SHORT COMMUNICATION

Cloning and Characterization of the Neural Isoformsof Human Dystonin

ARTHUR BROWN,*,1 GRATIEN DALPE,* MARTINE MATHIEU,* AND RASHMI KOTHARY*,†,2

*Institut du cancer de Montreal, Centre de Recherche L.-C. Simard, and †Department of Medicine,University of Montreal, Montreal, Quebec, Canada

Received May 18, 1995; accepted July 31, 1995

as the gene responsible for the sensory neuropathy, aDystonia musculorum (dt) is a hereditary neurode- targeted mutation in the bpag1 gene in mouse was

generative disease in mice that leads to a sensory recently reported to result in a line of mice that sufferataxia. We have identified and cloned a gene encoded from dystonia musculorum (7). Based on the roleat the dt locus. The product of the dt gene, dystonin, is ascribed to bpag1 in skin and on the role ascribed toa neural isoform of a hemidesmosomal protein bullous actin-binding proteins such as dystrophin, we have pos-pemphigoid antigen 1 (bpag1). To investigate the po- tulated that mutations in dystonin lead to neurodegen-tential role of dystonin in human neuropathies, we eration due to the disruption of the actin and/or neuro-have cloned the neural-specific 5* exons of the human filament networks.DT gene that together with the previously cloned

We have previously mapped the human DT gene toBPAG1 sequences comprise human dystonin. Thechromosome 6p12 (2). In view of the molecular datamouse and human dystonin genes demonstrate theconcerning the mouse dt gene, it is not surprising tosame spectrum of alternatively spliced products, andnote that 6p11–p12 is also the map position for humanthe amino acid sequences of the neural-specific exonsBPAG1 (3). The human BPAG1 gene sequence has beenin the mouse and human genes are over 96% identical.

q 1995 Academic Press, Inc. previously reported (16). To be able to evaluate fullythe potential involvement of DT in human neuropa-thies, we have cloned and sequenced the neural-specific

Mutations in the mouse dystonia musculorum (dt) portions of the human dystonin isoforms.gene result in a progressive degeneration of sensory To clone DT cDNAs, 106 phage clones from a humanneurons (4–6). Behaviorally, these mice suffer from a retinal cDNA library (12) were screened with a 5* por-sensory ataxia and die by the age of weaning. We have tion of the mouse dystonin-2 cDNA that includes thecloned a gene encoded at the dt locus (1). This gene, dystonin N-terminal actin-binding domain (1). Severaldystonin, is expressed in the neurons that degenerate hybridizing clones were isolated, and one clone, Hum-in dt/dt mice and is mutated in two spontaneous dt Ret7, was selected for further investigation. Sequencemutants and in an insertional dt mutant. Sequence analysis, which was carried out by the dideoxy-chainanalysis of the dt cDNAs indicates that at least two termination method (Sequenase Version 2, USB), indi-isoforms of the message exist. These two isoforms are cated that HumRet7 was the human homologue togenerated by alternative splicing of exons at the 5* end mouse dystonin-2; it was subsequently used to screenof the gene. These isoforms have two important fea- a human fetal brain cDNA library (Clontech). Threetures in common. First, they have an N-terminal do- positive phage clones, Hfbr1–3, were isolated and se-main with significant sequence similarity to the actin quenced. Two of these clones, Hfbr1 and Hfbr3, hadbinding domains of dystrophin, a-actinin and b-spec- sequences corresponding to the mouse dystonin-2 iso-trin. Second, both dystonin isoforms have the bullous form, while Hfbr2 had a 3-kb insert that contained se-pemphigoid antigen 1 (bpag1) coding regions at their

quence homologous to the mouse dystonin-1 isoformC-terminus. In support of our assignment of dystoninand stretched far into the coding region of BPAG1. Thepartial nucleotide and predicted amino acid sequences

Sequence data from this article have been deposited with the of the neural-specific dystonin exons for both of theEMBL/GenBank Data Libraries under Accession Nos. U31850 andhuman isoforms are given in Fig. 1.51.

1 Present address: Neurobiology Laboratory, The Salk Institute, This sequence analysis indicates that as in mouse,10010 North Torrey Pines Road, La Jolla, California 92037. there are two human dystonin isoforms. Both isoforms

2 To whom correspondence should be addressed at the Institut du splice into the second exon of BPAG1 and contain ancancer de Montreal, Centre de Recherche L.-C. Simard, Montreal,N-terminal actin-binding domain. These isoforms differP.Q., Canada H2L 4M1. Telephone: (514) 876-6811. Fax: (514) 876-

5476. E-mail: kotharyr@ere.umontreal.ca. from one another by virtue of alternatively spliced ex-

777 GENOMICS 29, 777–780 (1995)0888-7543/95 $12.00

Copyright q 1995 by Academic Press, Inc.All rights of reproduction in any form reserved.

/ m4055$3686 09-18-95 18:42:09 gnma AP-Genomics

SHORT COMMUNICATION778

/ m4055$3686 09-18-95 18:42:09 gnma AP-Genomics

SHORT COMMUNICATION 779

pathological features similar to those seen in dt mice.Specifically, diseases such as Friedreich’s ataxia (14),olivopontocerebellar atrophies (OPCAs) (11), and in-fant onset spinocerebellar ataxia (10) are all sensoryneuropathies characterized by the loss of large myelin-ated fibers. The Friedreich’s ataxia gene has beenmapped to chromosome 9q13–q21 (9), and in additionat least four different genes for different forms of OPCAhave been mapped (8, 13, 15, 17). If a human diseasehomologous to mouse dystonia musculorum exists, ei-ther it is a form of spinocerebellar degeneration thathas yet to be mapped or there may be heterogeneity ina patient population with an already mapped gene. Inany case, the cloning and characterization of the hu-man dystonin gene will allow us and others to assess itspotential role in diseases of the human nervous system.

ACKNOWLEDGMENTS

We thank Dr. Jeremy Nathans for the use of the human retinalcDNA library. This work was supported in part by grants from theDystonia Medical Research Foundation, the Muscular Dystrophy As-sociation of Canada, and the Medical Research Council of Canada.A.B. was supported by a Canderel fellowship. G.D. is supported bya studentship from the Fonds pour la Formation de Chercheurs etl’Aide a la Recherche (Quebec). R.K. is a Medical Research Councilof Canada Scholar and a Chercheur Boursier of the Fonds de laRecherche en Sante du Quebec.FIG. 2. Amino acid sequence comparison of the N-terminal por-

tions of the mouse and human dystonin isoforms. Amino acid identityat any position is indicated by a dash. The mouse and human dys- REFERENCEStonin-1 and dystonin-2 isoforms are 98 and 96% identical, respec-tively, in this region. 1. Brown, A., Bernier, G., Mathieu, M., Rossant, J., and Kothary,

R. (1995). The mouse dystonia musculorum gene is a neuralisoform of bullous pemphigoid antigen 1. Nature Genet. 10: 301–ons at their 5* ends. Based on the size of the dystonin 306.

transcript in mouse, we estimate that less than 1 kb 2. Brown, A., Lemieux, N., Rossant, J., and Kothary, R. (1994).of cDNA sequence remains to be cloned to complete the Human homolog of a mouse sequence from the dystonia muscu-human dystonin-2 isoform. Unfortunately, this isoform lorum locus is on chromosome 6p12. Mamm. Genome 5: 434–

437.has proven itself to be very G/C rich and has been3. Copeland, N. G., Gilbert, D. J., Sawamura, D., Giudice, G. J.,difficult to clone despite attempts to do so by 5* RACE

Chu, M.-L., Jenkins, N. A., and Uitto, J. (1993). Chromosomalprocedures and screens of mouse, human, and chicklocalization of mouse bullous pemphigoid antigens, BPAG1 andcDNA libraries (Ref. 1 and unpublished observations). BPAG2: Identification of a new region of homology between

We are confident, however, that we have cloned the mouse and human chromosomes. Genomics 15: 180–181.entire 5* coding region of dystonin-1, as there is an in- 4. Duchen, L. W. (1976). Dystonia musculorum—An inherited dis-

ease of the nervous system in the mouse. Adv. Neurol. 14: 353–frame stop codon 15 bp upstream from the putative365.start codon. Protein sequence comparisons between the

5. Duchen, L. W., Falconer, D. S., and Strich, S. J. (1963). DystoniaN-terminal portions of the mouse and the human dys-musculorum. A hereditary neuropathy of mice affecting mainlytonin-1 and -2 isoforms reveal 98 and 96% sequencesensory pathways. J. Physiol. (London) 165: 7–9.identity (Fig. 2). This evolutionary conservation pre-

6. Duchen, L. W., and Strich, S. J. (1964). Clinical and pathologicaldicts that these neural-specific exons, which are spliced studies of an hereditary neuropathy in mice. Brain 87: 367–into the BPAG1 gene, must fulfill a specific function in 378.neural tissues. 7. Guo, L., Degenstein, L., Dowling, J., Yu, Q-C., Wollmann, R.,

Perman, B., and Fuchs, E. (1995). Gene targeting of bpag1:Several hereditary human neuropathies have clinico-

FIG. 1. Partial nucleotide and predicted amino acid sequences of (a) the human dystonin-1 isoform and (b) the human dystonin-2isoform. The in-frame stop codon upstream of the first methionine in dystonin-1 is indicated by an asterisk. The double underlined sequenceis a domain with sequence similarity to the actin-binding domains of dystrophin, b-spectrin and a-actinin. The sequence shown in boldfaceis the BPAG1 coding sequence starting from the beginning of its second exon (marked by the open triangle). As the BPAG1 sequence haspreviously been reported (16), we present only a portion of it here and represent the remaining sequence by dotted lines. The closed trianglesmark the position at which the two dystonin isoforms share sequence identity. The nucleotide sequence data reported in this paper havebeen submitted to GenBank under Accession Nos. U31850 and U31851.

/ m4055$3686 09-18-95 18:42:09 gnma AP-Genomics

SHORT COMMUNICATION780

Abnormalities in mechanical strength and cell migration in 13. Orr, H. T., et al. (1993). Expansion of an unstable trinucleotideCAG repeat in spinocerebellar ataxia type 1. Nature Genet. 4:stratified epithelia and neurological degeneration. Cell 81: 233–

243. 221–226.8. Gispert, S., et al. (1993). Chromosomal assignment of the second 14. Ouvrier, R. A., McLeod, J. G., and Conchin, T. E. (1982).

locus for autosomal dominant cerebellar ataxia (SCA2) to chro- Friedreich’s ataxia. Early detection and progression of periph-mosome 12q23–24.1. Nature Genet. 4: 295–299. eral nerve abnormalities. J. Neurol. Sci. 55: 137–145.

9. Hanauer, A., et al. (1990). The Friedreich’s ataxia gene is as- 15. Ranum, L., Schut, L. J., Lundgren, J. K., Orr, H. T., and Living-signed to chromosome 9q13–21 by mapping of tightly linked ston, D. M. (1994). Spinocerebellar ataxia type 5 in a familymarkers and shows linkage disequilibrium with D9S15. Am. J. descended from the grandparents of President Lincoln maps toHum. Genet. 46: 133–137. chromosome 11. Nature Genet. 8: 280–284.

10. Koskinen, T., et al. (1994). Infantile onset spinocerebellar ataxia 16. Sawamura, D., Li, K.-H., Chu, M.-L., and Uitto, J. (1994). Hu-with sensory neuropathy: A new inherited disease. J. Neurol. man bullous pemphigoid antigen (BPAG1): Amino acid se-Sci. 121: 50–56. quences deduced from cloned cDNAs predict biologically im-

portant peptide segments and protein domains. J. Biol. Chem.11. McLeod, J. G., and Evans, W. A. (1981). Peripheral neuropathy266: 17784–17790.in spinocerebellar degenerations. Muscle Nerve 4: 51–61.

12. Nathans, J., Thomas, D., and Hogness, D. S. (1986). Molecular 17. Takiyama, Y., et al. (1993). The gene for Machado–Joseph dis-ease maps to human chromosome 14q. Nature Genet. 4: 300–genetics of human color vision: The genes encoding blue, green,

and red pigments. Science 232: 193–202. 304.

/ m4055$3686 09-18-95 18:42:09 gnma AP-Genomics