical anticipation.

Most recently, a family has been reported that has two major branches that both descend from the paternal grandparents of President Abraham Lincoln that has dominantly inherited spinocerebellar ataxia (SCA type 5). The gene locus for this disorder has been mapped to the centromeric region of chromosome 11 ^[36].

It is truly remarkable to see the rapid progress that is occurring in this field. A genomic classification (table) has brought order and clarity not achieved previously.