10. Strafella AP, Paus T, Fraraccio M, Dagher A. Striatal dopaminerelease induced by repetitive transcranial magnetic stimulation ofthe human motor cortex. Brain 2003;126:2609–2615.

11. Pogarell O, Koch W, Popperl G, Tatsch K, Jakob F, Zwanzger P,Mulert C, Rupprecht R, Moller HJ, Hegerl U, Padberg F. Striataldopamine release after prefrontal repetitive transcranial magneticstimulation in major depression depression: preliminary results ofa dynamic [123I] IBZM SPECT study. J Psychiatr Res 2006;40:307–314.

12. Angelucci F, Oliviero A, Pilato F, Saturno E, Dileone M, VersaceV, Musumeci G, Batocchi AP, Tonali PA, Di Lazzaro V. Trans-cranial magnetic stimulation and BDNF plasma levels in amyotro-phic lateral sclerosis. Euroreport 2004;15:717–720.

13. Maeda F, Keenan JP, Tormos JM, Topka H, Pascual-Leone A.Modulation of corticospinal excitability by repetitive transcranialmagnetic stimulation. Clin Neurophysiol 2000;111:800–805.

14. Thobois S, Hassoun W, Ginovart N, Garcia-Larrea L, Le CavorsinM, Guillouet S, Bonnefoi F, Costes N, Lavenne F, Broussolle E,Leviel V. Effect of sensory stimulus on striatal dopamine release inhumans and cats: a [(11)C]raclopride PET study. Neurosci Lett2004;368:46–51.

15. Strafella AP, Ko JH, Monchi O. Therapeutic application of trans-cranial magnetic stimulation in Parkinson’s disease: the contribu-tion of expectation. Neuroimage 2006;31:1666–1672.

Distinct Distribution of AutosomalDominant Spinocerebellar Ataxia

in the Mexican Population

Elisa Alonso, MD1 Leticia Martınez-Ruano, BS1

Irene De Biase, MD, PhD,2 Christopher Mader, MD,3

Adriana Ochoa, MS,1 Petra Yescas, MS,1

Roxana Gutierrez, BS,4 Misti White, BS,5

Luıs Ruano, MD,6 Marcela Fragoso-Benıtez, MD,7

Tetsuo Ashizawa, MD,5

Sanjay I. Bidichandani, MBBS, PhD,2,8 andAstrid Rasmussen, MD, PhD1,2*

1Department of Neurogenetics and Molecular Biology,Instituto Nacional de Neurologıa y Neurocirugıa Manuel

Velasco Suarez, Mexico City, Mexico; 2Department ofBiochemistry and Molecular Biology, University ofOklahoma Health Sciences Center, Oklahoma City,

Oklahoma, USA; 3Medicine Faculty, Universidad NacionalAutonoma de Mexico, Mexico City, Mexico; 4Chemistry

Faculty, Universidad Nacional Autonoma de Mexico, MexicoCity, Mexico; 5Department of Neurology, The University ofTexas Medical Branch, Galveston, Texas, USA; 6Division ofNeurology, Instituto Nacional de Neurologıa y NeurocirugıaManuel Velasco Suarez, Mexico City, Mexico; 7Faculty of

Medicine, Universidad La Salle, Mexico City, Mexico;8Department of Pediatrics, University of Oklahoma Health

Sciences Center, Oklahoma City, Oklahoma, USA

Abstract: Dominant ataxias show wide geographic varia-tion. We analyzed 108 dominant families and 123 sporadicataxia patients from Mexico for mutations causing SCA1–3,6–8, 10, 12, 17 and DRPLA. Only 18.5% of dominantfamilies remained undiagnosed; SCA2 accounted for half(45.4%), followed by SCA10 (13.9%), SCA3 (12%), SCA7(7.4%), and SCA17 (2.8%). None had SCA1, 6, 8, 12 orDRPLA. Among sporadic cases, 6 had SCA2 (4.9%), and 2had SCA17 (1.6%). In the SCA2 patients we identified 6individuals with the rare (CAG)33 allele, 2 of whom showedearly onset ataxia. The distribution of dominant ataxiamutations in Mexicans is distinct from other populations.© 2007 Movement Disorder Society

Key words: ataxia; autosomal dominant; Mexicanpopulation.

The autosomal dominant cerebellar ataxias (ADCA)are a complex group of progressive neurodegenerative

*Correspondence to: Dr. Astrid Rasmussen, Department of Bio-chemistry and Molecular Biology, University of Oklahoma HealthSciences Center, 975 NE 10th St. BRC 458, Oklahoma City, Oklahoma73104. E-mail: astrid-rasmussen@ouhsc.edu

Received 8 January 2007; Revised 6 February 2007; Accepted 15February 2007

Published online 11 April 2007 in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/mds. 21470

1050 E. ALONSO ET AL.

Movement Disorders, Vol. 22, No. 7, 2007

diseases. So far, 28 genetic loci have been associatedwith ADCA: SCA1–8, 10–28 and DRPLA.1 Wide geo-graphic variations have been observed for the distribu-tion of each SCA subtype. SCA3 is the most frequentADCA in Western Europe, North America, Brazil, andChina, and has been estimated to be the cause of ataxiain 21% of patients worldwide. This is followed by SCA2(15%), SCA6 (15%), and SCA1 (6%). The prevalence ofSCA7 and SCA8 is slightly lower (3–5%) but they arealso widely distributed. Some variants are very rare ornonexistent in most populations, as is the case for SCA10and SCA12.1,2 Notable examples of founder effects havebeen reported, as in the case of Brazil and Portugal wherethere is a very high frequency of SCA3,3 Cuba with alarge cluster of SCA2,4 and the almost exclusive local-ization of DRPLA in Japan.5

In Mexico, previous studies of ADCA patients led tothe identification of the SCA10 mutation, which hassince only been reported in Mexican or Brazilian fami-lies.6–8 Similarly, when we analyzed a large series ofpatients with autosomal recessive and sporadic ataxia formutations in the Friedreich ataxia gene (FXN), we foundthat it is eight times less frequent than in Europeans, andthat most patients remained undiagnosed even after ex-tensive workup.9,10

The distribution of ADCA subtypes in Mexico is notknown, but our previous findings suggested that theMestizo population may be different from other popula-tions. We therefore studied a large series of MexicanMestizo patients with autosomal dominant and sporadicataxia to describe the relative frequencies of the spino-cerebellar ataxias and DRPLA, and to devise an efficientmolecular diagnostic strategy for testing of patients ofMexican origin.

MATERIALS AND METHODS

Our cohort consists of 682 Mexican Mestizo individ-uals: 559 individuals belonging to 108 autosomal dom-inant families and 123 sporadic cases. In all cases, ac-quired causes of ataxia, such as alcoholism, wereexcluded. Friedreich ataxia was excluded in the spo-radics. The control group consisted of 100 unrelatedMexican Mestizos without manifestations of neurologicdisease.

The Mexican Mestizo designation was based on birthin Mexico of all patients, their parents, and at least 3 of4 grandparents. None of the patients were self-describedas belonging to any native Mexican ethnic (Indian) pop-ulations. Informed consent and genetic testing were per-formed as per guidelines of the Institutional ReviewBoard at INNN.

Polymerase chain reaction and gel electrophoresis wascarried out to detect the repeat microsatellite expansionsthat cause SCA1–3, 6–8, 10, 12, 17 and DRPLA usingstandard protocols.

RESULTS

Abnormal repeat expansions were detected in 88 of108 index patients (80.6%; Table 1), which represents arelatively high rate of positive molecular diagnosis forADCA. The most frequent mutation was SCA2 (45.4%),followed by SCA10 (13.9%), SCA3 (12%), SCA7(7.4%), and SCA17 (2.8%) (Table 1). Additionally, wedetected mutations in 8 sporadic patients (6.5%; Table1). No mutations were identified in the remaining 18.5%of ADCA families and 93.5% of sporadic cases. Theclinical features of the analyzed families were similar tothose described in the literature for each SCA variant.

The large number of families with SCA2 allowed us toperform a detailed study of intergenerational instabilityof the repeat in 37 transmissions. There were almostequal instances of transmission via the paternal (n � 17)and maternal (n � 20) germline. Paternal transmissionresulted in more variability in repeat length; rangingfrom �10 to �16 triplet repeats/transmission versusonly �3 to �9 triplet repeats via the maternal germline.The number of parent-child transmissions available forthe other diseases was much lower and were not ana-lyzed further.

While size range of both the normal and expandedSCA2 alleles in our patients and controls are similar toother populations, we found 6 individuals with the(CAG)33 allele. These alleles are rare, and have beenassociated with late onset ataxia, for example, at 60 and86 years in the family reported by Fernandez et al.11

Three of our patients with the (CAG)33 allele were symp-

TABLE 1. Distribution of ADCA mutations in the MexicanMestizo population

ADCA (n � 108)Sporadic

(n � 123)

Expanded allelerange (repeats)

Indexcases

Proportion(%)

Indexcases

Proportion(%)

SCA1 0 0 0 0 —SCA2 49 45.4 6 4.9 36–56SCA3 13 12.0 0 0 66–88SCA6 0 0 0 0 —SCA7 8 7.4 0 0 46–84SCA8 0 0 0 0 —SCA10 15 13.9 0 0 920–4140SCA12 0 0 0 0 —SCA17 3 2.8 2 1.6 50–61DRPLA 0 0 0 0 —Unknown 20 18.5 115 93.5 —

DISTINCT DISTRIBUTION OF ADCA IN MEXICAN POPULATION 1051

Movement Disorders, Vol. 22, No. 7, 2007

tomatic, displaying the typical progressive cerebellarataxia with slow saccades and hyporreflexia as well ascerebellar atrophy and signs of peripheral neuropathy.While 1 had an age of onset at 65 years, 2 had a relativelyearly onset of disease in their 20’s. The other 3 patientswere asymptomatic, but they were �30 years old.

DISCUSSION

Our prior experience with Friedreich ataxia andSCA10 in the Mexican population led us to hypothesizethat the distribution of the ADCA mutations may also beunique. To address this hypothesis, we analyzed a largecohort of ADCA and sporadic ataxia patients and deter-mined the relative frequency of each SCA. Our goal wasto design an efficient and cost-effective molecular diag-nostic testing strategy for ataxia patients of Mexicanorigin.

The proportion of SCA2 (45.4%) is second only to onereport from East India by Sinha et al.,12 who found it tobe the cause of ADCA in 57% of their patients. Simi-larly, SCA2 is relatively common in Italy, ranging from24 to 47% of their cases.13–15 The very elevated fre-quency of SCA2 in the Cuban population is restricted tothe Holguın province, where a founder effect has beendocumented.16 However, SCA2 showed a much lowerfrequency in two Spanish reports (15 and 4.5%).17,18 Thisis interesting, given what is known about the geneticadmixture profile of Mexican Mestizos, which includesmostly Native American, and European (Spanish), with aminor component of African genes.9,19 The observationsthat SCA2 is common in Mexican Mestizos but rela-tively uncommon in the Spanish, and the admixture withSpanish genes, which is estimated to be 13.0 to 24.4%,9

indicate that the SCA2 mutation is likely to be present inthe Native American population.

Analysis of our SCA2 pedigrees indicate that theremay be two different mechanisms that have produced the(CAG)33 allele (Fig. 1). In the autosomal dominant fam-ilies, they are presumably the result of a contraction ofthe expanded parental allele. Such a contraction waspreviously noted in a Cuban SCA2 pedigree.20 More-over, at least in one of the families, the (CAG)33 alleleseems to have arisen from an allele that is relativelyunstable, having mutated from (CAG)37 to (CAG)40 inthe previous generation as well (Fig. 1A). However, inthe 2 sporadic patients, a likely explanation for the originof the (CAG)33 repeat is a de novo expansion of the(CAG)22 allele of one of the parents. A potential caveat,which we have not been able to rule out, is that the lattermay be due to non-paternity.

SCA3, which is the most common ADCA worldwide,2

contributed to only 12% of our families. SCA7, which

accounted for 7.4% of our families, was more commonthan in most populations, with the exception of a Koreanreport (17%),21and a Finnish study (12%).22 However, itis likely that in our cohort, the relatively high proportionof SCA7 may be due to a founder effect, since 6 of the7 families originated from the same geographic area.

SCA17 is rare, accounting for 0 to 1% of ADCAfamilies in several populations,15,23,24 and 1–3% of Hun-tington disease-like patients.24 However, we found that2.8% of all Mexican ADCA families had SCA17 (i.e.3.4% of families with mutations in known genes), as wellas 2 sporadic patients (1.6%). These families originatedfrom different regions of the country and showed widevariation in the size of the expanded alleles, ranging from50 to 61 repeats. Whereas the range of alleles reported inthe literature spans 45 to 66 triplets, the allele with 61repeats in our cohort is the third largest reported to date,and was also associated with childhood onset of thedisease.23,25 We have not yet tested for SCA17 in ourseparate cohort of HD-like patients, but it will be inter-esting to see if the frequency of SCA17 is also higherthan in other populations.

Lack of mutations in SCA12 and DRPLA was notsurprising, since these have been mainly described inIndian and Japanese populations, respectively.1,2 How-ever, SCA1, 6, and 8 are present in most populations,including the Spanish.17,18 The mutational profile inMexicans also differs from the only other Latin-Ameri-can population that has been analyzed to some extent. In

FIG. 1. Three pedigrees with a (CAG)33 allele at the SCA2 locus. Theexpanded/normal allele sizes are shown directly under each individual;age of onset of disease is shown within parentheses and an asteriskdenotes the present age in asymptomatic individuals. (A) ADCA familywith three (CAG)33 alleles: The (CAG)40 allele in IV-1 likely con-tracted to (CAG)33 upon transmission to V-1 whose mother (notshown) is homozygous for the (CAG)22 allele. (B) ADCA family witha (CAG)33 allele that likely arose from the contraction of a (CAG)43

allele (from II-1 to III-2). The mother of III-2 (not shown) has two(CAG)22 normal alleles. (C) Sporadic case with the (CAG)33 allele thatapparently arose via a de novo expansion.

1052 E. ALONSO ET AL.

Movement Disorders, Vol. 22, No. 7, 2007

the Brazilian population, SCA3 contributes to the vastmajority of cases, SCA2 and SCA7 are uncommon, andSCA17 has only been identified once. However, theBrazilian population is the only population other than theMexican to have SCA10.3,8

We have used these data to establish a tiered-diagnos-tic approach in our laboratory: all ADCA and sporadicataxia patients are first assessed for SCA2 mutations, andthen for SCA10, SCA3, SCA7, and SCA17. They areonly tested for SCA1, 6, 8, 12, and DRPLA if theyremain undiagnosed after the first two tiers have beenanalyzed.

The Instituto Nacional de Neurologıa y Neurocirugıais a public reference center for neurological diseases thatserves the uninsured population of central and southernMexico. However, it is the only public hospital in thecountry offering molecular genetic testing for SCAs, andtherefore receives samples from all over Mexico. Wetherefore believe that our patient sample is largely rep-resentative of the Mexican Mestizo population. Our re-sults, which are useful for the ataxia patients in Mexico,will also be useful in the design of testing strategies forthe large Mexican diaspora outside Mexico.

Acknowledgments: This work was supported by grants toAR from Consejo Nacional de Ciencia y Tecnologıa (M30790and SALUD-2003-C01–028), to SIB from the National Insti-tutes of Health (NS047596), Muscular Dystrophy Association,Oklahoma Center for the Advancement of Sciences and Tech-nology (OCAST), and Friedreich Ataxia Research Alliance(FARA), and to TA from the National Institutes of Health(NS041547). We are grateful for the enthusiastic participationof our patients and their families. Dr. Aurelio Jara provided thenormal control samples.

REFERENCES

1. Bird TD. Hereditary ataxia overview. In: GeneReviews: Geneticdisease online reviews at GeneTests-GeneClinics. Seattle: Univer-sity of Washington. Available at http://www.geneclinics.org/. (ac-cessed January 2007).

2. Schols L, Bauer P, Schmidt T, Schulte T, Riess O. Autosomaldominant cerebellar ataxias: clinical features, genetics and patho-genesis. Lancet Neurol 2004;3:291–304.

3. Silveira I, Miranda C, Guimaraes L, et al. Trinucleotide repeats in202 families with ataxia: a small expanded (CAG)n allele at theSCA17 locus. Arch Neurol 2002;59:623–629.

4. Velazquez-Perez L, Santos FN, Garcıa R, Paneque HM, Hecha-varria PR. Epidemiology of Cuban hereditary ataxia. Rev Neurol2001;32:606–611.

5. Tsuji S. Dentatorubral-pallidoluysian atrophy. In: GeneReviews:Genetic disease online reviews at GeneTests-GeneClinics. Seattle:University of Washington. Available at http://www.geneclinics.org/. (accessed January 2007).

6. Matsuura T, Yamagata T, Burgess DL, et al. Large expansion ofthe ATTCT pentanucleotide repeat in spinocerebellar ataxia type10. Nat Genet 2000;26:191–194.

7. Rasmussen A, Matsuura T, Ruano L, et al. Clinical and geneticanalysis of four Mexican families with spinocerebellar ataxia type10. Ann Neurol 2001;50:234–239.

8. Teive HA, Roa BB, Raskin S, et al. Clinical phenotype of Brazilianfamilies with spinocerebellar ataxia 10. Neurology 2004;63:1509–1512.

9. Gomez M, Clark RM, Nath SK, et al. Genetic admixture ofEuropean FRDA genes is the cause of Friedreich ataxia in theMexican population. Genomics 2004;84:779–784.

10. Rasmussen A, Gomez M, Alonso E, Bidichandani SI. Clinicalheterogeneity of recessive ataxia in the Mexican population. J Neu-rol Neurosurg Psychiatry 2006;77:1370–1372.

11. Fernandez M, McClain ME, Martınez RA, et al. Late-onset SCA2:33 CAG repeats are sufficient to cause disease. Neurology 2000;55:569–572.

12. Sinha KK, Worth PF, Jha DK, et al. Autosomal dominant cerebel-lar ataxia: SCA2 is the most frequent mutation in eastern India.J Neurol Neurosurg Psychiatry 2004;75:448–452.

13. Filla A, Mariotti C, Caruso G, et al. Relative frequencies of CAGexpansions in spinocerebellar ataxia and dentatorubropallidoluy-sian atrophy in 116 Italian families. Eur Neurol 2000;44:31–36.

14. Cellini E, Forleo P, Nacmias B, et al. Clinical and genetic analysisof hereditary and sporadic ataxia in central Italy. Brain Res Bull2001;56:363–366.

15. Brusco A, Gellera C, Cagnoli C, et al. Molecular genetics ofhereditary spinocerebellar ataxia: mutation analysis of spinocere-bellar ataxia genes and CAG/CTG repeat expansion detection in225 Italian families. Arch Neurol 2004;61:727–733.

16. Orozco G, Nodarse A, Cordoves R, Auburguer R, Estrada R. Studyof 225 patients with autosomal dominant cerebellar ataxia: pre-sumed founder effect in the province of Holguin, Cuba. In:Plaitakis J, editor. Hereditary ataxias, 1st ed. Berlin: Springer-Verlag; 1989. p 345–360.

17. Pujana MA, Corral J, Gratacos M, et al. Spinocerebellar ataxias inSpanish patients: genetic analysis of familial and sporadic cases.The Ataxia Study Group. Hum Genet 1999;104:516–522.

18. Cabrero DM, Cristobal JH, Duque SC, et al. Distribution of dom-inant hereditary ataxias and Friedreich’s ataxia in the Spanishpopulation. Med Clin (Barc) 2000;115:121–125.

19. Lisker R, Ramırez E, Gonzalez-Villalpando C, et al. Racial ad-mixture in a Mestizo population from Mexico City. Am J HumBiol 1995;7:213–216.

20. Santos N, Aguiar J, Fernandez J, et al. Molecular diagnosis of asample of the Cuban population with spinocerebellar ataxia type 2.Biotecnologia aplicada 1999;16:219–221.

21. Bang OY, Huh K, Lee PH, Kim HJ. Clinical and neuroradiologicalfeatures of patients with spinocerebellar ataxias from Korean kin-dreds. Arch Neurol 2003;60:1566–1574.

22. Juvonen V, Hietala M, Kairisto V, Savontaus ML. The occurrenceof dominant spinocerebellar ataxias among 251 Finnish ataxiapatients and the role of predisposing large normal alleles in agenetically isolated population. Acta Neurol Scand 2005;111:154–162.

23. Koide R, Kobayashi S, Shimohata T, et al. A neurological diseasecaused by an expanded CAG trinucleotide repeat in the TATA-binding protein gene: a new polyglutamine disease? Hum MolGenet 1999;8:2047–2053.

24. Toyoshima Y, Onodera O, Yamada M, et al. Spinocerebellar ataxiatype 17. In: GeneReviews: Genetic disease online reviews atGeneTests-GeneClinics. Seattle: University of Washington. Avail-able at http://www.geneclinics.org/. (accessed January 2007).

25. Maltecca F, Filla A, Castaldo I, et al. Intergenerational instabilityand marked anticipation in SCA17. Neurology 2003;61:1441–1443.

DISTINCT DISTRIBUTION OF ADCA IN MEXICAN POPULATION 1053

Movement Disorders, Vol. 22, No. 7, 2007