ROCKEFELLER MEDICAL LIBRak. INSTITUTE OF NEUROLOGY THE NATIONAL HOSPI7? QUEEN SQUARE, LONDON,
A Molecular Genetic Study of Inherited Movement Disorders
Paul Richard Jarman
A thesis submitted to the University of London for the degree of Doctor of Philosophy September 1998
Institute of Neurology University of London
1 ProQuest Number: 10610917
All rights reserved
INFORMATION TO ALL USERS The quality of this reproduction is dependent upon the quality of the copy submitted.
In the unlikely event that the author did not send a com plete manuscript and there are missing pages, these will be noted. Also, if material had to be removed, a note will indicate the deletion. uest
ProQuest 10610917
Published by ProQuest LLC(2017). Copyright of the Dissertation is held by the Author.
All rights reserved. This work is protected against unauthorized copying under Title 17, United States C ode Microform Edition © ProQuest LLC.
ProQuest LLC. 789 East Eisenhower Parkway P.O. Box 1346 Ann Arbor, Ml 48106- 1346 ABSTRACT
This thesis describes a molecular genetic study of four dominantly inherited movement disorders: paroxysmal dystonic choreoathetosis (PDC), hereditary geniospasm, primary torsion dystonia (PTD) and dopa-responsive dystonia (DRD). The principal methodology employed in the study of these disorders was genetic linkage analysis. Sets of highly polymorphic microsatellite markers were used to map the subchromosomal location of the disorders as the first step in a positional cloning strategy for disease gene identification.
1. Paroxysmal dystonic choreoathetosis. A large British family with PDC was ascertained, and a detailed description of clinical features obtained. A genome-wide search for genetic linkage was performed, and linkage of the PDC gene in this family to markers on the distal long arm of chromosome 2 was confirmed (maximum LOD score 8.7). Fine genetic mapping using closely spaced markers allowed refinement of the candidate region to a 3.8 cM interval. Two putative candidate genes mapping to this region were evaluated using intragenic polymorphisms for linkage analysis.
2. Hereditary geniospasm. Two large families with hereditary geniospasm were ascertained, and one family studied as described above. Linkage of a gene for geniospasm to genetic markers on the proximal long arm of chromosome 9 was found with a maximum
LOD score of 5.24. The gene (GSM1) was mapped to a 2.1 cM interval. Geniospasm in a second family was excluded from this region, indicating genetic heterogeneity.
3. Prim ary torsion dystonia. A large Australian family with PTD was studied but genetic linkage was not observed with any of the markers analysed. Exclusion of dystonia in this family and two other families with PTD, from known PTD loci on chromosomes 8 and 18 indicates the existence of at least one more genetic locus for PTD.
4. Dopa-responsive dystonia. Mutation analysis of the GTP cyclohydrolase I gene (GCH1) was performed in seven patients diagnosed with anticholinergic-responsive PTD who were suspected of having DRD. Three novel GCH1 mutations were identified in two patients, one of whom was a compound heterozygote.
These findings are discussed and future directions of study for identification of the disease genes involved are suggested.
2 CONTENTS
TITLE PAGE 1 ABSTRACT 2 CONTENTS 3 LIST OF TABLES 9 LIST OF FIGURES 10 ABBREVIATIONS 12 ACKNOWLEDGEMENTS 14
CHAPTER 1. INTRODUCTION 15
1.1 STRATEGIES FOR GENE IDENTIFICATION 16 1.2 FUNCTIONAL CLONING 16 1.3 POSITIONAL CLONING 17 1.4 GENETIC LINKAGE ANALYSIS 18 1.4.1 Principles of linkage analysis 18 1.4.2 Markers for genetic mapping 22 1.4.3 Testing for linkage - the LOD score method 24 1.4.4 Linkage disequilibrium 26 1.5 FROM MAP POSITION TO GENE - PHYSICAL MAPPING 26 1.6 THE CANDIDATE GENE APPROACH 28 1.7 MUTATION ANALYSIS OF CANDIDATE GENES 29 1.8 APPLICATION OF MOLECULAR GENETICS TO NEUROLOGICAL DISEASE 30 1.8.1 Application of molecular genetics to the study of movement disorders 31 1.9 OBJECTIVES OF THIS THESIS 32
CHAPTER 2. MATERIALS AND METHODS 34
2.1 OUTLINE OF CHAPTER 34 2.2 DNA EXTRACTION 34 2.3 MEASUREMENT OF DNA CONCENTRATION AND DILUTION OF DNA 35 2.4 LINKAGE ANALYSIS 35 2.4.1 The polymerase chain reaction 36 2.4.1.1 Fluorescence-based PCR. 3 6
3 2.4.1.2 Radioactive PCR. 3 7 2.4.2 Oligonucleotide primers for microsatellite markers 38 2.4.3 Agarose gel electrophoresis 39 2.4.4 Polyacrylamide gel electrophoresis 39 2.4.5 Polyacrylamide gel electrophoresis of fluorescently labelled PCR products 39 2.4.5.1 Polyacrylamide gel preparation 40 2.4.5.2 Pooling of PCR products for loading 40 2.4.5.3 Gel loading and electrophoresis conditions 42 2.4.6 Data analysis for fluorescently labelled PCR products 42 2.4.6.1 Initial data processing 42 2.4.6.2 Genotyping using the Genotyper programme 43 2.4.7 Polyacrylamide gel electrophoresis of radioactively labelled PCR products 43 2.4.8 Computational linkage analyses 45 2.5 MUTATION ANALYSIS 46 2.5.1 Amplification of exons by PCR 46 2.5.2 DNA sequencing 46 2.5.2.1 Cycle sequencing using Taq DNA polymerase 47 2.5.3 Detection of mutations by restriction enzyme analysis 48 2.6 BUFFERS AND SOLUTIONS 48 2.6.1 10xTBE(IL) 48 2.6.2 10 x TE (lOmM Tris/lmM EDTA) (I L) 48 2.6.3 4% acrylamide gel mix for 377 sequencer (50 ml) 48 2.6.4 6% acrylamide gel mix for radioactive gels (500 ml stock solution) 49
CHAPTER 3. A CLINICAL AND GENETIC STUDY OF A FAMILY WITH PAROXYSMAL DYSTONIC CHOREOATHETOSIS 50
3.1 OUTLINE OF CHAPTER 50 3.2 INTRODUCTION 50 3.2.1 The paroxysmal dyskinesias 50 3.2.2 Paroxysmal dystonic choreoathetosis 51 3.2.2.1 Clinical features ofparoxysmal dystonic choreoathetosis 52 3.2.3 Other paroxysmal dyskinesias 56 3.2.3.1 Paroxysmal kinesigenic choreoathetosis 56 3.2.3.2 Paroxysmal exercise-induced dystonia 56 3.2.4 Pathophysiology of paroxysmal dyskinesias 57
4 3.2.5 Rationale for the study 61 3.3 PAROXYSMAL DYSTONIC CHOREOATHETOSIS IN THE ‘B’ FAMILY 61 3.3.1 Clinical features of index cases 63 3.3.1.1 Case V-9 63 3.3.1.2 Case VI-1 65 3.3.2 Clinical features in other family members 65 3.3.2.1 Age at onset 66 3.3.2.2 Attack phenomenology 66 3.3.2.3 Attack frequency 66 3.3.2.4 Attack duration 66 3.3.2.5 Precipitants 67 3.3.2.6 Sleep benefit and other relieving factors 61 3.3.2.7 Diurnal pattern 61 3.3.2.8 Diplopia 68 3.3.2.9 Treatment 68 3.3.2.10 Clinical investigations 68 3.4 GENETIC LINKAGE ANALYSIS IN THE ‘B’ FAMILY - METHODOLOGY 68 3.5 RESULTS OF GENOME-WIDE LINKAGE ANALYSIS 69 3.5.1 PDC in th e ‘B’ family is linked to chromosome 2q 70 3.6 RESULTS OF HAPLOTYPE ANALYSIS OF CHROMOSOME 2q MARKERS 70 3.7 CANDIDATE GENE LINKAGE ANALYSIS 74 3.8 DISCUSSION 75 3.8.1 Comparison with other PDC families 75 3.8.2 Linkage analysis 75 3.8.3 Analysis of candidate genes 77 3.8.3.1 CHRND 11 3.8.3.2 SCL4A3 78 3.8.3.3 5-HT2B is a candidate gene 79 3.9 FUTURE DIRECTIONS OF STUDY 79
CHAPTER 4. LINKAGE ANALYSIS IN FAMILIES WITH HEREDITARY GENIOSPASM
4.1 OUTLINE OF CHAPTER 81 4.2 INTRODUCTION 81 4.2.1 Aims of the study 84 4.3 HEREDITARY GENIOSPASM FAMILIES AND METHODS 84 4.3.1 Family 1 84 4.3.2 Family 2 85
5 4.3.3 Methods 85 4.4 RESULTS OF THE GENOME-WIDE SEARCH FOR THE HEREDITARY GENIOSPASM LOCUS IN FAMILY 1 88 4.5 RESULTS OF HAPLOTYPE ANALYSIS IN FAMILY 1 89 4.6 RESULTS OF LINKAGE ANALYSIS USING CHROMOSOME 9 MARKERS IN FAMILY 2 89 4.7 DISCUSSION 95 4.8 FUTURE DIRECTIONS OF STUDY 97
CHAPTER 5. LINKAGE ANALYSIS IN FAMILIES WITH PRIMARY TORSION DYSTONIA
5.1 OUTLINE OF CHAPTER 99 5.2 AN INTRODUCTION TO PRIMARY TORSION DYSTONIA 99 5.2.1 Clinical features and inheritance of dystonia 100 5.2.1.1 Primary torsion dystonia 102 5.2.1.2 Secondary dystonias 102 5.2.2 The molecular genetic basis of dystonia 104 5.2.2.1 DYT1 106 5.2.2.2 DYT2 107 5.2.2.3 DYT3 108 5.2.2.4 DYT4 108 5.2.2.5 DRD (formerly DYT5) 108 5.2.2.6 DYT6 108 5.2.2.7 DYT7 109 5.3 AIMS OF THE STUDY 109 5.3.1 Does dystonia in families with non-DYTl PTD map to the DYT6 or DYT7 loci? 109 5.3.2 Genetic mapping of the DYT4 locus 110 5.4 PTD FAMILIES AND METHODS 110 5.4.1 Family 1 111 5.4.2 Family 2 111 5.4.3 Family 3 112 5.4.4 DNA analysis 112 5.4.5 Microsatellite markers analysed 117 5.4.6 Linkage analysis 118 5.4.6.1 Exclusion mapping o f the DYT6 and DYT7 loci 118 5.4.6.2 Mapping o f the DYT4 locus 118
6 5.5 RESULTS OF LINKAGE ANALYSIS AT DYT6 AND DYT7 LOCI 118 5.6 RESULTS OF GENOME-WIDE SEARCH FOR THE DYT4 LOCUS 119 5.7 DISCUSSION 127 5.7.1 Exclusion of the DYT6 and DYT7 loci in three PTD families 127 5.7.2 Genome-wide search for the DYT4 locus 129 5.8 FUTURE DIRECTIONS OF STUDY 131 5.8.1 Non-DYTl PTD families 131 5.8.2 Screening for DYT1 mutations in British patients 131
CHAPTER 6. MUTATION ANALYSIS OF GTP CYCLOHYDROLASE I IN PATIENTS WITH DYSTONIA RESPONSIVE TO ANTICHOLINERGIC DRUGS
6.1 OUTLINE OF CHAPTER 133 6.2 INTRODUCTION 133 6.2.1 Dopa-responsive dystonia 133 6.2.2 Anticholinergic-responsive dystonia 134 6.3 AIMS AND DESIGN OF THE STUDY 136 6.4 PATIENTS AND CONTROLS 137 6.5 RESULTS OF GCH1 SEQUENCING AND RESTRICTION ENZYME ANALYSIS 140 6.6 DISCUSSION 145
CHAPTER 7. CONCLUSIONS AND FUTURE DIRECTIONS OF STUDY
7.1 GENERAL CONCLUSIONS 149 7.1.1 Mapping a gene for paroxysmal dystonic choreoathetosis 149 7.1.2 Mapping a gene for hereditary geniospasm 150 7.1.3 Linkage studies in primary torsion dystonia 150 7.1.4 GCH1 mutation analysis in patients with anticholinergic-responsive dystonia 150 7.2 FUTURE DIRECTIONS OF STUDY 151
APPENDICES
APPENDIX 1. RESULTS OF TWO-POINT LINKAGE ANALYSIS IN THE ‘B’ FAMILY 153 APPENDIX 2. RESULTS OF TWO-POINT LINKAGE ANALYSIS IN HEREDITARY GENIOSPASM - FAMILY 1 156
7 APPENDIX 3. RESULTS OF TWO-POINT LINKAGE ANALYSIS AT THE DYT6 AND DYT7 LOCI 163 APPENDIX 4. RESULTS OF GENOME-WIDE SEARCH FOR THE DYT4 LOCUS 168
REFERENCES 184
PUBLICATIONS RESULTING FROM WORK PRESENTED IN THIS THESIS 209
8 TABLES
Table 3.1: Summary of clinical features of 12 families with PDC including the ‘B’ family 53-54 Table 3.2: Core features of familial paroxysmal dystonic choreoathetosis 55 Table 3.3: Results of two-point LOD scores between PDC and chromosome 2q markers in the ‘B’ family 72 Table 3.4: Results of two-point LOD scores between PDC and candidate gene polymorphisms in the ‘B’ family 72 Table 4.1: Two-point LOD scores between hereditary geniospasm in family 1 and microsatellite markers on chromosome 9. Penetrance of geniospasm set at 80% 90 Table 4.2: Two-point LOD scores between hereditary geniospasm in family 1 and microsatellite markers on chromosome 9. Penetrance of geniospasm set at 100% 91 Table 4.3: Two-point LOD scores between hereditary geniospasm in family 2 and microsatellite markers spanning GSM1 candidate interval. Penetrance set at 80% 92 Table 4.4: Genes mapping to 9ql3-21 98 Table 5.1: Classification of dystonia by distribution 101 Table 5.2: Selected genetic causes of secondary dystonia 103 Table 5.3: Genetic loci for the primary dystonias 105 Table 5.4: Summary of selected clinical features in DYT6/DYT7 kindreds and the PTD families studied 116 Table 5.5: Summary of microsatellite maker numbers and density at DYT6/DYT7 loci 117 Table 6.1: Summary of clinical features of index cases studied and GCH1 mutations identified 139 Table 6.2: Summary of GCH1 mutations identified and resulting restriction site changes 141
9 FIGURES
Figure 1.1: Schematic representation of the positional cloning paradigm 19 Figure 2.1: Illustration of a typical Genescan run 41 Figure 2.2: Illustration of allele sizing using the Genotyper programme 44 Figure 3.1: Basal ganglia motor circuits 58 Figure 3.2: The ‘B’ family 62 Figure 3.3: Genetic map of chromosome 2q showing selected microsatellite markers and PDC candidate regions 71 Figure 3.4: Chromosome 2q haplotypes in the ‘B’ family 73 Figure 4.1: Hereditary geniospasm - family 1 86 Figure 4.2: Hereditary geniospasm - family 2 87 Figure 4.3: Genetic map of the pericentromeric part of chromosome 9 showing selected microsatellite markers and GSM1candidate region 93 Figure 4.4: Chromosome 9q haplotypes in family 1 94 Figure 5.1: Primary torsion dystonia - family 1 113 Figure 5.2: Primary torsion dystonia - family 2 114 Figure 5.3: Primary torsion dystonia - family 3 115 Figure 5.4: Multipoint LOD score graph for selected markers at the DYT6 locus in family 1 120 Figure 5.5: Multipoint LOD score graph for selected markers at the DYT7 locus in family 1 121 Figure 5.6: Multipoint LOD score graph for selected markers at the DYT6 locus in family 2 122 Figure 5.7: Multipoint LOD score graph for selected markers at the DYT7 locus in family 2 123 Figure 5.8: Multipoint LOD score graph for selected markers at the DYT6 locus in family 3 124 Figure 5.9: Multipoint LOD score graph for selected markers at the DYT7 locus in family 3 125 Figure 5.10: Chromosome 19 haplotypes in family 3 126
10 Figure 6.1: The tetrahydrobiopterin synthesis pathway and its relationship to the synthesis of dopamine Figure 6.2: Pedigrees of index patients with anticholinergic-responsive dystonia Figure 6.3: Sequencing electropherogram showing an A -> T transversion in exon 6 of GCH1 in patient 2 Figure 6.4: Sequencing electropherogram showing a C -» T transition in exon 1 of GCH1 in patient 2 Figure 6.5: Sequencing electropherogram showing a G -> A transition in exon 1 of GCH1 in patient 7 ABBREVIATIONS
A Adenine ADNFLE Autosomal dominant nocturnal frontal lobe epilepsy APS Ammonium persulphate Asn Asparagine Asp Aspartic acid bp Base pair C Cytosine cDNA Complementary DNA CEPH Centre d’Etudes du Polmorphisme Humaine CHLC Cooperative Human Linkage Centre CHRND Delta subunit of the nicotinic acetylcholine receptor cM CentiMorgan CNS Central nervous system CT Computerised tomography scan dNTP Deoxynucleoside triphosphate ddNTP Dideoxynucleoside triphosphate DNA Deoxyribonucleic acid DRD Dopa-responsive dystonia DYT 1 - 7 Dystonia gene symbols 1 -7 EDTA Ethylene diamine tetraacetate EEG Electroencephalogram FPD1 Familial paroxysmal dyskinesia, locus 1 (alias: PDC) G Guanine GCH1 GTP cyclohydrolase gene GDB Genome database GSM1 Hereditary geniospasm, locus 1 HIAA 5-hydroxyindoleacetic acid 5-FIT 5-hydroxytryptamine http hypertext transmission protocol HVA Homovanillic acid
12 kb Kilobase Leu Leucine LOD Logarithm of odds ratio Lys Lysine Mb Megabase MRS Magnetic resonance spectroscopy OD Optical density OMIM On-line Mendelian Inheritance in Man database PCR Polymerase chain reaction PDC Paroxysmal dystonic choreoathetosis PED Paroxysmal exercise-induced dystonia PKC Paroxysmal kinesigenic choreoathetosis Pro Proline PTD Primary torsion dystonia RFLP Restriction fragment length polymorphism SCL4A3 Anion exchanger subtype, gene symbol SNP Single nucleotide polymorphism T Thymine TBE Tris-borate EDTA solution TE Tris EDTA solution TEMED N, N, N ’, N’ - tetramethylethylenediamine WD Wilson’s disease Z LOD score
13 ACKNOWLEDGEMENTS
Several people have made it possible for me to carry out the work described in this thesis. Mary Davis initiated me into the mysteries of linkage analysis and instructed me in computational linkage analysis with great patience. She also gave me an enthusiasm for genetics which I will retain long after I have forgotten how to use MLINK. Mary Sweeney, Tom Warner, Lee Williams, Martin Brockington and Cathy Woodward faced the uphill struggle involved in explaining to me the rudiments of laboratory work and the need to clear up the resulting mess. Toby Nygaard kindly worked on the multipoints with a keen eye for detail. The Wellcome Trust provided funding in the form of a Research Training Fellowship. Professor Anita Harding kindled my interest in molecular genetics and planned this project. She was a source of inspiration and will be greatly missed. Nick Wood ‘inherited’ me and has guided me through to the finishing line with encouragement, enthusiasm and good humour. He has always been generous with his time during a busy period. I am very grateful for his support and supervision, although I still struggle to read his handwriting. Above all I wish to thank my wife, Jenna, without whose unfailing support, strength, and endless patience with me during the ups and downs of research I could not have completed this work. Her editing skills have turned my random use of punctuation into something resembling the English language. During the time it has taken me to write this thesis, our daughter Emily has learnt to walk and talk. I am grateful to her for helpful discussions of the finer points of animal noises and for helpfully reorganising my papers from time to time. I attribute all typos to her little fingers bashing away at this keyboard.
14 CHAPTER ONE
INTRODUCTION
The discovery of human genes and the mutations within these genes which result in disease is of fundamental importance to genetics and medicine. The ability of researchers to isolate and clone the genes underpinning human disease is the central tenet of the doctrine of ‘the new genetics’ (Comings, 1979). It has led to insights into the underlying biological mechanisms of disease, the development of molecular diagnostic and predictive testing, and may ultimately enable the development of rational forms of treatment, including, it is hoped, gene therapy. It is no exaggeration to assert that ‘the new genetics’ has revolutionised both genetics and medicine (Collins, 1992).
The human genome contains approximately 3 x 109base pairs of DNA. Identification of a point mutation in a gene, with no prior knowledge about its function, may be compared to searching for a needle in a haystack. Technological and methodological advances over the last two decades, however, have paved the way for the molecular characterisation of many disease genes. The availability of increasingly dense and polymorphic genetic maps of the human genome, in tandem with advances in the analysis of the markers comprising these maps, have made possible an exponential increase in the numbers of disease genes isolated. In 1992 some 13 disease genes had been positionally cloned; by late 1995 more than 60 had been cloned based on their position, and more than 500 genes had been mapped to specific chromosomal locations (Collins, 1995; Lander and Kruglyak, 1995). At the time of writing, few important monogenic diseases remain to be mapped, and the genes for many have now been fully characterised.
This chapter reviews the strategies available for gene identification, with particular emphasis on the technique of genetic linkage analysis, which forms the basis of much of the work described in this thesis.
15 1.1 STRATEGIES FOR GENE IDENTIFICATION
Four general strategies exist for identification of disease genes: functional cloning, positional cloning, positional candidate approaches and position-independent candidate approaches. The suitability of each approach depends on the degree of understanding of the molecular pathology of the disease, the availability for study of patients and families with the disease, and the completeness of the available gene m ap.
1.2 FUNCTIONAL CLONING
Functional cloning makes use of fundamental information about the function of a disease gene, such as the presumed biochemical defect, to identify the gene without reference to its chromosomal map position. In most cases this approach relies on demonstration of an abnormal or deficient protein and subsequent purification of the protein. Amino acid sequencing of the protein allows complementary oligonucleotides to be synthesised and used to probe cDNA libraries. Once a cDNA fragment containing the gene of interest is identified, this may be used to screen genomic DNA to isolate the disease gene. An alternative approach makes use of a monoclonal antibody raised against the purified protein to screen expression vectors transfected with cDNA fragments, one of which may express the disease-associated protein. Functional complementation of mutant yeast strains with defined biochemical defects by human DNA fragments is also used as a means of gene identification.
Functional cloning was the first method of gene identification to be developed and has been employed to identify many disease genes including the genes for phenylketonuria (Robson et al, 1982) and factor VIII (Gitschier et al, 1984). However, functional cloning is critically dependant on a high level of prior understanding of the molecular pathology of a disease in order to predict the abnormal protein involved. Unfortunately, for the vast majority of single gene disorders, such detailed insights about the basic gene defect are not available, and this limits the application of the technique.
16 1.3 POSITIONAL CLONING
Since the early 1980s the strategy of positional cloning has emerged to identify genes for diseases where little or nothing is known about the nature of the disease gene product. The term positional cloning refers to the isolation of a gene solely on the basis of its chromosomal map position. The principle upon which positional cloning is based is the initial localisation of the gene to a specific subchromosomal region, followed by successive narrowing of this candidate interval, which eventually results in the location of the gene itself. The initial localisation of the gene is usually achieved using the technique of genetic linkage analysis, described below (section 1.4). Alternatively, this information may be derived by other means, such as loss of heterozygosity screening in the case of genes involved in oncogenesis, or more commonly, the identification of cytogenetic abnormalities, such as chromosomal deletions or translocations, which have been used to pinpoint disease genes with great accuracy.
The first success of the positional cloning technique was the identification of the gene for chronic granulomatous disease on the X chromosome in 1986 (Royer Pokora et al, 1986). Many of the first disease genes to be mapped were on the X chromosome because it was possible to localise the gene to this chromosome on the basis of an X-linked inheritance pattern, and because cytogenetically visible deletions and translocations involving the X chromosome frequently result in a clinical phenotype in males due to the haploid status of the X chromosome. A substantial proportion of positional cloning successes have in fact relied on fortuitous identification of chromosomal rearrangements in individual patients. Before the isolation of the genes for neurofibromatosis type 2 and Huntington’s disease in 1993 (Rouleau et al, 1993; Huntington's disease collaborative research group, 1993), only one gene - the cystic fibrosis gene (Kerem et al, 1989; Rommens et al, 1989) - had been identified by positional cloning without the aid of cytogenetic rearrangements to narrow the field of search. Despite the difficulties, the positional cloning approach has resulted in numerous successes and is now considered a standard means of gene identification, although the more efficient positional candidate approach (discussed in section 1.6) looks set to become the prevailing method of gene cloning.
17 The paradigm employed for positional cloning of a disease gene is illustrated in schematic form in figure 1.1. In the absence of visible cytogenetic rearrangements to locate a gene, families in which the disease segregates are recruited and studied using genetic markers, with the aim of identifying a marker linked to the disease gene. Fine genetic mapping follows, to assign the gene to as small a genetic region as possible between defined markers. The resolution achieved by genetic mapping depends upon the number of informative meioses available in the pedigree material. In practice, it is unusual to be able to localise a gene to a smaller genetic interval than one centimorgan (cM), defined as the genetic distance over which two loci will be separated by recombination in 1% of meioses. Often, however, the candidate region containing the gene is larger. The minimum genetic localisation required for success in a purely positional cloning strategy is considered to be approximately 2-3 cM. It should be appreciated that genetic and physical distances are not linearly related, but that 1 cM corresponds to approximately 1 million base pairs (Mb) of DNA. A region of this size may contain a substantial number of genes; it is estimated that there are an average of 20-25 genes per Mb, throughout the human genome (Fields et al, 1994). Cloning within a candidate region using the techniques of physical mapping is an expensive and time-consuming endeavour. Thus, reduction of the size of the genetic region to which a gene is mapped at the outset of a positional cloning project is of paramount importance.
1.4 GENETIC LINKAGE ANALYSIS
Genetic linkage analysis is the first step in the positional cloning and positional candidate approaches to gene identification; the success of both are dependent to a large degree on the exactness of initial genetic mapping. Gene mapping by linkage analysis forms the basis of much of the work described in this thesis.
1.4.1 Principles of linkage analysis
Genetic linkage analysis makes use of the exception to Mendel’s Law of Independent Assortment which states that alleles at different genetic loci assort at random during
18 Figure 1.1: Schematic representation of the positional cloning paradigm. The steps by which it is possible to map a disease gene and 43 11 f i VO Td • 3 T m ON O S nd m 'd ’55 .2 .2 ■*-> *3 3 T td td cd O 03 o £ o 3 4 l - I
For closely linked loci (where 0 < 0.05-0.1), it is reasonable to assume that the probability of more than one recombination occurring between the loci is small. In these circumstances the recombination fraction is equal to the genetic map distance between the loci, thus two loci showing recombination in 1% of meioses (0 = 0.01) are approximately 1 cM apart. For more distant loci, however, multiple crossovers may occur which are not ‘seen’ by analysis of recombinants, leading to underestimation of genetic distance. The non-linear relationship between genetic distance and recombination frequency is described by mapping functions, such as
20 Haldane’s function (Haldane, 1919) which is required to transform 0 into additive map distance:
co = -V2 In (1 - 26)
or,
0 = V2 [1 - exp(-2co)],
where co is the map distance in cM. However, Haldane’s function assumes, falsely, that crossovers are randomly distributed along the chromosome. In fact, the phenomenon known as interference inhibits the formation of crossovers in the vicinity of an existing crossover. Mapping functions allowing for this phenomenon, such as that described by Kosambi (1944), may be more accurate.
Ultimately, it is the physical position of loci on the DNA molecule and the distance between loci measured in base pairs of DNA that is of greatest interest to researchers attempting to clone genes. The relationship between genetic and physical distance, however, is not constant but depends on the recombination rate which in turn varies quite considerably both between chromosomal regions and between males and females. However, if a total genetic map length of approximately 3,000 cM in man is assumed, this corresponds on average to 1 Mb of DNA per 1 cM of genetic distance in the 3 x 109 bp human genome.
For the purpose of localisation of disease genes, linkage analysis between two loci - the disease gene locus and anonymous DNA marker loci - is performed to determine whether the two co-segregate and are therefore in close physical proximity. Since the sub-chromosomal location of markers is known, the position of linked genes within the framework of the genetic map may be deduced. There are a number of prerequisites for linkage mapping of diseases. It is self-evident that the aetiology of the disease must have a genetic component; predominantly non-genetic diseases, such as vertically transmissible infectious diseases, may be mistaken for genetic disorders. Ideally, the disease should be monogenic (i.e. caused by a single disease gene) and the mode of inheritance established by segregation analysis. If these criteria are not met, for example in the case of polygenic diseases caused by the interaction of multiple disease-susceptibility genes with one another and with
21 environmental factors, model-free methods of analysis, such as the affected sibling pair method, may be employed (Ott, 1996). However, a full discussion of non- parametric methods of linkage analysis is beyond the scope of this thesis. Clinical ascertainment of pedigrees in which the disease studied is segregating is of fundamental importance, and may, to a large extent, determine the success of a linkage study. Identification of large pedigrees with numerous affected individuals is desirable in order to maximise the power of the study to detect linkage. In practice, large pedigrees are often not available, particularly when studying recessive disorders. Smaller families may then be grouped together for linkage analysis, but locus heterogeneity (i.e. different disease genes resulting in indistinguishable phenotypes) is a commonly encountered problem which complicates this approach. Correct phenotypic designation of affected and unaffected status is an essential requirement for successful linkage analysis, and should be decided at the outset of the study in order to minimise bias. Finally, linkage analysis requires the use of polymorphic genetic markers which are discussed in the next section.
1.4.2 Markers for genetic mapping
The tools needed for genetic mapping in humans have only become available during the last two decades. Markers used for mapping must fulfil three requirements: they must be polymorphic (two, or preferably more, common alleles in the population), their chromosomal location must be known, and it must be easy to type the marker (i.e. to distinguish between alleles). Initial attempts at mapping relied on the use of easily typed protein polymorphisms such as blood group proteins and classical HLA antigens, but as the chromosomal location of these proteins was not known linkage groups could be established but the diseases could not be mapped to a specific autosome. It was not until the 1980s that a comprehensive array of human DNA markers, in the form of restriction fragment length polymorphisms (RFLPs), became available (Botstein et al, 1980). The usefulness of RFLPs in genetic mapping is limited by their biallelic nature (i.e. the presence or absence of a restriction enzyme cleavage site) which results in low heterozygosity (50% maximum), and by the labour-intensive methodology required to type large numbers. Although RFLPs are still useful in genetic mapping, particularly when typed with the aid of the
22 polymerase chain reaction (PCR), they have been largely superseded in linkage projects by variable number tandem repeat markers (VNTRs) such as microsatellites.
A microsatellite is a genomic sequence in noncoding DNA consisting of a mono-, di, tri-, or tetranucleotide, repeated in multiple tandem copies. The number of repeat copies is usually highly variable between individuals in a population, relatively stable between generations, and inherited in a mendelian fashion. Microsatellites are therefore extremely useful as genetic markers, particularly since they are abundant - CA dinucleotide repeats occur on average every 18-28 kb (Stallings et al, 1991), dispersed throughout the genome, and may be easily typed by PCR (Weber and May, 1989; Silver, 1992). Amplification of a short segment of DNA using PCR primers situated in non-repetitive DNA surrounding a microsatellite, produces multiple copies of a DNA fragment which may be sized by electrophoresis on a high resolution polyacrylamide gel, thus distinguishing between alleles with differing repeat numbers (see chapter 2). Systematic identification of microsatellites and careful ordering of these relative to one another in the genome by observation of recombinations in large pedigrees has enabled comprehensive genetic marker maps to be constructed for humans and several model organisms. The first phase of the Human Genome Project, completed in 1996, was the construction of a comprehensive high resolution genetic map of the human genome based on over 5,000 microsatellites with an average spacing of 1.6 cM and an average heterozygosity of 70% (Dib et al, 1996). The advent of such maps, along with the technology for large-scale, rapid typing of markers, has made systematic disease linkage analysis possible, and provided a framework which is a prerequisite for construction of physical maps of the human genome and ultimately for complete genomic sequencing.
The most common type of human genetic variation is the single nucleotide polymorphism (SNP), a variation between two alternative bases occurring at a single position. SNPs are highly abundant, occurring on average at least once per kb (Wang et al, 1998). When the two alternative bases both occur at appreciable frequency in the population SNPs may be used as biallelic genetic markers. Although SNP maps are at an early stage of development they have the potential to supersede microsatellite markers for mapping. Genotyping large numbers of microsatellite
23 markers for linkage analysis is both costly and time-consuming. Unlike microsatellites which require gel-based methods to distinguish between alleles on the basis of length, SNPs are more amenable to automated genotyping using new technologies such as hybridisation to high density DNA probe arrays known as ‘DNA chips’ (Goffeau, 1997; Wang et al, 1998). Although individually less informative than microsatellites, the potential for rapid high throughput genotyping, and the predicted availability of SNP maps with a marker density much greater than 1 per cM, are likely to give SNPs significant advantages over current genetic markers (Jordan and Collins, 1996; Kruglyak, 1997).
1.4.3 Testing for linkage - the LOD score method
The direct method of linkage analysis depends upon observing and counting recombinants and nonrecombinants to directly calculate 0 (Ott, 1991). In human genetics, however, the information needed to unambiguously identify the position of crossovers is often lacking, due to unknown phase and incomplete penetrance, for example, and an indirect statistical method is generally used. The most frequently used method is the likelihood method, which calculates the likelihood (L), that the observed genotypes could have arisen from two linked loci at a given value of 0, against the likelihood of observing the same genotypes for unlinked genes (L(0.5) - the null hypothesis of independent assortment). When the ratio of L(0)/L(O.5) is greater than 1, it indicates that linkage is more likely than non-linkage to explain the observed genotypes. This odds ratio is usually expressed as a decimal logarithm, called a LOD score (for Log of ODds ratio), in order that scores from different families may be summed (Ott, 1991).
Z(0) = loglo[L(0)/L(O.5)] where Z is the LOD score at a given value of 0. The LOD score is calculated for various values of 0 using computer programs such as MLINK (Lathrop and Lalouel, 1988; Cottingham et al, 1993; Schaffer et al, 1994) to obtain the value of 0 associated with the highest LOD score. This provides an estimate of the genetic distance between the two loci studied, along with a measure of the weight of the data in favour of the hypothesis of linkage (Ott, 1991). Calculation of LOD scores requires
24 specification of the mode of inheritance, the penetrance of the disease gene and estimation of the disease gene frequency.
Positive LOD scores indicate evidence for linkage and negative LOD scores indicate absence of linkage. The critical value considered to indicate significant evidence of linkage is Z = ^ 3. Conversely, linkage can be rejected if the LOD score < - 2. Although a LOD score at the Z = 3 threshold indicates an odds ratio of 1000:1 in favour of the data being explained by the hypothesis of linkage, it should be appreciated that this corresponds to a conventional probability of linkage at the p = 0.05 level of significance. This can be explained by the prior probability of linkage between any two randomly chosen loci (the probability that they are close enough for linkage to be detected) which is estimated to be in the order of only 0.02.
Linkage analysis between two loci is a powerful technique for the localisation of disease genes. The technique depends upon pairwise linkage analysis between a disease locus and each of a large number of genetic markers. In practice genome- wide searches for linkage usually involve genotyping family members from one or more affected pedigrees, using a panel of microsatellite markers spaced evenly throughout the genome. Panels of microsatellite markers designed specifically for this purpose are available. These consist of markers selected on the basis of high heterozygosity and reliable amplification using the polymerase chain reaction (PCR). The density of markers screened in a genome-wide search for linkage will depend on the estimated maximum LOD score of the pedigrees studied. Once genetic linkage between a marker and the disease is identified, additional fine mapping may be undertaken, using closely spaced markers in the area of interest. This allows informative flanking recombinations between markers and disease to be identified by reconstruction of chromosomal haplotypes in family members. A disease gene can then be assigned to the genetic interval between the two closest flanking recombinations identified.
Pairwise analysis may be complemented by multilocus (or multipoint) linkage mapping, wherein three or more loci may be considered simultaneously. This technique is particularly useful for determination of probable marker order for a series of linked loci when constructing genetic maps, and is also helpful for exclusion mapping of disease loci to fully exclude a disease locus from the vicinity of a set of
25 markers. Multipoint analysis between closely spaced loci yields more information than serial pairwise analysis and may therefore be helpful in a region where pairwise scores are equivocal.
1.4.4 Linkage disequilibrium
While linkage refers to the relationship between two loci, the term linkage diseqilibrium is used to describe the non-random association of alleles at linked loci. When a new mutation creating a disease allele is first introduced into a population it occurs on a chromosome with a unique combination of surrounding alleles (a haplotype). During subsequent generations the original characteristic haplotype will be altered by random recombination, but the closer an allele to the disease allele, the greater the likelihood that these alleles will co-segregate through subsequent meioses. It follows that diseases which have originated relatively recently in a population (in generational terms) will be associated with a unique surrounding haplotype inherited from the founder. The more recent the common disease origin, the greater the genetic distance over which patients will share a common haplotype. These considerations make analysis of linkage disequilibrium a useful tool for high resolution genetic mapping since the technique provides indirect information on a far greater number of recombinations than may be observed in any pedigree-based linkage study. In practice it is possible to map genes at a resolution of less than 1 cM by the identification of markers that are in strong linkage disequilibrium with the disease allele (Xiong and Guo, 1997).
1.5 FROM MAP POSITION TO GENE - PHYSICAL MAPPING
Genetic mapping allows a disease gene to be localised within the framework of a genetic marker map. Physical mapping refers to the process of localisation of the gene of interest to a specific position on the DNA molecule. The genetic and physical genome maps have been partially united to provide integrated maps containing a selection of microsatellites and other sequence tagged sites such as expressed sequence tags (ESTs -which are fragments of cDNA sequence most of which have been mapped to radiation hybrid or other physical maps) ordered relative to one another (Cox and Myers, 1996; Schuler et al, 1996). Nevertheless, the transition from knowledge of the genetic map position of a gene to its exact physical position is a
26 laborious process. Characterisation of the gene for Huntington’s disease, for example, followed some ten years after initial linkage was established (Gusella et al, 1983; Huntington's disease collaborative research group, 1993). An understanding of the techniques used for physical mapping is necessary to appreciate the scale of the undertaking required in moving from genetic map position to final gene identification.
The principle of physical mapping for the purposes of positional cloning rests upon the construction of an ordered series of overlapping, cloned DNA fragments spanning the candidate region - known as a contig. Genes within the contig, any of which may be disease gene candidates, are then systematically identified and screened for disease-associated mutations. A variety of cloning vectors are available for this purpose, each of which typically carries a different sized insert. These include yeast artificial chromosomes (YACs - inserts of up to 2 Mb), the bacteriophage PI (70-100 kb inserts) and plasmids containing cos sequences capable of being introduced into E. coli cells by the bacteriophage lambda (cosmids - up to 50 kb inserts). The choice of vector depends upon the contig resolution required, as well as a number of technical factors. A hierarchy of contigs of increasing resolution is often constructed to cover large areas. Initial clones from the area of interest may be identified by screening DNA libraries for clones containing the microsatellite markers close to which the gene has been mapped. The contig may then be extended by creating probes from the end sequences of clones and using these to identify overlapping fragments (a process called chromosome walking), or by ‘fingerprinting’ of random clones to identify overlapping sequences. Whichever technique is used, the construction of a contig is highly labour-intensive despite increasing automation of the process. It is sometimes possible to bypass this stage by making use of pre existing contigs of the area, produced as part of the Human Genome Project strategy for genome sequencing and available through public databases.
Assembly of the contig is followed by identification of coding sequences of genes from the clones, a process termed transcript mapping. A wide variety of methods have been developed for transcript mapping, including ingenious techniques such as exon trapping (Church et al, 1994) which makes use of splice sites flanking exons to express cloned exons in cell culture, as well as more straightforward approaches such
27 as simple insert sequencing. Transcripts identified in the contig are often fragments of genes which may subsequently be fully characterised. Analysis of tissue expression patterns and putative function of novel genes as well as mutation analysis in patients are required to verify whether any gene identified is the cause of the disease studied.
1.6 THE CANDIDATE GENE APPROACH
As the Human Genome Project nears completion, the base pair sequence of an increasing proportion of the human genome is becoming available. As a result, an increasing number of the estimated 60,000 - 80,000 human genes (Fields et al, 1994) are being identified and placed on the gene map (Genome Database at http://www.gdb.org/). The difficulties of a pure positional cloning approach may sometimes be circumvented by directly testing previously characterised genes which are considered to be putative disease-causing genes. Genes may be considered candidates either by virtue of their chromosomal position in the vicinity of the disease locus, or because the function of the gene is such that it may be envisaged to play a part in the pathophysiology of the disease, or preferably both. The most frequently used is the positional candidate approach, which is dependant upon knowledge of the genetic map position of the disease locus and of the genes previously assigned to this region. Some appreciation of the molecular pathology of the disorder being studied is also required. This strategy is much more efficient than pure positional cloning and has already resulted in numerous successes (Collins, 1995). The positional candidate approach looks likely to completely supersede positional cloning as the human gene map is completed.
If no suitable candidate genes are known to exist in the vicinity of a disease locus, syntenic chromosomal regions of model organisms such as the mouse may be searched for previously mapped candidate genes. Genes identified in this way are particularly convincing candidates if they are known to be associated with a mouse phenotype having features in common with the human disease. Such a situation was encountered in the case of Charcot-Marie-Tooth disease type 1A (CMT1 A) which was known to be associated with duplications on chromosome 17p, syntenic to mouse chromosome 11. Subsequently, mutations in the peripheral myelin protein 22
28 gene (PMP22) in the Trembler mouse (a model of CMT1 A) led to identification of PMP22 mutations and duplications in CMT1A patients (Scherer and Chance, 1995).
When a predicted pathophysiological link between a disease and a known gene is sufficiently convincing, it is possible to screen the gene directly for mutations in patients, without first performing linkage analysis to map the disease locus. This approach has been termed the position-independent candidate gene approach.
1.7 MUTATION ANALYSIS OF CANDIDATE GENES
Regardless of the approach employed to identify putative disease genes, disease- causing mutations must be demonstrated in patients to confirm the identity of a disease gene. The most widely applied method of mutation screening is by genomic DNA sequencing of the candidate gene. Differences in sequence between patients and controls may represent pathogenic mutations. Alternatively, sequence variants may be simple neutral polymorphisms of no pathological significance. Distinguishing between these two alternatives is essential. Features suggesting a pathogenic mutation include: large-scale deletion or rearrangements involving a gene, stop codon or frameshift mutations, non-synonymous changes (altering the amino acid specified by a codon), non-conservative amino acid substitutions (substitution of one amino acid by another with different chemical properties) and changes in amino acid residues highly conserved throughout evolution (suggesting an important role in protein function). In addition, putative mutations should segregate with the disease within a family and other mutations in the same gene may be identifiable in unrelated families; in some instances a single mutation may be associated with disease in unrelated families and even in different species, strengthening the case for a pathogenic role. The final requirement when studying dominantly inherited disorders is that putative disease-causing mutations should be absent in ethnically-matched controls. By convention, any sequence variant present in 1% or more of the population is termed a polymorphism, but rare variants present at lower frequency may not be identified in controls if insufficient numbers are screened. The ultimate distinction between pathogenic and neutral variants however can only be made by demonstrating changes in protein function consistent with models of disease pathogenesis. This might include demonstration of changes in the oxygen dissociation curve of haemoglobin due to globin gene mutations (Weatherall,
29 1985), or patch clamp recording from single channels, for example, in hyperkalaemic periodic paralysis where genes encoding sodium channel subunits carry mutations (Hudson et al, 1995). Relatively recent advances have permitted the creation of transgenic animal models of human disease by introduction of mutation-carrying human genes and subsequent functional rescue by reinsertion of the wild-type gene. Studies of this type, performed to study the Clock gene for example (Antoch et al, 1997), provide unambiguous evidence of the pathogenic role of mutations.
1.8 APPLICATION OF MOLECULAR GENETICS TO NEUROLOGICAL DISEASE
As many as 30,000 genes may be expressed in the human central nervous system. It is perhaps not surprising then, that of the approximately 5,000 mendelian disorders of humans listed by McKusick, a significant proportion have a partly or wholly neurological phenotype (McKusick, 1994; or via: http://www.gdb.org/). For these reasons, and because the pathophysiology of neurological disorders may sometimes be difficult to elucidate due to the complexity and inaccessibility of brain tissue, molecular genetic techniques have been applied extensively to the study of neurological disorders (Harding, 1993a; Baraitser, 1997). This approach has met with considerable success - a recent review of disorders predominantly affecting muscle, peripheral nerve or spinal cord for which the causative genes have been identified or mapped, lists 96 disorders (Kaplan and Fontaine, 1997). This takes no account of the rapidly growing list of disorders affecting the brain for which genes have already been mapped or cloned (on-line Mendelian Inheritance in Man, via: http://www.gdb.org/).
Identification of neurological disease genes is of importance for several reasons. Perhaps most importantly, characterisation of a disease gene often provides biological insights into the function of that gene in health and disease. Greater understanding of pathophysiology as a result of cloning studies is essential to pave the way for the development of rational forms of therapy. It is also hoped that study of relatively uncommon monogenic variants of diseases such as Alzheimer’s disease and amyotrophic lateral sclerosis will help to clarify the molecular pathogenesis of the far more common sporadic forms of these diseases. However, although many new disease-causing genes have been identified in recent years, understanding of the
30 processes linking genotype to final disease phenotype is beginning to emerge only relatively slowly.
Cloning of disease genes has provided clinicians with the means for both diagnostic testing and predictive testing of at-risk individuals and foetuses. These advances have simplified genetic counselling and increased its accuracy, thus directly benefiting patients (MacMillan and Harper, 1994; Bird and Bennett, 1995). The nosology and classification of neurological diseases has also been clarified by characterisation of diseases at the molecular level. This is exemplified by genotype-phenotype studies of the hereditary ataxias, as a result of which, it is now clear that several genes may result in clinically indistinguishable phenotypes and that a single locus may cause more than one phenotypic manifestation (Rosenberg, 1995). Future classifications of these disorders will thus be based on genetic locus as well as phenotype.
Many of the recent advances in the field of neurogenetics have been based on the elucidation of single gene disorders. Although these are important, many of the commoner neurological disorders are not caused by single genes but are thought to result from the interaction of environmental factors with multiple susceptibility genes. Diseases such as Parkinson’s disease, multiple sclerosis, stroke, epilepsy and Alzheimer’s disease, are all believed to have a genetic component. Studies based on use of non-parametric linkage methods and analysis of candidate genes in these conditions are underway (Bell and Lathrop, 1996; Giinel and Lifton, 1996; Wood, 1997).
1.8.1 Application of molecular genetics to the study of movement disorders
Movement disorders are neurological syndromes in which there is either an excess of movement (hyperkinesias) or a paucity of voluntary or automatic movements, unrelated to weakness or spasticity (hypokinesias). The involuntary movements which comprise the hyperkinesias may be classified into five main categories: tremor, chorea, myoclonus, tics and dystonia. Parkinson’s disease and a number of other akinetic-rigid syndromes with parkinsonian features constitute the hypokinesias (Marsden, 1996).
31 Involuntary movements may occur as part of the clinical phenotype of numerous genetically determined neurological disorders. In many such disorders the involuntary movements are accompanied by other neurological or non-neurological features, producing a distinct clinical syndrome or disease. However, several movement disorders may occur in isolation, without associated abnormalities on clinical examination or routine investigation. Some such disorders appear to have a genetic basis, and a molecular genetic approach has begun to identify some of the genes responsible for these disorders. Two genes causing Parkinson’s disease inherited as a mendelian trait, a-synuclein and parkin, have recently been reported (Polymeropoulos et al, 1997; Kitada et al, 1998; Kruger et al, 1998) and there are numerous genetic polymorphisms reported to be associated with an increased risk of developing sporadic idiopathic Parkinson’s disease (reviewed by Bandmann et al, 1998). Study of a large American kindred of Czech ancestry and Icelandic kindreds with essential tremor have resulted in identification of two genetic loci for tremor on chromosomes 2p22-p25 and 3ql3 (Higgins et al, 1997; Gulcher et al, 1997). The molecular genetic basis of the dystonias has likewise been the subject of intensive investigation and is discussed in chapter 5.
1.9 OBJECTIVES OF THIS THESIS
The author has attempted to apply molecular genetic techniques to the study of families with a number of inherited movement disorders. Families with a variety of dominantly inherited forms of dystonia, and families with an unusual movement disorder confined to the muscles of the face, named hereditary geniospasm were studied. An introduction to each disorder is given in the individual chapters. Kindreds were first followed up and clinically ascertained by the author, and DNA collected for analysis. Genetic linkage analysis, using microsatellite markers spread throughout the genome to localise disease genes, was performed in families with paroxysmal dystonic choreoathetosis (chapter 3), hereditary geniospasm (chapter 4) and primary torsion dystonia (chapter 5). Where possible, candidate gene analysis was also undertaken after identification of linkage and subsequent fine genetic mapping. Multipoint exclusion mapping at the dystonia loci on chromosomes 8 and 18 was performed in three families with primary torsion dystonia (chapter 5). Mutation analysis of the GCH1 gene was performed in individuals with
32 anticholinergic-responsive dystonia (chapter 6). These studies were undertaken with the aim of genetic mapping of the responsible disease genes as a first step in the process of gene characterisation as described above.
33 CHAPTER TWO
MATERIALS AND METHODS
2.1 OUTLINE OF CHAPTER
This chapter describes the materials used and experimental methodology employed in this study. The methodological stages of linkage analysis are described including extraction of DNA from blood, generation of DNA fragments using the polymerase chain reaction, separation of fragments by polyacrylamide gel electrophoresis, scoring of genotypes and computational linkage analysis. The DNA sequencing techniques and restriction enzyme analysis employed in mutation analysis of the GTP cyclohydrolase I gene are also described. Buffers and solutions used are described at the end of the chapter. The clinical features of the families and patients studied are given in the relevant results chapters. All chemical reagents were supplied by BDH unless otherwise specified.
2.2 DNA EXTRACTION
DNA was extracted from venous blood samples (collected in EDTA) using the Nucleon II system (Scotlab). To extract leucocytes, 10 mis of blood was added to 40 mis of reagent A (to lyse erythrocytes), and the mixture was shaken for four minutes before centrifugation (Beckman, model GS-6R centrifuge) at 2,600 rpm for five minutes (to pellet the leucocytes). The supernatant was discarded, 2 ml of reagent B added (to lyse leucocytes) and the pellet gently resuspended. 500 pi of 5 M sodium perchlorate was then added (to deproteinise the mixture) and the mixture shaken for 10 minutes at room temperature and for 15 minutes in a 65°C water bath. DNA was extracted by the addition of 2 ml of chloroform (at -20°C) and the contents shaken for one minute before centrifugation for two minutes at 2,000 rpm (Beckman, model TJ- 6 centrifuge). Nucleon suspension (300 pi) was added and the mixture centrifuged as above for five minutes at 2,000 rpm. The aqueous DNA-containing phase was taken and DNA precipitated by the addition of two equal volumes of absolute ethanol, the
34 mixture was then gently inverted. DNA was then hooked out into a tube containing 0.5 -1 ml of sterile 1 x TE.
2.3 MEASUREMENT OF DNA CONCENTRATION AND DILUTION OF DNA
DNA concentration was estimated by measurement of optical density (OD) using a spectrophotometer (Cecil, model CE202). OD measurements were performed at a wavelength of 260 nm using quartz cuvettes. Measurements were calibrated using distilled water. DNA dilutions of 5 pi of DNA in 1 ml of distilled water (1 in 200) were used for OD measurements. For linkage analysis using fluorescently tagged primers and microtitre plates, DNA was diluted to a concentration of 10 ng/pl using 1 x TE, and stored in covered deep-well titre plates (Beckman) at 4°C when in frequent use, and at -20°C when used infrequently. Concentrations of all other DNA samples were adjusted to approximately 50 - 100 ng/pl.
2.4 LINKAGE ANALYSIS
The technique of genetic linkage analysis was used extensively throughout this study. The principles of linkage analysis are described in detail in chapter 1. Briefly, linkage analysis refers to the mapping of a disease gene to a sub-chromosomal location. The technique makes use of polymorphic microsatellite DNA markers to identify a chromosomal region segregating with the disease in one or more families. Markers are amplified from genomic DNA of family members using the polymerase chain reaction (PCR) and separated using polyacrylamide gel electrophoresis. Family members may be genotyped for each marker according to the number of tandem repeats present (which are determined by the length of the PCR fragments). Segregation of a marker allele with the disease phenotype in a family suggests linkage between the microsatellite marker and disease gene.
During the course of this study, the techniques available for genotyping of microsatellite markers in genome-wide searches for linkage underwent considerable evolution and refinement (Ziegle et al, 1992; Mansfield et al, 1994; Reed et al, 1994). These advances necessitated several modifications of the experimental techniques used during the study in order to maximise genotyping efficiency and improve
35 accuracy. Thus, use of panels of fluorescently tagged primers superseded radioactive labelling of PCR products, microtitre plates replaced use of individual tubes for PCR reactions, and computerised semi-automated genotyping techniques superseded manual scoring of autoradiographs. These new techniques are capable of significantly increasing genotyping throughput and improving accuracy (Ziegle et al, 1992; Schwengel et al, 1994). As identification of linkage for a mendelian disorder using a genome-wide search usually requires the generation of thousands of genotypes, such improvements are of great benefit in linkage studies.
2.4.1 The polymerase chain reaction
The polymerase chain reaction is a method for the selective enzymatic synthesis of a specific DNA sequence using two oligonucleotide primers hybridising to opposite strands of DNA flanking the target sequence. Repeated cycles of DNA denaturation, primer annealing and extension of the annealed primers by the thermostable DNA polymerase derived from Thermus aquaticus (Taq), results in the selective amplification of the target DNA fragment. The extension products of each cycle are used as templates in subsequent cycles, resulting in exponential accumulation of target DNA. The requirements of PCR are as follows: DNA template, oligonucleotide primers, deoxynucleoside triphosphates (dNTPs), Taq DNA polymerase, a magnesium-containing buffer for the enzyme and a thermal cycler. Both primers and dNTPs must be present in excess. Two methods of visualising and sizing the final PCR product were used at different times in this study: firstly the labelling of one primer from each pair with a fluorecent dye detectable by laser using an automated DNA sequencer, and secondly radioactively labelling the DNA product by incorporation of [a- 32 PJdCTP and subsequent autoradiography. The former technique was used more extensively in this study than the latter.
2.4.1.1 Fluorescence-based PCR.
PCR reactions using fluorescently-labelled primers were carried out in final reaction volumes of 20 pi in 96-well microtitre plates (Micro Test III, Falcon). The reaction mixture consisted of: 0.2 mM each of dATP, dCTP, dGTP and dTTP (2 pi of 10 x dNTP solution, Promega); 2 pi of GeneAmp 10 x magnesium-free PCR Buffer II (Perkin Elmer); 1.5 mM MgCl2 (1.2 pi of 25 mM MgCl2 solution, Perkin-Elmer); 10
36 ng of each primer; autoclaved and filtered distilled water to make up reaction mixture volume to 15 pi; 0.5 units of DNA polymerase added last (AmpliTaq Gold™ 5units/pl, Perkin-Elmer). The reaction mixture was prepared at room temperature and aliquotted into microtitre plate wells using an eight channel pipette (Scotlab). Fifty ng (5 pi) of template DNA was then added to each well and overlaid with approximately 25 pi of paraffin oil (BDH) to prevent evaporation. Microtitre plates were then centrifuged at 1000 rpm for 30 seconds (Beckman GS-6R centrifuge).
PCR reactions were performed using a Hybaid Omnigene thermal cycler. Reaction mixes were first heated to 95°C for 11 minutes to activate the AmpliTaq Gold™; subsequent cycling conditions were:
94°C (denaturation) 30 seconds 48 - 56°C (annealing) 30 seconds 72°C (extension) 30 seconds repeated for 28-33 cycles
PCR conditions were optimised for each primer pair to determine the optimal annealing temperature and the number of cycles required to produce an approximately constant PCR yield. As PCR is a sensitive technique capable of amplifying very small quantities of template, great care was taken to avoid contamination during set-up and all reagents and materials used were sterile. A negative control (omitting template DNA) was always included in experiments.
2.4.1.2 Radioactive PCR.
Linkage analysis performed during the first year of this study made use of radioactive labelling of PCR products. PCR reactions were performed in 0.5 ml microcentrifuge tubes in a volume of 50 pi. Reactions were performed as above with the following differences: dCTP was reduced to 20 pM; 0.1 pi of [a-32P]dCTP (Amersham International) was added to the reaction mix; 5 pi of 10 x Taq DNA polymerase reaction buffer (Promega) was used in place of GeneAmp buffer (giving a final MgCl2 concentration of 1.5 mM); 1.25 units of Taq DNA polymerase (Promega) was added to each reaction after an initial five minute denaturing step
37 (‘hotstart’). After the addition• of [a- 32 P]dCTP to the reaction, appropriate precautions were taken for the handling of radioactive materials.
2.4.2 Oligonucleotide primers for microsatellite markers
For genome-wide linkage searches, a panel of fluorescently labelled primers designed for this purpose were used in all but one family studied (family 3, described in chapter 5). This marker panel was made available by the Medical Research Council of the UK. The panel contains 254 dinucleotide repeat marker loci covering all 22 autosomes and the X chromosome with an average marker spacing of 13 centiMorgans (Reed et al, 1994). 80% of the markers in the panel are derived from the Genethon genetic map (Dib et al, 1996) and the remaining markers are selected from a variety of other marker maps for genetic regions where Genethon markers are not available or are insufficiently polymorphic. A fluorescent dye moiety (6-FAM, TET or HEX) is attached to the 5’ end of one of each pair of PCR primers. The panel could be divided into 14 sets, each of which consisted of markers not overlapping in size or dye colour. Each set was capable of being run simultaneously on a single polyacrylamide gel (section 2.4.5). This marker panel was used for exclusion mapping (chapter 5) as well as for genome-wide linkage searches (chapters 3 and 4).
Additional microsatellite markers used for high resolution genetic mapping and exclusion mapping after February 1996 were analysed using custom-made fluorescently labelled primers. Primers were manufactured by Perkin-Elmer with a 5’ 6-FAM, TET or HEX dye on one of each primer pair. Markers used were dinucleotide (CA)n repeats from the Genethon genetic map (Dib et al, 1996), tetranucleotide repeats from the Utah marker development group (The Utah marker development group, 1995), and tri- and tetranucleotide repeats from the Cooperative Human Linkage Centre (CHLC) (Sheffield et al, 1995). In all cases, markers with the highest heterozygosities possible were selected for use in order to maximise informativeness.
38 For linkage analysis in one family (chapter 5, family 3), a set of conventional primers for microsatellite markers from the Genethon genetic map (Dib et al, 1996) were used. PCR products were radioactively labelled (section 2.4.1.2).
2.4.3 Agarose gel electrophoresis
To check for the presence of a PCR product of the desired size and quantity, 5 pi aliquots of reaction mix from four randomly selected wells (as well as the negative control) was visualised by electrophoresis on ethidium bromide stained agarose gels. 5 pi of reaction mix was added to 2 pi of agarose gel loading buffer (fix buffer consists of 40% w/v sucrose, 0.25% bromophenol blue) and electrophoresed at 50 V through a 3.2% agarose minigel (Flowgen Instruments Ltd) for 30 - 60 minutes. A 100 bp size standard (Gibco) (1 pi) was run alongside the PCR products to enable estimation of their size. Ethidium bromide staining of the agarose gel (1 mg / ml, Sigma) permitted direct visualisation of DNA products using transillumination with ultraviolet light.
2.4.4 Polyacrylamide gel electrophoresis
Electrophoresis through a polyacrylamide gel is an effective means of separating small DNA fragments with high resolution, allowing fragments differing in size by as little as 1 bp to be separated. Denaturing polyacrylamide gels are polymerised in the presence of an agent such as urea which suppresses base pairing in nucleic acids. Denatured (single stranded) DNA migrates through these gels at a rate that is determined by fragment size and almost completely independent of base sequence and composition, permitting sizing of fragments according to distance travelled through the gel (smaller fragments migrate further than larger ones).
2.4.5 Polyacrylamide gel electrophoresis of fluorescently labelled PCR products
PCR products produced using fluorescently tagged primers were sized by electrophoresis through a denaturing 4 % polyacrylamide gel in an automated DNA sequencer (Applied Biosystems, model 377). During electrophoresis, a section of the gel furthest from the loading comb is scanned by a laser causing each dye moiety
39 (attached to one oligonucleotide primer incorporated into PCR fragments) to emit light of a known wavelength as it migrates past the laser. A size standard consisting of DNA fragments of known size, labelled with the fluorescent dye TAMRA, is run in each lane to allow accurate sizing of PCR fragments. This method of DNA sizing has the great advantage over radioactive methods (see below) that markers of non overlapping size and dye composition may be multiplexed in each lane so maximising efficiency and increasing sample throughput. As many as 24 microsatellites may be run in each lane although in practice a maximum of 20 markers were run simultaneously during this study. Moreover, a comparison of fluorescence-based genotyping with autoradiographic techniques suggests that the former results in fewer genotyping errors than the latter (Schwengel et al, 1994). An example of a typical Genescan gel is shown in figure 2.1.
2.4.5.1 Polyacrylamide gel preparation
36 cm well-to-read glass plates were used with the 377 sequencer. Plates were cleaned with detergent and rinsed with distilled water. The dry plates were assembled in the 377 cassette prior to pouring the gel. The catalysts TEMED (Sigma) (35 pi) and freshly prepared 10% ammonium persulphate solution (APS) (Sigma) (250 pi) were added to 50 ml of 4% acrylamide gel mix (section 2.6.3) to start polymerisation. The mix was then taken up into a 50 ml syringe and carefully introduced into the notch between the front and back plates, spreading evenly between the glass plates. A spacer was inserted into the upper notch between the plates and the gel left for two hours to polymerise. After polymerisation, the upper spacer was removed and a 36 well shark’s tooth comb was carefully inserted in its place. The cassette and plates were then placed in the 377 sequencer and the plates checked for background fluorescence using Genescan (version 2.0.2) software (Applied Biosystems). Heating plate and buffer chambers were assembled and 1.3 L of 1 x TBE buffer (section 2.6.1) added before pre-running the sequencer until the gel temperature reached 50°C. Samples were then loaded.
2.4.5.2 Pooling o f PCR products for loading
Up to 20 non-overlapping microsatellite markers, amplified from a single DNA sample, were run simultaneously in each lane (multiplexed). PCR products from each
40 Figure 2.1: Illustration of a typical Genescan run showing the gel image obtained. Samples have been run in 36 lanes (indicated at the top of the figure) with 7 microsatellite markers amplified in each lane. These are visible as green (TET), blue (6- FAM) and yellow (HEX) bands on the gel. The internal size standard run in each lane is visible as red bands. Lane 1 contains the marker ‘wand’ visible as a white stripe.
1 2 3 4 5 6 7 8 9 10 1112 1314 15 16 17 1819 20 21 22 23 24 25 26 27 28 293031 32 33 34 3536 A AAA M AAAA AA M flA AMAAAAA AAAA AMA A A M
41 DNA sample were first pooled according to the dye they contained as follows: 6- FAM - 4 pi; TET - 4 pi; HEX - 10 pi. These volumes were adjusted according to the yield of the PCR as determined on agarose gel electrophoresis (section 2.4.3). Pooling was performed in microtitre plates using an eight channel pipette. 2.5 pi of pooled product from each well was then aliquotted into a fresh microtitre plate and an equal volume of loading mix added. The loading mix consisted of 100 pi of deionised formamide, 20 pi of loading buffer (blue dextran, 50mg/ml, EDTA 25 mM, Perkin Elmer) and 24 pi of Genescan 350-TAMRA or 500-TAMRA size standard (depending on the anticipated size of the largest PCR fragment) (Applied Biosystems). The final mix of pooled product and loading mix was denatured at 95°C for 2 minutes in a Hybaid thermal cycler and then placed immediately onto ice before loading.
2.4.5.3 Gel loading and electrophoresis conditions
Wells were carefully flushed with 1 x TBE buffer immediately prior to loading. Alternate (odd-numbered) wells were loaded with 1.8 pi of final mix using a P2 pipette (Gilson) and Sorenson MiniFlex 0.2 mm flat tips (Anachem). Great care was taken to avoid spill-over into adjacent wells. Electrophoresis at 3,000 V for two minutes ensured that samples were run into the gel before even-numbered lanes were loaded. Loading of alternate lanes made it possible to distinguish adjacent lanes in the final gel image and improved the ability of the software to track lanes correctly. Total run time was two hours. A maximum of 36 samples could be run in adjacent lanes. Where more than 36 DNA samples from a single family were genotyped for a given marker (necessitating two electrophoresis runs), five samples from the first run were included in the second run as a control, to ensure gel-to-gel consistency of scoring.
2.4.6 Data analysis for fluorescently labelled PCR products
2.4.6.1 Initial data processing
Data collected during the electrophoresis run were analysed automatically using the Genescan programme in order to size DNA fragments separated by electrophoresis. Automatic lane tracking was checked using the gel image, and adjusted lane by lane
42 where necessary. To ensure accurate sizing, the automatic designation of peak sizes for internal lane size standards was manually checked in each lane. Markers were only sized if there were two size standard bands of greater size, and two of smaller size, present in the lane.
2.4.6.2 Genotyping using the Genotyper programme
Fragment size data collected using the Genescan programme were then analysed using the Genotyper (version 1.1) programme (Applied Biosystems) as described in the programme manual. The Genotyper programme labels fluorescent peaks with fragment size (to 0.01 of a base) and filters out background peaks. Manual adjustments are required by scrolling through all electropherograms to ensure that alleles are correctly labelled. Although time-consuming, this step is extremely important as labelling of incorrect (non-allele) peaks is a major source of genotyping error if not manually checked. Peaks in each marker range are grouped into discrete alleles and sequentially numbered from smallest to largest. Genotypes were scored blind without reference to the family pedigree to minimise bias. An example of allele sizing of a microsatellite marker using the Genotyper programme is illustrated in figure 2.2.
2.4.7 Polyacrylamide gel electrophoresis of radioactively labelled PCR products
Denaturing 6% polyacrylamide gels were used to separate radioactively labelled PCR products. In contrast to the fluoresence-based methodology described above, only one or two markers could be run simultaneously in each lane. 21 x 50 cm glass plates were used with apparatus supplied by Biorad. Plates were thoroughly cleaned with distilled water and absolute ethanol before assembly. The upper plate was periodically treated with a small volume of SigmaCote (Sigma) to prevent gels from adhering to this plate when disassembled. The bottom of the plates was sealed by allowing the apparatus to stand in a casting tray and polymerising 40 mis of acrylamide mix in the tray for approximately 30 minutes. To 90 mis of acrylamide gel mix was added 90 pi of TEMED and 90 pi of 25% APS. The mix was then carefully poured between the plates and a comb inserted into the upper edge of the gel.
43 Figure 2.2: Illustration of allele sizing using the Genotyper program. The microsatellite marker D8S260 has been amplified in a small family (parents in lanes 27 and 28; children in lanes 29 and 30). Fluorescent peaks have been labelled with allele sizes in base pair units after manual editing. ‘Stutter peaks’ of size n-2 and n-4, due to imperfect replication, are visible and are indicated by asterisks in lane 29.
•2 7 Ti I1 \ ' I I V
199.03 204.82
•2 8 A / V A
205.10 207.16 •2 9
* * A / V
204.82 PCR products were prepared for loading by adding 5 pi of PCR reaction mix to 2 pi of loading buffer (80% deionised formamide, lmg/ml bromophenol blue, lmg/ml xylene cyanol, 10 mM EDTA). Samples were denatured at 95°C for two minutes and placed onto ice immediately prior to loading. 1 x TBE preheated to 50°C was used to fill the buffer chambers and the gel comb removed. Wells in the acrylamide were flushed prior to loading samples (7 pi of each sample). Samples were run for the desired length of time determined by the relative positions of the bromophenol blue and xylene cyanol dyes.
Once electrophoresis was complete, the apparatus was disassembled with the polyacrylamide gel remaining on the inner glass plate. The gel was transferred to 3 mm filter paper (Whatman) and covered with Saran wrap prior to drying at 80°C for one hour usinga vacuum gel dryer (Biorad). Autoradiography with Fuji RX film was performed with intensifying screen (Genetic Research Instrumentation) at -70°C for 12-72 hours. Autoradiographs were manually scored.
2.4.8 Computational linkage analyses
Pairwise linkage analyses were performed using the FASTLINK 3.OP version of MLINK, or version 5.1 of MLINK from the LINKAGE program package (Lathrop and Lalouel, 1988; Cottingham et al, 1993; Schaffer et al, 1994). An autosomal dominant model of inheritance was used for all families with a disease allele frequency set at 0.0001. Disease penetrance varied according to the disorder studied and these are given in the individual results chapters. Marker allele frequencies were assumed to be equal.
Multipoint linkage analyses of families with primary torsion dystonia (chapter 5) were performed using the VITESSE program (O'Connell and Weeks, 1995) with the assistance of Dr T.G. Nygaard, Columbia University, New York, USA. A conservative ‘affecteds-only’ methodology was used for the exclusion study described in chapter 5, in order to avoid bias resulting from inclusion of possibly affected individuals or incorrect estimation of penetrance or age of onset.
45 2.5 MUTATION ANALYSIS
The GTP cyclohydrolase I gene (GCH1) was sequenced to identify mutations in patients with anticholinergic responsive dystonia (chapter 6). Where mutations were identified, restriction enzyme digests were performed to confirm genotype in patients, and to screen controls for the sequence variant. All six exons of GCH1 were amplified separately by PCR. Both sense and antisense strands of all six GCH1 exons were sequenced for all patients, and a normal control sample was always included in sequencing runs. Exon sizes were as follows: exon 1- 343 bp, exon 2-110 bp, exon 3-56 bp, exon 4-32 bp, exon 5-85 bp, exon 6 -124 bp.
2.5.1 Amplification of exons by PCR
Exons were individually amplified by PCR in order that the DNA products of these reactions could be sequenced. PCR reactions were performed as described in section 2.4.1.2 with the following differences: reactions were ‘cold’ (non-radioactive), a volume of 25 pi was used, and annealing and extension steps were performed at 60°C for 30 cycles. Oligonucleotide primer sequences were as reported previously (Ichinose et al, 1995a; Bandmann et al, 1996). DNA from PCR reactions was then treated to remove contaminating protein and amplification primers using the Wizard Prep™ DNA purification system (Promega).
2.5.2 DNA sequencing
The Sanger method of sequencing makes use of enzymatic DNA synthesis from template, in the presence of base-specific dideoxynucleotide chain terminators (Sanger et al, 1977). Dideoxynucleoside triphosphates (ddNTPs) are analogues of dNTPs which may be incorporated into an extending DNA molecule but which cause termination of chain synthesis as a result of the absence of a 3 ’ hydroxyl group. DNA synthesis starts from an oligonucleotide sequencing primer in the presence of DNA polymerase, dNTPs and a lower concentration of fluorescently labelled ddNTPs. A population of DNA molecules of different sizes are generated, which may be distinguished by length and terminal ddNTP (each of the four ddNTPs is labelled with a different fluorophore group). These may be separated by polyacrylamide gel electrophoresis using an automated DNA sequencer.
46 2.5.2.1 Cycle sequencing using Taq DNA polymerase
This method was used to sequence the GCH1 gene. The DNA resulting from amplification of GCH1 exons by PCR (section 2.5.1) was used as template for sequencing reactions. The Taq DyeDeoxy Terminator sequencing kit (Applied Biosystems), which utilises AmpliTaq DNA polymerase, was used for sequencing. Taq DNA polymerase is fast and processive, and because of the high temperature at which it is used, non-specific primer binding and secondary structure problems in template DNA are reduced. The kit utilises dITP to minimise band compressions on the sequencing gel. All four termination reactions are carried out in a single tube.
Each sequencing reaction contained 7 pi of purified DNA product, 9.5 pi of terminator premix (1.58 pM A-DyeDeoxy, 94.74 pM T-DyeDeoxy, 0.42 pM G- DyeDeoxy, 47.37 pM C-DyeDeoxy, 78.95 pM dITP, 15.79 pM of dATP, 15.79 pM of dTTP, 15.79 pM of dCTP, 168.42 mM Tris-HCl (pH 9.0), 4.21 mM (NH4)2S04, 42.1 mM MgCl2, 0.42 units/pl AmpliTaq DNA polymerase), and 3.2 pmol of primer. Sequencing primers were identical to those used for initial amplification. A final volume of 20 pi was reached by addition of sterile water and the mix overlaid with paraffin oil. Using a thermal cycler, the mixture was subjected to 25 cycles of 96°C for 30 seconds, 50°C for 15 seconds and 60°C for 4 minutes. Once completed, the tubes were placed on ice and a phenol chloroform DNA extraction performed. The sequencing reaction mix was made up to 100 pi with water and the paraffin oil removed. An equal volume of phenol/ chloroform (Sigma)/ water (in the ratio 68:14:18 by volume) was added and mixed by vortex for 30 seconds, then spun on a microfuge for 2 minutes. The aqueous phase was removed and the procedure repeated. 15 pi of 3 M sodium acetate and 300 pi of absolute ethanol was then added to the separated aqueous phase. The samples were placed on ice for 15 minutes and then microfuged for 10 minutes. The supernatant was discarded and the remaining pellet washed in 70% ethanol and then dried. This was then resuspended in 5pl of loading buffer (5:1 (v/v) mixture of deionized formamide and 50 mM EDTA) and denatured at 94°C before loading onto a 6% acrylamide denaturing gel. Prior to loading wells were flushed with TBE solution and alternate lanes loaded. Gels were run on a 373 automated DNA sequencer (Applied Biosystems) for 12 hours in 1 x TBE. The Sequence Editor program (Applied Biosystems) was used to compare test
47 sequence with control and published sequence (GenBank accession numbers U15256-9).
2.5.3 Detection of mutations by restriction enzyme analysis
Creation or abolition of restriction enzyme cutting sites by base changes identified during sequencing were used to confirm the presence of the mutation in patients and to screen controls for the mutation. GCH1 exons containing putative mutations were amplified using PCR as described in section 2.5.1. For digestions, 8 pi of PCR product was incubated with 1 pi of enzyme and 1.2 pi of the appropriate enzyme buffer for 1 hour (Alu-I at 37°C, Taq-I at 65°C, Dde-I at 37°C; all supplied by Promega) in microcentrifuge tubes. The digested products were then separated on agarose gels as described in section 2.4.3 and the pattern of bands obtained used to identify the presence or absence of cutting sites. A control containing uncut PCR product (1 pi of water substituted for the restriction enzyme) was always run with digested products and a digested sample of control (wild type) DNA was always included.
2.6 BUFFERS AND SOLUTIONS
Unless stated otherwise, buffers and solutions were prepared using distilled water. Chemicals were supplied by BDH unless otherwise stated.
2.6.1 10 x TBE (I L)
Tris (Sigma) 121.1 g Boric acid 61.8 g (anhydrous) EDTA (Sigma) 7.4 g
2.6.2 10 x TE (lOmM Tris/lmM EDTA) (I L)
Tris 1.21 g EDTA 0.37 g
2.6.3 4% acrylamide gel mix for 377 sequencer (50 ml)
Urea (Fison’s) 18 g Water 27.5 ml
48 40% acrylamide solution (Biorad, 19:1 acrylamide:bisacrylamide) 5 ml Amberlite resin (Sigma, deionising) 0.5 g 10 x TBE 5 ml Filter TBE through Whatman filter system (0.45 jam filter). Stir remaining constituents on a magnetic stirrer until dissolved. Filter onto TBE.
2.6.4 6% acrylamide gel mix for radioactive gels (500 ml stock solution)
Urea 230 g 40% acrylamide solution (19:1 acrylamide:bisacrylamide) 75 ml 10 x TBE 50 ml
Volume adjusted to 500 ml with water. Filtered and stirred as above. Stored for up to four weeks at 4°C.
49 CHAPTER THREE
A CLINICAL AND GENETIC STUDY OF A FAMILY WTTH PAROXYSM AT
DYSTONIC CHOREOATHETOSIS
3.1 OUTLINE OF CHAPTER
This chapter describes a clinical and genetic linkage study undertaken in a large British family with paroxysmal dystonic choreoathetosis (PDC) - the ‘B’ family. The ascertainment of the family from two apparently unrelated index cases and the clinical features of affected family members are described. A genome-wide search for genetic linkage using microsatellite markers was performed. Fine genetic mapping using closely spaced microsatellite markers was performed where positive LOD scores suggested linkage. Haplotypes were constructed for an area segregating with the PDC phenotype in order to identify flanking recombinations. Polymorphic tandem repeats within putative candidate genes mapping to this area were analysed. The significance of the results presented in this chapter are discussed and future directions of study suggested.
3.2 INTRODUCTION
3.2.1 The paroxysmal dyskinesias
The paroxysmal dyskinesias are an unusual group of hyperkinetic movement disorders characterised by paroxysmal attacks of a mixed movement disorder, encompassing elements of dystonia, chorea and ballism in varying proportions (Fahn, 1994). Patients are entirely normal between attacks and clear consciousness is preserved throughout attacks. The current classification of these disorders divides them into three groups according to the duration and precipitation of attacks: paroxysmal dystonic choreoathetosis, paroxysmal kinesigenic choreoathetosis (PKC) and paroxysmal exercise-induced dystonia (PED)(Lance, 1977; Demirkiran and Jankovic, 1995).
50 3.2.2 Paroxysmal dystonic choreoathetosis
PDC is a rare and distinctive disorder characterised by attacks of involuntary movements consisting of dystonia, chorea and ballism intermixed (Fahn, 1994). The first clear description of a patient with PDC was by Mount and Reback in 1940, who named the condition ‘familial paroxysmal choreoathetosis’ (Mount and Reback, 1940). They reported a young man with attacks of choreoathetosis, ‘wild flinging movements of the upper extremity’ (ballism), and ‘torsion spasm’ (dystonia), with onset in infancy. Twenty-seven other family members over five generations were reported to be similarly affected with an autosomal dominant pattern of inheritance. Walker (1981) provided a follow-up of the descedants of Mount and Reback’s index case, who suffered from an identical movement disorder. Forssman may be credited with the second description of a family with typical PDC. In 1961, he provided a detailed and meticulous clinical description of involuntary movements in several generations of a large Swedish family, in a paper entitled: ‘A hereditary disorder characterised by attacks of muscular contractions, induced by alcohol amongst other factors’. Forssman, who was unaware of the earlier report of Mount and Reback, rejected the previous diagnoses of epilepsy, hysteria, and paramyotonia given to family members, and suggested instead that this was a hitherto unknown disease which he called ‘hereditary intermittent tetanus’. Other large families with PDC were later described, including those of Richards and Barnett (1968) who introduced the term ‘paroxysmal dystonic choreoathetosis’ to emphasise the dystonic component of the movements, and Lance (1963, 1977) who helped distinguish PDC from PKC, and added the exercise-induced form to the classification.
Most reported cases of PDC are familial with autosomal dominant inheritance and high penetrance. However, sporadic cases of a paroxysmal movement disorder similar to PDC have also been reported (Bressman et al, 1988; Fahn, 1994; Demirkiran and Jankovic, 1995). Many cases of sporadic PDC are secondary to another neurological lesion or disease, and have been attributed to disorders as diverse as multiple sclerosis, stroke, trauma, encephalitis, perinatal hypoxia, hypoparathyroidism and basal ganglia calcification (Fahn, 1994; Demirkiran and Jankovic, 1995). From a case series of 25 patients, Bressman (1988) concluded that a psychogenic aetiology was the most common cause of sporadic PDC. Demirkiran and Jankovic (1995), in their own series, made a similar finding.
51 3.2.2.1 Clinical features of paroxysmal dystonic choreoathetosis
There have been 11 families with clinically well-characterised PDC have been reported in the English-language literature since 1940 (Mount and Reback, 1940; Forssman, 1961; Weber, 1967; Richards and Barnett, 1968; Lance, 1977; Tibbies and Barnes, 1980; Kurlan et al, 1987; Byrne et al, 1991; Fouad et al, 1996; Schloesser et al, 1996; Fink et al, 1997). These comprise 124 affected individuals in total, of whom clinical details are available for 77. The clinical features of these families are summarised in table 3.1. From these cases it is possible to propose a core phenotype of clinical features which characterise familial PDC (table 3.2). These may be of use for diagnosis of additional families, and to standardise phenotype in genetic studies. Since the completion of this study a further two families with PDC have been reported, both of which conform to the classical PDC phenotype (Hofele et al, 1997; Raskind et al, 1998).
Attacks of involuntary movements in patients with PDC are often preceded by a sensation of tightness or paraesthesiae in a limb, followed after a few minutes by the onset of movements usually in a hand or foot. Typically, symptoms progress to a hemi- body distribution affecting ipsilateral arm, leg and face, and may subsequently generalise to affect all limbs, trunk and face. Speech is involved with slurring dysarthria or anarthria. The duration of attacks may be as short as 10 minutes or as long as 24 hours in some cases, but is more typically 30 minutes to 3 hours. Attacks may be precipitated by a variety of stimuli including emotion, alcohol and caffeine. Many patients report that attacks may be aborted by sleep, but drug therapy is frequently unrewarding, although benzodiazepines (in particular clonazepam) have been used with success in some patients (table 3.1). The onset of symptoms is usually during early childhood years, often in the first year of life. The frequency of attacks is very variable, but may range from three attacks daily to one attack per year, with a tendency in many reports for attacks to become less frequent with advancing years (table 3.1). Males outnumber females by approximately 1.3:1 (table 3.1). As with all the paroxysmal dyskinesias, patients remain fully aware throughout attacks and return to normal between them. Not surprisingly, as awareness of the condition is not widespread, it is often erroneously diagnosed as epilepsy (including pseudo-seizures), a psychiatric disorder or hysteria. There are no estimates of the prevalence of PDC but the scarcity of reported cases suggests that this is a rare condition.
52 Table 3.1: Summary of clinical features of 12 families with PDC including the ‘B’ family. a " pa ~ £ . PQ 0£ ■a 4 M O 2 « > ON VO eo \ o 4 - 1 VO ovvo *» £ £ " •a < 4J 4J — « g s N — (N S3 co o s = -s r co a js •T3 ■*r ;0 2 < A A V >-> A S I a i ■2 o o ■ W + + + + + + + + + + + • + + • + + + • + « + + + + + + + + + + + + * + + + + + + + + + +++++++ —« o x o _ s e — .5 « E OO }»- B « _ « 2 + + + • + + + • < UH f U W ffi U H U + + + > < + + • io m Table 3.1 (continued): Summary of clinical features of 12 families with PDC including the ‘B’ family CQH •■2 V - — CQ " £ h-) n pi cl .2 ■o CQ £ 'i q >> i - i O'* 5 I £ CO 5 ^ 2 u c l-H — o V Os v) 2> w5 L- n e ^ vo vo OS jj 3 « * c/3 Pi U Q 3 . r * 's 1 W B « - l .2 8 « 9 U -2 e « " V W C V 3 + r + i « « + i + i + I » « p 8- c S £ ■a "O CJ X o E 6 (U o 5 a jx £ js 3 T3 .s T3 M £ < -* + o s 3 O CQ > * £ 1 i E CD E ex x> tx 1 CQ a> s-l a> s 0 .s .s ;3 *> rs *3 .S eta .X .s < ta 3 CQ 4> cr CD c O CQ 00 ex "3 TJ X X *o • Episodic. No abnormality between attacks. • Mixed movement disorder. Chorea, dystonia, ballism, in any combination. • Speech frequently involved. Dysarthria or anarthria. • No disturbance of consciousness. • Early age at onset. Usually early childhood. • Attack duration >5 mins. Often > 1 hour. • Precipitating factors may include: emotion, stress, alcohol, caffeine, tiredness, hunger, cold, exercise. Never precipitated by sudden movement. • Attack frequency never > x3/ day. Usually decreases with age. • Marked sleep benefit and diurnal variation often present. • Family history: autosomal dominant with high penetrance. 55 The phenomenology of attacks in patients with sporadic PDC is similar to that of patients with familial PDC, but considered together, patients with sporadic PDC differ from familial cases in several important respects: the age at onset in this group is usually older than in familial cases (early adult life rather than childhood), the excess of males over females in familial cases is reversed in the sporadic group, and attacks may be more frequent or more prolonged in some sporadic cases as compared to their familial counterparts. Furthermore, characteristic clinical features of familial PDC, such as sleep responsiveness and precipitation by caffeine or alcohol, are rarely present in sporadic cases (Bressman et al, 1988; Fahn, 1994; Demirkiran and Jankovic, 1995). These observations suggest that sporadic PDC may represent a non-genetic phenocopy of the familial form, rather than being the result of new mutations in a PDC gene. This view is supported by two other lines of evidence: firstly that a PDC-like phenotype may occur as a result of a variety of non-genetic diseases (symptomatic PDC), and secondly that the high penetrance of the PDC gene means that a family history is almost always apparent (although this observation might conceivably represent ascertainment bias). However, it remains possible that genetic heterogeneity underlies the clinical differences between familial and sporadic groups. This issue will only be finally resolved with the identification of the PDC gene or genes. 3.2.3 Other paroxysmal dyskinesias 3.2.3.1 Paroxysmal kinesigenic choreoathetosis PKC is the commonest of the three forms of paroxysmal dyskinesia. The condition is characterised by brief attacks of chorea, dystonia and ballism, similar in phenomenology to those of PDC but typically lasting only a few seconds and always less than five minutes. In contrast to PDC, attacks are precipitated by sudden movement (thus the term kinesigenic) or startle and usually respond well to treatment with anticonvulsant drugs. Patients experience numerous (sometimes 100 or more) attacks each day (Kertesz, 1967; Fahn, 1994). Inheritance is considered to be autosomal dominant with reduced penetrance in about two thirds of cases (Harding, 1993b); the remaining cases are apparently sporadic. 3.2.3.2 Paroxysmal exercise-induced dystonia Lance (1977) described members of a family over three generations who developed attacks of dystonia lasting 5-30 minutes following prolonged exercise. Subsequently, a second 56 family and fifteen sporadic cases with PED have been reported (Plant et al, 1984; Nardocci et al, 1989; Wali, 1992; Demirkiran and Jankovic, 1995; Bhatia et al, 1997). PED differs from both PKC and PDC in the following respects: the phenomenology of attacks is different (dystonia is the only abnormal movement in most cases of PED, ballism and chorea are usually absent), precipitation is always by prolonged exercise rather than initiation of movement, alcohol, caffeine, and finally, the duration of attacks is intermediate between that of PKC and PDC. These considerations merit classification of PED as a separate disorder rather than a forme fruste of PDC or PKC. Only the eventual identification of the genes responsible for the paroxysmal dyskinesias will allow their classification to be clarified on the basis of genotype rather than clinical features. 3.2.4 Pathophysiology of paroxysmal dyskinesias With regard to pathophysiology, little has changed in our understanding of PDC since 1968 when Richards and Barnett, discussing its aetiology, commented ‘the ultimate answer probably lies within the chemistry of the basal ganglia’ (Richards and Barnett, 1968). The unilateral nature of many attacks with involvement of the ipsilateral side of the face suggests a central rather than peripheral site of dyskinesia generation. Several lines of evidence support the view that the basal ganglia are the structures involved in the generation of the abnormal movements seen in PDC. Structural, ischaemic or ablative lesions of the basal ganglia and their projections are sufficient to cause dystonia, chorea or ballism (Bhatia and Marsden, 1994), and positron emission tomography (PET) studies have demonstrated increased blood flow to the contralateral basal ganglia in a patient with paroxysmal hemi-dystonia (Perlmutter and Raichle, 1984). The neuronal circuitry of the basal ganglia is complex but current concepts of the projections from the striatum to the globus pallidus and beyond envisage two major pathways: the ‘direct’ pathway supports cortically generated movement while the ‘indirect’ pathway eventually inhibits it (Albin et al, 1989) (figure 3.1). Progressive reduction of cortical inhibition by reduced activity in the indirect pathway is believed to result in chorea and finally ballism. The early selective loss of the GABA/enkephalin-containing striato- pallidal neurons of the indirect pathway is thought to be the cause of chorea and ballism in Huntington’s disease (Reiner et al, 1988). The same neuronal pathway would be well 57 Figure 3.1: Basal ganglia motor circuits. Schematic representation of basal ganglia circuitry showing the direct and indirect striatal efferent pathways. Solid lines indicate excitatory pathways and broken lines indicate inhibitory pathways. Details of striatal synapses and receptors of the two pathways are depicted below. BS/SC: brainstem/spinal cord; GPe: external segment of the globus pallidus; GPi/SNr: internal segment of the globus pallidus/substantia nigra pars reticulata; SNc: substantia nigra pars compacta; STN: subthalamic nucleus; THAL: thalamus. From Ferre et al, 1992. CEREBRAL CORTEX STRIATUM THAL G Pi/SN r placed to play a central role in the dyskinesias of PDC. Striato-pallidal neurons co-localise dopamine (D2) and adenosine (A2) receptors. The effect of dopamine in the striatum is to inhibit firing of the neurons making up the indirect pathway while activating the direct pathway. Adenosine, in contrast, acts through A2 receptors to inhibit transduction at dopamine receptors and so enhances indirect pathway activity (Ferre et al, 1992; Ferre et al, 1993; Ferrari et al, 1997). The recent recognition that adenosine plays a central role in the mediation of sleep (Porkka-Heiskanen et al, 1997) is also intriguing given the effect of sleep deprivation in precipitating attacks, and the remarkable restorative effects of sleep in this disorder. Any model of PDC pathophysiology must explain the effects of caffeine, alcohol and emotion in precipitating attacks. Caffeine is a competitive antagonist at adenosine receptors, and most pharmacological effects of adenosine can be suppressed by relatively low concentrations of caffeine, such as those attained after drinking 1-3 cups of coffee (Nehlig et al, 1992). Antagonism of endogenous adenosine at the level of the striato- pallidal neuron, with consequent inhibition of the indirect pathway, forms a theoretical basis for the effect of caffeine in precipitating attacks in PDC. In addition, caffeine has been shown to have variable effects on dopamine release in different brain regions, although it may decrease dopamine release in the striatum (Nehlig et al, 1992). Alcohol is capable of inducing dopamine release in the striatum in experimental animals (Yoshimoto et al, 1991), and could therefore inhibit the indirect pathway through this mechanism. Finally, the limbic loop from the cingulate gyrus and amygdala to the nucleus accumbens of the ventral striatum is a possible anatomical substrate by which the potent effect of emotion on paroxysmal movement disorders such as PDC might be mediated. The primary defect in PDC is not known, however. It has been suggested that either an abnormal surge of dopamine release (caused by precipitants such as alcohol), or postsynaptic dopamine receptor upregulation, may make patients with PDC susceptible to attacks (Fink et al, 1997). The proposed mechanism of dyskinesia generation according to such a model is an increase in dopaminergic transmission causing choreoathetosis, which may be followed by a period of relative dopamine deficiency causing prolonged dystonia. The evidence supporting this hypothesis is mostly indirect, however. Postmortem studies of the brains of two patients with PDC have shown no neuropathological abnormality (Lance, 1977). 59 Other genetic neurological disorders with intermittent or paroxysmal manifestations, such as the periodic paralyses and episodic ataxias, have been shown to result from mutations in membrane ion channel genes. Hyperkalaemic and hypokalaemic periodic paralysis are associated with sodium channel (SCN4A) and calcium channel (CACNL1A3) mutations respectively (Ptacek et al, 1991 and 1994); episodic ataxia types 1 and 2 are due to mutations in potassium (KCNA1) and calcium channel (CACNL1A4) genes (Browne et al, 1994; Ophoff et al, 1996). By analogy with the paroxysmal nature of these disorders, it has been suggested that the paroxysmal dyskinesias may share a similar disease mechanism, and that disorders of one or more ion channels might underlie these disorders (Fouad et al, 1996). There is no direct evidence to support this view at present, but the observation that a potassium channel gene cluster maps to the vicinity of the locus for an atypical form of paroxysmal choreoathetosis (CSE) is intriguing (Auburger et al, 1996). The relationship of paroxysmal dystonias, in particular PKC, to epilepsy has been a matter of considerable debate. The paroxysmal nature of attacks of involuntary movements bears clear similarities to epilepsy, although the dystonic and choreoathetoid nature of the movements is unlike the tonic or clonic movements usually associated with generalised or focal motor seizures. Electroencephalogram (EEG) recordings between attacks in patients with PKC or PDC have shown no consistent abnormality but there have been a number of reports of paroxysmal epileptiform EEG discharges during attacks of PKC (Perez-Borja et al, 1967; Kato and Araki, 1969; Hirata et al, 1991). Hirata et al argued that the origin of the abnormalities they observed (a generalised 5-Hz discharge with prolonged spike duration) suggested a subcortical origin of discharge. The excellent therapeutic response of PKC to anticonvulsants, and the coexistence of PKC and epilepsy in some patients (Hudgins and Corbin, 1966) support the contention that PKC may be a form of epilepsy. This viewpoint is reinforced by a recent description of four French families in whom both paroxysmal dystonia (probably of the kinesigenic type, although this is not clear from the published clinical details) and benign infantile convulsions appear to be different phenotypic manifestations of the same genetic disorder, linked to a locus at the pericentromeric region of chromosome 16 (Szepetowski et al, 1997). However, it should be noted that EEG abnormalities are not universal during attacks of PKC (Kato et al, 1970) and EEG abnormalities have not been described during attacks in patients with PDC. Moreover, the sparing of consciousness and the lack of a postictal state are constant 60 features of all forms of paroxysmal dyskinesia. Thus, although there is evidence supporting a relationship between epilepsy and PKC, there is little evidence at present that PDC has an epileptogenic basis. Ultimately such semantic distinctions may prove to be sterile, as a definition of epilepsy as a paroxysmal discharge of cerebral neurons may well be expected to encompass paroxysmal disorders such as those discussed in this chapter. Perhaps more relevant are questions concerning the anatomical and physiological substrates and the molecular events which underpin the disorder. 3.2.5 Rationale for the study PDC is a disabling condition which is refractory to most forms of treatment. As the underlying pathophysiological basis of the disorder is little understood, a genetic linkage analysis approach followed by positional cloning is likely to be the most effective means of disease gene identification. Identification of the PDC gene will improve understanding of the molecular pathology of PDC, and may be expected to lead to insights into the function of the basal ganglia in health as well as disease, and to the development of more rational forms of treatment for PDC and other movement disorders. 3.3 PAROXYSMAL DYSTONIC CHOREOATHETOSIS IN THE ‘B’ FAMILY Two index cases with PDC were identified during the course of this study, one from the records of the National Hospital for Neurology and Neurosurgery, London, the other referred by Dr S Hodgson at Guy’s Hospital, London. The two index cases were from different regions of the UK, and were initially thought to be unrelated. Each had an extensive family history of similarly affected relatives with clear autosomal dominant inheritance. A common ancestral surname suggested that the two families may have been related. Follow-up of relatives with the help of the oldest surviving member of the family (III-13), aged 85, made it possible to trace the ancestry of both index cases to a common ancestor who lived in south-east London during the 19th century. This individual was remembered as suffering from involuntary movements typical of PDC. It was also possible to trace a third branch of the family with a single affected member (IV-19), but a fourth branch of the family with at least one affected member (IV-1) could not be traced. The pedigree is illustrated in figure 3.2. 61 Figure 3.2: The 'B' family pedigree, excluding individuals younger than eight years of age. The index cases are indicated by arrows. Affected status is denoted by filled symbols. The 36 family members used during the initial genome screen are indicated by asterisks. Additional family members genotyped for markers in areas of interest are indicated by double asterisks. Birth order, and gender of some individuals, have been altered to make the pedigree anonymous. ■T -■ $ $ * o = o s ) t . o ' s VO (N In all, 27 family members were affected by PDC over 6 generations, 20 of whom were still living at the outset of this study. The author interviewed 48 family members and spouses and obtained a detailed history of symptoms. Family members younger than eight years of age were not included in the study and do not appear in figure 3.2. Individuals were designated as affected or unaffected on the basis of the presence or absence of a history of typical attacks. Physical and neurological examinations were performed in all but five family members (three affected and two unaffected). Attacks were witnessed by the author in five individuals and recorded on videotape in four. The medical records of affected family members were requested. Venous blood samples for DNA extraction were obtained from 46 family members and spouses. The male to female ratio of affected individuals is 1.7:1. There are eight instances of male- to-male transmission which excludes the possibility of X-linked inheritance. There are no obligate, non-manifesting, gene carriers in the family. 3.3.1 Clinical features of index cases 3.3.1.1 Case V-9 A 35-year-old Caucasian male. His mother dated the earliest recognisable attack of involuntary movements to the age of two months. Episodes are preceded by a prodromal sensation of tingling in the skin and tightness of the muscles accompanied by a feeling of inner tension and restlessness. The onset is with choreoathetoid movements starting in the hand or foot. These spread after about 10 minutes to involve both ipsilateral limbs and are accompanied by dystonic posturing of the limbs. Speech becomes progressively more slurred during an attack, frequently resulting in anarthria at the height of an attack. The ipsilateral half of the face is usually involved, causing grimacing movements and often causing the patient to dribble saliva from the comer of his mouth, or on occasions to bite his tongue. Attacks may remain unilateral, affecting the right or left sides with approximately equal frequency. If severe, attacks progress to become generalised, affecting all limbs and the muscles of the neck and trunk causing chorea, dystonia and ballism intermixed. The patient feels warm and perspires profusely during a severe attack but does not experience pain. The attack reaches a peak approximately 30 minutes after the onset of symptoms, and the average duration of symptoms is between 30 minutes and 2 hours 63 (range: 20 minutes to 7 hours). The severity of symptoms may wax and wane during the attack. Afterwards he feels exhausted and will often sleep. He remains fully aware with no disturbance of consciousness during the episode, and he returns to normal between attacks. The only intervention capable of cutting short an attack is sleep. He is often able to abort an attack completely after the onset of the prodrome or during its early stages by sleeping for as short a period as five minutes, after which he will awake having returned to normal. If he waits until the attack is fully evolved, he is usually unable to sleep and so can not abort the attack which will then run its course. During an attack, he finds that vigorous exercise such as press-ups or weight lifting will ameliorate the dyskinesia but only while he is engaged in the exercise. Drinking large quantities of cold fluids and allowing himself to relax or distract himself from the attack is also helpful. Several factors are capable of precipitating an attack, including: extremes of emotion (particularly anger or excitement), stress, tiredness, hunger, intercurrent illness, alcohol and caffeine (one or two cups of strong coffee or tea may precipitate an attack after a delay of one to three hours). Sudden movement, however, never precipitates an attack. Episodes may occur without apparent precipitants, however, and there is a clear diurnal pattern, with attacks occurring most frequently in the late afternoon or evening. The frequency and duration of attacks of PDC has decreased with increasing age. Attacks were at their most frequent in childhood and adolescence, often occurring on a daily basis and usually lasting between one and four hours. They now occur approximately once per month and there is a tendency for clustering of attacks. Despite the disabling nature of these attacks, the patient has been able to continue to work as a driver. He very rarely experiences attacks while driving, and he attributes this to avoiding tiredness and hunger, and to the concentration required. He has tried phenytoin and clobazam without improvement. Clonazepam taken after a precipitant such as an alcoholic drink is of some benefit but is not helpful taken during an attack. Previous investigations have included computerised tomography (CT) of the brain and EEG recordings both between and during attacks, all of which were normal. Physical and neurological examination is normal. 64 3.3.1.2 Case VI-1 A 22-year-old Caucasian female. Onset was at the age of approximately six months with paroxysms of stiffening, stillness and crying. Episodes begin as hemi-dystonia with involuntary dystonic posturing of the arm and leg, usually on the left side. The foot becomes plantar flexed and inverted, adopting an equinovarus posture; the hand is clenched with a flexed wrist; the arm extended at the elbow and internally rotated at the shoulder. Torticollis is usual. Most attacks progress to become generalised affecting all limbs within 30 minutes, and involve a mixture of dystonia, chorea and ballism. She is unable to walk during an attack. Speech is affected, causing slurring dysarthria or anarthria. Dystonia is often so severe that the shoulder dislocates anteriorly, and on one occasion the right clavicle spontaneously fractured during a dystonic spasm. Consciousness in not impaired during attacks, and she has no symptoms between attacks. Most episodes last from 2 to 12 hours but may last as long as 24 hours on occasions. Attacks occur up to three times daily and there has been no reduction in frequency with increasing age. Precipitants include emotion, stress, tiredness, hunger, and caffeine (including consumption of caffeine-containing carbonated drinks and chocolate bars). Small amounts of alcohol have not precipitated attacks. There is no apparent diurnal pattern but attacks are more frequent and severe around the time of menstruation. Several drugs have been tried without benefit, including: phenytoin, sodium valproate, diazepam, benzhexol, baclofen and clonazepam. Physical examination is unremarkable between attacks. Investigations have included CT of the head and two EEG recordings between attacks, all of which were normal. 3.3.2 Clinical features in other family members Eighteen family members, other than the two index cases, reported attacks of abnormal movements typical of PDC at interview. A summary of selected clinical details for these individuals is given below. No family members suffered from exertional cramps which have previously been suggested to be a forme fruste of PDC (Schloesser et al, 1996). 65 3.3.2.1 Age at onset In all cases the onset of symptoms was in early childhood. Exact age at onset is not known for many family members, most of whom had had symptoms as long as they were able to remember. However, symptom onset is known to have been at the age of two months and six months in two individuals, and in the second year of life in seven others. 3.3.2.2 Attack phenomenology Witnessed attacks in two individuals (V-19 and VI-14) consisted of generalised choreoathetosis and dystonic posturing with involvement of speech. Most affected individuals described similar attacks consisting of a mixed movement disorder with elements of chorea, dystonia and ballism. Three subjects described attacks consisting largely of sustained dystonic posturing without apparent chorea. Attacks started in a limb in all cases, and usually progressed to hemi-dystonia and choreoathetosis. Some attacks spontaneously dissipated after just a few minutes, before symptoms progressed beyond the hand or foot. At some time all affected individuals had experienced generalised attacks affecting all limbs. Speech was involved during severe attacks in all cases. There was considerable person-to-person variation in the usual severity of attacks. Many individuals were, or had been, severely disabled by the condition while others felt it interfered only slightly with their lives. 3.3.2.3 Attack frequency All adult cases reported a marked decline in attack frequency as they grew older, which was often described as ‘growing out o f the attacks. For most adults the maximum frequency of attacks was within the range once per day to once per week, and occurred during childhood or adolescence. The oldest surviving family member aged 85 experienced attacks approximately once or twice per annum and these were always mild. One woman aged 67 had not had an attack for seven years, since moving from a house in which a long- running dispute with a neighbour had resulted in one or two attacks per month. 3.3.2.4 Attack duration Attack duration varied from 10 minutes at the shortest to as long as 12 hours. The majority of episodes fell within the range of 30 minutes to two hours. Attack duration was inversely 66 related to patient age with all adults reporting a decline in average attack duration as they grew older. 3.3.2.5 Precipitants All eighteen cases reported stressful situations and extremes of emotion, such as anger and excitement, as regular precipitants of an attack. These were the most important precipitating factors for many individuals. Of the 18 cases, 13 found caffeine to be a precipitant. Alcohol was a precipitating factor in all those individuals who did not abstain from it, but the amount of alcohol required to trigger an attack varied greatly from person to person. One individual found that drinking very modest amounts of alcohol in a social situation made him more relaxed and consequently less likely to suffer an attack. Sleep deprivation could precipitate attacks in all but three individuals. Hunger and intercurrent illness were other commonly reported precipitants. Two individuals found that cold could act as a trigger, and three reported prolonged exercise to be an infrequent precipitant. Attacks were more likely to occur around the time of menstruation in three female patients. No subjects reported any effect of smoking on the duration or frequency of episodes, but one man described an attack of unprecedented severity at the time he was giving up smoking. Sudden movement or startle were never precipitating factors. 3.3.2.6 Sleep benefit and other relieving factors With the exception of one of the index cases, all adults reported the same remarkable response to sleep. In all cases as little as five to ten minutes of deep sleep shortly after the onset of symptoms would be sufficient to abort an attack. Where sleep was not possible, for example as a result of rapidly worsening involuntary movements, there was little that could be done to alter the course of an attack. Five subjects found vigorous exercise during an attack to be helpful if the attack was not severe. 3.3.2.7 Diurnal pattern Most affected individuals were aware of a diurnal fluctuation in predisposition to attacks. Subjects were most susceptible to attacks in the afternoon and evening. Very few attacks occurred in the morning. No patients or spouses reported attacks during sleep. 67 3.3.2.8 Diplopia Diplopia was associated with attacks in two individuals: in one subject there was horizontal separation of images during severe attacks, in another diplopia had occurred on only two occasions. 3.3.2.9 Treatment In general, drug treatment was not successful in most members of this family. Clonazepam was felt to be of modest benefit in three of the six patients to whom it had been administered. Other drugs which had been tried and found not to be effective included: acetazolamide (one patient); phenytoin (one patient); sodium valproate (four patients); carbamazepine (one patient); vigabatrin (one patient); ethosuximide (one patient); diazepam (two patients); baclofen (one patient); benxhexol (one patient) and salbutamol via metered dose inhaler which made attacks worse (one patient). 3.3.2.10 Clinical investigations EEGs were recorded in nine patients, in two cases during attacks. All were normal apart from bilateral non-specific abnormality in one case. Three patients had had CT scans of the head which showed no abnormality. Magnetic resonance imaging of the brain was normal in one patient. 3.4 GENETIC LINKAGE ANALYSIS IN THE ‘B’ FAMILY - METHODOLOGY To localise a gene for PDC in the ‘B’ family, a systematic search for genetic linkage was performed using a panel of highly polymorphic, fluorescently labelled, microsatellite markers spanning the genome (as described in section 2.4). The limited understanding of the pathophysiological basis of PDC means that possible candidate genes for PDC are difficult to predict. The genome search was therefore performed using microsatellite markers distributed randomly throughout the genome. However, markers in the vicinity of voltage-gated ion channels were given priority in view of the suggestion that ion channel mutations may be the cause of this paroxysmal disorder. During the initial genome search, a set of 36 individuals from the ‘B’ family were genotyped for microsatellite markers. By limiting the sample set to 36, all samples could be analysed in the same electrophoresis run for each marker set, thus maximising the 68 efficiency of genotyping without compromising the power to identify linkage. Where LOD scores were greater than 1.0, all available family samples were analysed. Two samples (individuals V-6 and V-22 in figure 3.2) did not become available until linkage had been established. Linkage analysis was performed using the FASTLINK 3.OP version of MLINK, from the LINKAGE program package (Section 2.4.8). An autosomal dominant model of inheritance was used with a disease penetrance of 80% by the age of eight. In the ‘B’ family, 26 out of 66 at-risk family members were affected, suggesting a penetrance of 79% in this pedigree. This figure depends on accurate recall by family members of affectation status of individuals in generation II. However, this figure is broadly in line with the 84% penetrance calculated by combining data from all the PDC kindreds with an unequivocal phenotype reported before the outset of this study (Mount and Reback, 1940; Forssman, 1961; Weber, 1967; Richards and Barnett, 1968; Lance, 1977; Tibbies and Barnes, 1980; Byrne et al, 1991). An estimated disease allele frequency of 0.0001 was used. Marker allele frequencies were assumed to be equal. 3.5 RESULTS OF GENOME-WIDE LINKAGE ANALYSIS Forty-four microsatellite markers distributed over 13 chromosomes were analysed initially. The results of pairwise LOD scores for these markers are given in Appendix 1. LOD scores are negative at 9= 0 - 0.05 for all markers and scores are more negative than -2.0 at 0= 0 - 0.01 for all markers, excluding linkage of PDC in the immediate vicinity of these markers. The highest LOD score amongst this set of markers is 0.92 at 0= 0.2 for the marker D7S502. During the course of this study, linkage of a gene for PDC to markers on the long arm of chromosome 2 was reported in two families by other researchers (Fink et al, 1996; Fouad et al, 1996). In the Polish-American family reported by Fink and co-workers (1996), the PDC locus (named FPD1) was assigned to a 14 cM interval (using map distances from the Genethon genetic map, Dib et al 1996, or on electronic server at http://www.genethon.fr) between the flanking markers D2S164 and D2S159. Fouad and others (1996) simultaneously reported an Italian kindred in whom PDC mapped to an overlapping interval of approximately 12 cM, between the markers D2S128 and D2S126. The clinical 69 phenotype of both families is consistent with PDC (see table 3.1). Consideration of the data from the two families together, suggests that the PDC locus is in the 8 cM interval between D2S164 and D2S126 on chromosome 2q (figure 3.3). 3.5.1 PDC in the ‘B’ family is linked to chromosome 2q Following identification of linkage of PDC to chromosome 2q, this region was studied in detail in the ‘B’ family using 15 closely spaced microsatellite markers mapping to the region. All but one of the markers were from the Genethon map (Dib et al, 1996), the marker D2S1242 was from the Utah map (The Utah marker development group, 1995). The markers studied and genetic distances between markers are shown in figure 3.3. PCR primers were fluorescently tagged and analysed as described in section 2.4. All 46 family members for whom DNA was available were genotyped for chromosome 2 markers. The results of pairwise LOD scores between PDC and chromosome 2q microsatellite markers are presented in table 3.3. All markers yielded positive LOD scores. Only markers D2S377 and D2S344 were relatively uninformative in this family, producing small positive scores. The highest LOD scores were obtained with the markers D2S1242 (Zmax= 8.7 at 9= 0), D2S2359 (Zmax= 8.12 at 0= 0) and D2S2250 (Zmax= 7.58 at 0= 0). These data robustly confirm the assignment of the PDC locus, FPD1, to chromosome 2q in the CB’ family. 3.6 RESULTS OF HAPLOTYPE ANALYSIS OF CHROMOSOME 2q MARKERS In order to identify the flanking recombinations defining the PDC locus in the ‘B’ family, haplotypes for chromosome 2q microsatellite markers were analysed. Haplotypes were constructed manually by minimisation of recombination events between markers. Haplotypes for seven markers are given in figure 3.4. Informative recombinations were observed in individuals V-5, V-9 and VI-14. Recombinations in individuals V-5 and V-9 place the disease gene telomeric to the marker D2S295. The next telomeric marker studied, D2S173, at 0.7 cM from D2S295, is homozygous for the affected parents of V-5 and V-9, and is therefore not informative for these meioses. A recombination in individual VI-14 indicates that the PDC gene is centromeric to the marker D2S344. Markers D2S173, D2S2250, D2S1242, D2S163, D2S120 and D2S2359 all segregate with PDC in this family. Thus, the PDC candidate interval in the ‘B’ family spans 3.8 cM between D2S295 and D2S344, at 2q36-q37 (figure 3.3). 70 Figure 3.3: Genetic map of part of distal chromosome 2q, showing selected microsatellite markers and PDC candidate regions.Markers D2S206 and D2S331 are telomeric to D2S159. Sex-averaged genetic distances are given in centimorgans according to the Genethon chromosome 2 map at: http://www.genethon.fr. Candidate intervals for PDC in the 'B' family and in the families described by Fouad et al (1996) and Fink et al (1996) are indicated. cen D2S157 4.5 cM D2S128 0.4 cM D2S143 4.5 cM D2S164 1.1 cM Fouad et al D2S295 11.7 cM 0.1 cM D2S173 0.7 cM D2S2250 'B' family 1.7 cM 3.8 cM D2S163 / D2S120 / D2S2359/ 1.3 cM D2S1242* D2S344 Fink et al 1.3 cM 14.1 cM D2S377 0.6 cM D2S126 1.3 cM D2S130 6 cM D2S159 q-ter * Position of marker D2S1242 according to the Marshfield integrated genetic map (at: http://www.marshmed.org/genetics/). 71 Table 3.3: Results of two-point LOD scores between PDC and chromosome 2q markers in the ‘B’ family. Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D2S157 -6.65 -2.78 0.08 1.85 2.31 2.23 1.69 0.9 D2S164 — o o 0.11 2.00 3.07 3.22 2.77 1.96 1.01 D2S295 -2.23 0.63 2.5 3.46 3.5 2.89 1.95 0.9 D2S173 1.87 1.86 1.83 1.66 1.45 1.03 0.61 0.24 D2S2250 7 .58 7.57 7.47 7.01 6.37 4 . 94 3.33 1.59 D2S1242 8.7 8.68 8.57 8.03 7.3 5.69 3.86 1.87 D2S163 5.08 5.07 5.0 4.68 4.24 3.26 2.18 1.04 D2S120 6.77 6.76 6.66 6.19 5.57 4 .23 2.77 1.26 D2S2359 8.12 8.11 7.99 7.45 6.72 5.13 3.39 1.6 o o D2S344 — o o -1.89 -0.9 -0.26 i 0.07 0.05 0.02 D2S377 -1.26 0.44 1.41 1.97 2.05 1.76 1.24 0.62 D2S126 0.81 2.5 3.77 4.13 3.97 3.23 2.23 1.09 D2S130 3.08 4.78 5.66 5.87 5.52 4.41 3.05 1.54 O CM O D2S206 — o o -1.96 1 1.22 1.57 1.56 1.18 0.64 D2S331 — o o -7.3 -3.38 -0.76 0.18 0.77 0.77 0.48 Table 3.4: Results of two-point LOD scores between PDC and candidate gene polymorphisms in the ‘B’ family. Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 SLC4A3 6.08 6.07 6.0 5.67 5.2 4 .08 2.76 1.34 VO 00 r" CHRND I -3.49 -1.52 -0.22 0.21 0.39 0.27 0.11 72 in to co co co y- c o IT) <0 CO CO CO i — CO h . m CO N- CO LO CO o > CO h » CM T ” N CO CM CO CO CM T3 LO CO CO CO CO 1 - CO B *5 CO CM LO 1 - T - LO cd X CO CM 1 - CO CM CM T - ^ D V3 co co t- in co r- ifl O jK cn° - o cn U £ < C/0 L* c d c o i n c o cm t - in . o § ^ c o c CO lO ^ CO CM r - Oh 5 CO CM CM i n CM T - O ’" T c o S ^ .s 0 0 CO CM CO CO T - CM CO CO CM CO CO T - T - 0 0 CO CM CO CO r-> CO CM CO CO CM LO . 2 2 g o 2 CO CO CM CO CO i - t - co co i- in co c o c o v- in co t- m ^ 2 S i n c o ■*- c o c o cm c o C° /3 £V Q W e * * ? O N 0) C '■ u N m c o cm c o c o t— t o c d d> x c r^ . c o cm c o c o cm i n CO e ^ • a .£ ^ d ) / — CO CO CM CO CO T - T - ^r^coe^m coi-m | CO CO CM CO CO T - r - rs .2 co N t - - t- CO CO r - S T3 vo ■ k CM CO CM (O CO CM t - N S j _ , 4—» r- i- •»- 1- T- r- 4t t 4- N 3 « X •4—1 CO o «u i^'cd e s S I CO CO CM CO ^ T - rj- w o c l , £ co to i- in co i— m 3.7 CANDIDATE GENE LINKAGE ANALYSIS After assigning the PDC locus to chromosome 2q36-q37, a database search of genes known to map to this cytogenetic region was performed to identify possible candidates for the PDC disease gene. The positional candidate approach, when successful, has the advantage of circumventing the process of physical mapping which follows identification of genetic linkage. As current understanding of the pathophysiological basis of PDC is limited, identification of candidate genes in this disorder is difficult. However, potential candidates include genes expressed in the brain and genes coding for ion channels or involved in neurotransmission or regulation of neuronal excitability. Two genes located in the vicinity of chromosome 2q36-q37 may be considered putative candidates for the PDC gene. The chloride/bicarbonate anion exchanger (gene symbol SLC4A3), a member of the family of solute transporters, has been mapped to the interval between D2S126 and D2S164 at 2q36 (Su et al, 1994). The delta polypeptide of the nicotinic acetylcholine receptor (gene symbol CHRND) has been mapped to the cytogenetic location 2q33-q34 (Beeson et al, 1990), although a more distal localisation has been proposed (Pasteris et al, 1993). Both genes contain polymorphic tandem repeat sequences suitable for linkage studies (Landa et al, 1994; Su et al, 1994). Linkage analysis involving these polymorphisms was performed using fluorescently tagged primers, as described in section 2.4. Two-point LOD scores between these markers and PDC are given in table 3.4. The CHRND polymorphism gives a LOD score of -7.68 at a recombination fraction of 0. The SLC4A3 polymorphism gives a LOD score of 6.08 at a recombination fraction of 0. There were no recombinations observed between PDC and SLC4A3. These data exclude CHRND as a candidate for PDC, but indicate that the PDC and SLC4A3 74 genes are in close physical proximity. These data are consistent with the contention that SLC4A3 may be the PDC gene. 3.8 DISCUSSION 3.8.1 Comparison with other PDC families This study identified a large British family with PDC. The family contains 27 affected members, 20 of whom were still living at the outset of the study, making this one of the largest families with PDC yet to be reported. The clinical phenotype in 20 affected family members was determined by interview. The description of attacks and their associated features is remarkably consistent between family members, many of whom did not know one another, and is typical of the clinical features of PDC described previously in other families (table 3.1). The careful clinical descriptions of attacks of PDC in classic early papers by Mount and Reback (1940), Forssman (1961), and Lance (1963), are almost identical to attacks described by members of the ‘B’ family. Features shared in common between the ‘B’ family and most other PDC families include: the phenomenology of the attacks themselves; the early childhood onset and tendency for improvement with age; the male preponderance; the duration of attacks; the wide range of precipitating factors; the remarkable sleep responsiveness of the condition, and the extensive family history indicating autosomal dominant inheritance with high penetrance. The peri-menstrual exacerbation of symptoms, diplopia during attacks, beneficial effect of exercise or drinking fluids, and occasional precipitation of attacks by cold or prolonged exercise are also occasional features in other PDC kindreds. 3.8.2 Linkage analysis The results of linkage analysis in the ‘B’ family confirm the assignment of the PDC gene, FPD1, to the distal long arm of chromosome 2. Fink et al (1996) mapped the FPD1 locus to a 14 cM interval between the flanking markers D2S164 and D2S159, and Fouad et al (1996) assigned the locus to an overlapping 12 cM interval between D2S128 and D2S126. Thus, the FPD1 locus is situated within the 8 cM interval between D2S164 and D2S126. Identification of critical chromosome 2q recombinations in the ‘B’ family allow further refinement of the FPD1 candidate interval. These recombinations place the locus in the 3.8 cM interval between the markers D2S295 and D2S344. The reduction in size of the FPD1 75 critical region, from 8 cM to 3.8 cM represents a significant refinement of the genetic map position of the locus which will facilitate positional cloning of the disease gene. Since this work was completed, two further families with PDC have been mapped to chromosome 2q, in both cases to an interval overlapping the candidate region identified in the ‘B’ family (Hofele et al, 1997; Raskind et al, 1998). Raskind et al mapped FPD1 to the 6.2 cM region bounded by D2S164 and D2S377, while Hofele et al mapped the locus to an 8.1 cM region between D2S143 and D2S2359. The data of Raskind et al do not help to localise the FPD1 gene any further, but Hofele et al have identified a new distal flanking recombination in their family (with D2S2359) which, in combination with the centromeric recombination in the ‘B’ family (D2S295), narrows the FPD1 interval to 2.5 cM. Several lines of evidence indicate that the assignment of the FPD1 locus to chromosome 2q is unlikely to be a false positive result. Firstly, the maximum LOD score achieved (Z=8.7 at 0= 0 for marker D2S1242) is considerably in excess of the conventional threshold for accepting linkage (Z = 3.0). Secondly, several adjacent markers give significant positive LOD scores suggesting that genotyping or data entry errors are unlikely to explain the result. Thirdly, the dramatic nature of the attacks reported by family members and their relatives at interview makes designation of affected status uncomplicated. This is important because correct assignment of phenotype to family members is essential for genetic linkage studies. A potential source of error, however, is the possibility that non-genetic phenocopies exist within the family, possibly with a psychogenic basis. The finding that all apparently affected individuals carry the disease haplotype makes this less likely. Fourthly, PDC attacks were witnessed by the author in two of the affected individuals in whom critical proximal and distal recombinants were identified ( V-9 and VI-14 ); objective evidence of affected phenotype in these individuals increases confidence in the mapping of the FPD1 locus in the ‘B’ family. Most importantly, the mapping of PDC in the ‘B’ family to the same locus as four other families with an almost identical phenotype is strongly in favour of the assignment of the disease locus to 2q in this family. The finding that PDC in the ‘B’ family maps to chromosome 2q, adds to the evidence for genetic homogeneity of PDC. The diverse geographical origins of the five families linked to chromosome 2q to date suggests that the families may not be related. However, three families with PDC, including the original family of Mount and Reback, originate from Scotland (Mount and Reback, 1940; Richards and Barnett, 1968; Walker, 1981; Byrne et 76 al, 1991) and it is therefore possible that these families are ancestrally related to the ‘B’ family and descended from a single founder. A dominantly inherited paroxysmal movement disorder with similarities to PDC, named ‘paroxysmal choreoathetosis/spasticity’ has been assigned to a locus on chromosome lp in a German kindred during the course of this study (Auburger et al, 1996). However, a constant spastic paraplegia was a prominent feature of the phenotype in this family, and the age of onset was relatively late in some individuals. The attacks also differed from those in the ‘B’ family and in others with classical PDC, in that headache was a feature and physical exercise a precipitant. Therefore, PDC and the complicated form of paroxysmal choreoathetosis described by Auburger et al appear to be distinct both clinically and genetically. 3.8.3 Analysis of candidate genes Linkage analysis was performed using dinucleotide repeat polymorphisms within two putative candidate genes mapping in or near the cytogenetic region 2q36-q37. A third candidate gene, encoding a 5-hydroxytryptamine receptor, was also identified. 3.8.3.1 CHRND An epilepsy syndrome characterised by motor seizures with hyperkinetic or tonic manifestations and the retention of awareness, named autosomal dominant nocturnal frontal lobe epilepsy (ADNFLE), shares features in common with PDC. The involuntary movements associated with ADNFLE may be mistaken for PDC in clinical practice (Scheffer et al, 1995) and the suggestion that PDC may itself be a form of epilepsy remains open to debate. Mutations in the gene encoding the alpha 4 subunit of the neuronal nicotinic acetylcholine receptor have been shown to cause ADNFLE in some families (Steinlein et al, 1995). The gene encoding the delta subunit of the nicotinic acetylcholine receptor is therefore also a possible candidate gene for FPD1. However, it should be noted that the delta subunit is a component of the neuromuscular junction acetylcholine receptor and has not to date been identified in neuronal acetylcholine receptors. The significantly negative LOD score between CHRND and PDC (-7.68 at 0= 0) indicates that CHRND is not close to the FPD1 locus and excludes it as a candidate. 77 3.8.3.2 SCL4A3 SLC4A3 is the third member of the band 3-related family of anion exchangers and is expressed in predominantly in brain and heart (Alper, 1991). The brain isoform of the protein is expressed on neurons throughout the brain, particularly in the deep pontine grey matter, the caudal medulla and, interestingly, a high level of expression is also seen in the substantia nigra of the mouse brain (Kopito et al, 1989). SLC4A3 is a membrane-bound protein which functions as a sodium-independent chloride/bicarbonate exchanger and an alkali extruder. It is believed to play a role in the regulation of intracellular pH, cell volume and in regulation of intracellular chloride concentration (Kopito et al, 1989). Transient changes in intracellular pH are known to affect neuronal excitability (Gruol et al, 1980). Impaired or altered function of the SLC4A3 protein might therefore be expected to lead to changes in intracellular pH and thus neuronal excitability, which has been suggested as a possible model of PDC pathophysiology (Ptacek, 1997). Other mechanisms by which altered chloride/bicarbonate exchange might affect neuronal firing may also be envisaged. GABA and glycine are important inhibitory neurotransmitters which open anion channels permeable to both chloride and bicarbonate ions. Chloride/bicarbonate exchange plays an important part in maintaining the intracellular chloride concentration above its equilibrium concentration (Grinstein, 1987). Thus, dysfunction of the exchanger might be expected to alter anion gradients across the cell membrane and so to affect inhibitory neurotransmission. Furthermore, changes in the chloride gradient may also have implications for the re-uptake of GABA into neurons by the GABA/sodium/chloride cotransporter, which is dependent on transmembrane sodium and chloride gradients. Finally, like its erythrocyte homologue, SLC4A3 may serve as a membrane anchor for the neuronal cytoskeleton, so contributing to the structural organisation of the neuron (Kopito et al, 1989). Linkage analysis using the SLC4A3 polymorphism yielded a LOD score of 6.08 at a recombination fraction of 0. There were no recombinations between PDC and the SLC4A3 polymorphism in the ‘B’ family. The absence of recombination between PDC and SLC4A3, its expression pattern in the central nervous system, and physiological importance in the regulation of intra-neuronal homeostasis, all provide support for a role of this gene in PDC. Further analysis of the SLC4A3 gene was not undertaken as part of this 78 study as although the cDNA sequence of the gene is known, its structural organisation (i.e. intron-exon boundaries) has not yet been described. 3.8.3.3 5-HT2B is a candidate gene 5-hydroxytryptamine (5-HT) is an important neurotransmitter in the central nervous system. The gene for the 5-HT receptor, type 2B (5-HT2B), has been demonstrated to map to the cytogenetic region 2q36.3-q37.1 by fluorescent in-situ hybridisation (Le Coniat et al, 1996). The major sites of expression of 5-HT2B mRNA are the liver and kidney, but weak expression is also observed in the brain (Le Coniat et al, 1996). Although 5-HT is not known to be involved in neurotransmission in the basal ganglia (Albin et al, 1989; Ferrari et al, 1997), this gene should nevertheless be considered a putative candidate for FPD1. The 5-HT2B receptor gene is not known to contain polymorphisms suitable for linkage analysis. 3.9 FUTURE DIRECTIONS OF STUDY SLC4A3 is a strong candidate for the PDC disease gene. Further investigation by means of mutation screening of SCL4A3 in PDC families is warranted. Comprehensive mutation analysis directly from genomic DNA is not possible until the intron-exon organisation of the gene is known. Determination of the genomic organisation of SLC4A3 and mutation analysis in PDC patients is currently underway in the laboratory of Dr L. Ptacek at the University of Utah. To date no SLC4A3 mutations have been identified by screening using single-strand conformation polymorphism analysis. Other potential candidate genes mapping to the vicinity of the PDC locus such as 5-HT2B should also be investigated by screening for mutations as part of a positional candidate approach. As the Human Genome Project progresses, and new genes on chromosome 2q are identified, further candidates will become available for study. The 2.5 cM FPD1 candidate region is likely to represent a relatively large physical distance of up to 3 Mb. A physical mapping project to identify the FPD1 gene within this interval would represent a considerable undertaking. Further refinement of the genetic localisation of PDC would greatly facilitate physical mapping of the gene. This could be achieved using two approaches: firstly, study of additional large families with PDC by means of linkage analysis, may identify critical recombinants within the candidate region and so 79 narrow the interval. Families with PDC are rare, but follow up of previously reported large families (table 3.1) would be worthwhile where possible. Secondly, searching for linkage disequilibrium between PDC and linked chromosome 2q markers in familial and sporadic PDC patients, particularly those with British ancestry, may refine the position of the gene. Linkage disequilibrium depends upon the existence of a founder mutation in apparently unrelated families and sporadic cases; if these have arisen as independent mutations, linkage disequilibrium will not occur. The Scottish ancestry of at least three reported families suggests that a common founder may link some families (section 3.8.2). Functional studies of PDC pathophysiology are underway at the Institute of Neurology which will complement genetic studies. These include a positron emission tomography study using the selective D2 receptor ligand n C-raclopride to measure striatal dopamine receptor binding. This study, performed in collaboration with Dr N. Tuijanski at the MRC cyclotron unit of the Hammersmith Hospital, has shown normal D2 receptor binding in one patient with PDC to date. Magnetic resonance spectroscopy (MRS) is being performed in patients with PDC both during and between attacks, in collaboration with Drs C. Davie and S. Taylor-Robinson at the Institute of Neurology and the Hammersmith Hospital. 31 Preliminary results of P MRS have demonstrated normal brain pH before, during, and after attacks of PDC. This finding is of particular interest in view of the genetic data presented in this chapter suggesting that SLC4A3 may be implicated in the pathogenesis of PDC, possibly by altering neuronal pH. If confirmed, this finding would not support the hypothesis that changes in neuronal pH, mediated by SLC4A3 for example, cause PDC. !H MRS has demonstrated a normal N-acetylaspartate (NAA) peak within the lentiform nucleus in a patient with PDC indicating neuronal preservation. Studies of neurotransmitter metabolites in cerebrospinal fluid during and between attacks are also underway in collaboration with Dr S. Heales at the Institute of Neurology. Initial results in a patient with PDC have shown a rise in CSF homovanillic acid (HVA, a dopamine metabolite) and hydroxyindoleacetic acid (HIAA, a 5-HT metabolite) during an attack of PDC as compared to baseline (HVA: 165 nmol/L at baseline, 537 nmol/L during attack, normal range 71-565 nmol/L; 5-HIAA: 88 nmol/L at baseline, 221 nmol/L during attack, normal range 58-220 nmol/L). It should be emphasised that these are all preliminary results of work in progress which will need to be extended and confirmed. Insights into pathophysiology which may eventually be gleaned from these studies will hopefully enable more accurate identification of candidate genes from those known to map to the disease locus. 80 CHAPTER FOUR LINKAGE ANALYSTS IN FAMILIES WTTH HEREDITARY GENIOSPASM 4.1 OUTLINE OF CHAPTER This chapter describes a genetic linkage study to map a gene for an unusual focal movement disorder affecting the chin and lower lip - hereditary geniospasm. The clinical features of two British families with hereditary geniospasm are described and the results of a genomic search for the geniospasm locus in one of these families are presented. The results of linkage analysis in a second family using markers linked to a geniospasm locus are also described. These findings are discussed and future directions of study suggested. 4.2 INTRODUCTION Hereditary geniospasm (on-line Mendelian Inheritance in Man - OMIM number: 190100) is a dominantly inherited movement disorder first described more than one hundred years ago. It is characterised by rapid, paroxysmal, rhythmic up-and-down movements confined to the chin and lower lip due to involuntary contractions of the mentalis muscle. These movements give the appearance of a rapid quivering of the chin and lower lip. Episodes of geniospasm last from seconds to hours, and may be triggered by emotion, anxiety, concentration, or may occur without apparent precipitants. The condition typically becomes manifest in infancy or early life, and episodes tend to reduce in frequency with advancing age. Although the condition is sometimes distressing and embarrassing, in most cases there are no associated abnormalities and the condition is considered benign. The first description of hereditary geniospasm was by Massaro in 1894 who reported a family containing 26 cases over five generations. Massaro noted that the onset of the condition was shortly after birth, that attacks were precipitated by emotion, and were of variable intensity and duration. He coined the term ‘geniospasm’ to describe the 81 condition. Several other families with a dominantly inherited disorder, variously described as ‘hereditary quivering of the chin’, ‘hereditary chin-trembling’ or ‘facial spasm’ have subsequently been reported in the English language (Stocks, 1922; Goldsmith, 1927; Grossman, 1957; Wadlington, 1958; Laurance et al, 1968; Johnson et al, 1971; Alsager et al, 1991; Danek, 1993; Gordon et al, 1993; Soland et al, 1996; Destee et al, 1997). In all, some 24 families with hereditary geniospasm have been reported from several countries, including families described in the German language literature, reviewed by Danek (Danek et al, 1991). The clinical features are almost identical in all reported families: involuntary movements confined to the chin and lower lip, onset soon after birth or in early childhood, precipitation particularly by stress, emotion or concentration, and dominant inheritance. The condition usually occurs in isolation, and the reported associations with hereditary motor and sensory neuropathy type I (Danek, 1993), otosclerosis (Grossman, 1957; Wadlington, 1958), nystagmus (Frey, 1930; Laurance et al, 1968) and rapid eye movement behavioural disorder (Blaw et al, 1989) are probably incidental associations. Clinical investigations, including EEGs, have revealed no definite abnormality in affected individuals. Needle EMG studies recorded from the mentalis muscle have shown polymorphic discharges with a frequency of 6 - 30 Hz. Geniospasm is considered benign but can lead to considerable embarrassment and distress in some affected individuals, and may result in significant psychosocial morbidity (Gordon et al, 1993; Soland et al, 1996; Destee et al, 1997). Severe tongue lacerations may occur as a result of nocturnal tongue biting in young children (Johnson et al, 1971; Blaw et al, 1989; Soland et al, 1996). Treatment is often unnecessary but injection of botulinum toxin into the mentalis muscle has been used with success (Gordon et al, 1993; Soland et al, 1996). Hereditary geniospasm may present to neurologists or paediatricians but most affected family members do not seek medical attention, and it therefore seems likely that the true prevalence of the condition is higher than the relatively small number of reported families would suggest. It is difficult to fit geniospasm into the nosology of movement disorders. In discussing geniospasm in 1922, Stocks commented that ‘there has been much confusion in the classification and nomenclature of affections of this type’, a statement which remains true today. Geniospasm resembles a form of tremor and some authors have classified it 82 as a variant of essential tremor (Hubble et al, 1989). Patients with essential tremor may develop tremor of the jaw and tongue, but in such cases tremor is not confined to the chin, and is always accompanied by severe tremor elsewhere, particularly the arms (Biary and Koller, 1987; Bain et al, 1994). Moreover, tremor of the jaw results from activation of the masseters rather than mentalis. Several other characteristics of geniospasm differ from those associated with tremor: the early age of onset, sometimes only minutes after delivery (Destee et al, 1997), differs from the adult onset typical of most forms of essential tremor; movements tend to remit rather than progress with advancing age, geniospasm may persist during sleep in contrast to tremor, and finally, the absence of rhythmic activation on EMG recordings is quite unlike the pattern of muscle activation which characterises tremor. Focal tremor affecting the jaw or face may occur as a manifestation of dystonia, but is usually accompanied by sustained dystonic movements in patients or their relatives, and does not affect the chin in isolation (Jedynak et al, 1991). Some authors have proposed that geniospasm represents a form of myoclonus, possibly a focal form of hereditary essential myoclonus. They cite the brief burst duration of 10 - 25 ms and the wave form on EMG recordings similar to that seen in rhythmical myoclonus (Destee et al, 1997). However, persistence during sleep and lack of response to alcohol would be atypical for essential myoclonus. The absence of cortical potentials on EEG back averaging is against a cortical origin. The important differences between geniospasm and essential tremor, dystonic tremor or essential myoclonus suggest that geniospasm cannot be classified with any of these disorders. As suggested by Danek (1993), the unique clinical features of hereditary geniospasm are probably sufficient to justify its classification as a distinct entity. Brief episodes of geniospasm may occur as a normal phenomenon in many individuals, for example as a prelude to crying. However, this phenomenon is distinct from familial geniospasm where episodes are usually more prolonged and may be triggered by a wide variety of precipitants. Hereditary geniospasm might therefore be considered to be an extreme form of a normal biological trait. The cause of hereditary geniospasm is not known. The mentalis muscle fibres are innervated by the ipsilateral facial nerve nucleus in the brain stem (Welt and Abbs, 1990). The bilateral nature of the movement affecting most of the mentalis muscle, suggests that 83 the abnormality may lie either within the muscle itself or at the level of higher cortical or midbrain centres. Other paroxysmal disorders such as episodic ataxia and periodic paralysis are caused by abnormalities of voltage gated ion channels (Ptacek et al, 1991 and 1994; Browne et al, 1994) and it may be that this is a possible mechanism for the generation of involuntary muscle contractions in this condition. 4.2.1 Aims of the study As the cause of hereditary geniospasm is unknown, a positional cloning strategy was adopted to attempt to identify the gene or genes responsible for this disorder. A genetic linkage study was undertaken in a large family with hereditary geniospasm by means of a genomic search using microsatellite markers. Mapping a genetic locus for hereditary geniospasm may facilitate classification of the condition among other inherited movement disorders, and would represent the first step in positional cloning of the responsible gene(s). It is to be hoped that the eventual identification of these gene(s) will help elucidate the pathophysiology underlying this unusual disorder. 4.3 HEREDITARY GENIOSPASM FAMILIES AND METHODS Two unrelated British families with hereditary geniospasm were studied. The clinical features of both families have been reported in detail previously (Soland et al, 1996). For this study, blood was collected from all available family members for DNA extraction and the clinical details were updated by interviewing family members. 4.3.1 Family 1 Family 1 (figure 4.1) contains 20 affected individuals over four generations, 18 of whom are still living. Forty-two family members were interviewed (by Dr M.B. Davis and the author) and DNA samples obtained from 36 individuals. Affected individuals experienced characteristic involuntary up-and-down movement of the tip of the chin with a superimposed quivering activity and movement of the lower lip. Episodes of geniospasm were frequently precipitated by stress, concentration, or extremes of emotion, but could occur spontaneously. Episodes were usually several minutes in duration. The onset of symptoms was in infancy or early childhood in all cases, and geniospasm was observed as early as hours after birth in one family member. All adults reported a marked tendency for attacks to become less frequent as they became older, 84 the highest frequency occurring during childhood and adolescence. Attacks of geniospasm were witnessed in most individuals. Affected individuals were entirely normal between attacks. Five affected individuals suffered from nocturnal involuntary tongue-biting during early childhood. There were no other associated neurological abnormalities. 4.3.2 Family 2 Family 2 (figure 4.2) contains 16 affected individuals over four generations, 11 of whom are still living. All family members were interviewed by the author. The index case described chin quivering since early childhood, precipitated by emotion and excitement, and sometimes severe enough to interfere with speech and drinking. Episodes of geniospasm could occur during sleep. He found the movements extremely embarrassing and socially inhibiting. Treatment with botulinum toxin injections had almost completely abolished the movements. Other family members were similarly but less severely affected. Episodes of geniospasm were witnessed by the author in two family members. The clinical descriptions provided by family members, and the episodes of geniospasm witnessed by the author were indistinguishable in the two families. The clinical features of geniospasm in the two families were also very similar to those described in previously reported families. 4.3.3 Methods A fluorescently labelled set of microsatellite markers (section 2.4.2) was used to undertake a genomic search for the hereditary geniospasm locus in family 1. DNA samples from family 2 became available at a late stage in the study. Areas yielding positive LOD scores in family 1 were subsequently studied for evidence of linkage in family 2. Detailed genetic mapping using additional closely spaced markers was performed around markers with LOD scores > 1.0. Additional microsatellite markers used for fine genetic mapping were mostly from the Genethon map (Dib et al, 1996). Details of all markers used are available from the Genome Database (GDB) at http://www.gdb.org/. 85 < z TJ o > Figure 4.2: Hereditary geniospasm - family 2. Family members from whom DNA was available for study are indicated by asterisks. The index case is indicated by an arrow. Birth order, and gender in some individuals, have been altered to make the • SP CX 4.4 RESULTS OF THE GENOME-WIDE SEARCH FOR THE HEREDITARY GENIOSPASM LOCUS IN FAMILY 1 One hundred and four microsatellite markers on 16 chromosomes were analysed in family 1. During the initial search, evidence supporting linkage of hereditary geniospasm to the marker D9S175 was obtained (Zmax= 2.5 at 0 = 0.1). There was no evidence to support linkage of the locus to any of the other markers screened (the highest LOD score obtained among all other markers in the initial search was 0.84 at 0 = 0 for D6S281). Pairwise LOD scores for all microsatellite markers are presented in appendix 2. Following initial linkage to D9S175, this region was examined in detail using fourteen closely spaced microsatellite markers surrounding D9S175. The Galton chromosome 9 map places these in the order: pter-D9S147E (9pter-qter) -interlocus distance not accurately determined (Reed et al, 1994) - D9S161 (9p21) - 3.8 cM - D9S43 (9p21) - 3.4 cM - D9S1791 (9pter-qter) - 5.8 cM - D9S200 (9pl2-p21) - 4.1 cM - D9S166 (9p21-q21) - 0.5 cM - D9S1806 (9pter-qter) - 0.2 cM - D9S1822 (9pter-qter) - 0.4 cM - D9S276 (9ql3-q22) - 0 cM -D9S1837 (9pter-qter) - 0.2 cM - D9S1876 (9pter- qter) - 1.3 cM - D9S175 (9ql3-q21) - 4 cM - D9S1834 (9pter-qter) - 5.8 cM - D9S153 (9ql3-q22) -15.5 cM - D9S257 (9ql3-q22) -qter (figure 4.3). Distances between markers are sex-averaged. 88 Pairwise LOD scores between chromosome 9 markers and hereditary geniospasm in family 1 are given in table 4.1 (at 80% penetrance). Maximum LOD scores were obtained with the markers: D9S1822 (Zmax= 4.7 at 0 = 0), D9S1837 (Zmax= 5.24 at 0 = 0) and D9S1876 (Zmax= 4.91at 0 = 0). LOD scores were also calculated with the penetrance of geniospasm set at 100% (table 4.2). Although maximum LOD scores were lower for most markers at 100% penetrance, particularly at the lower recombination fractions, LOD scores > 3 were obtained for three markers, supporting linkage to chromosome 9 (D9S1822 Zmax= 3.6 at 0 = 0.1, D9S1837 Zmax= 4.12 at 0= 0.1 and D9S1876 Zmax= 3.77 at 0= 0.1). These data allow the locus for hereditary geniospasm to be assigned to the proximal long arm of chromosome 9 in family 1. The nomenclature committee of the Human Genome Organisation have assigned this locus the gene symbol ‘GSM1’. 4.5 RESULTS OF HAPLOTYPE ANALYSIS IN FAMILY 1 Manually reconstructed haplotypes for nine chromosome 9 markers are shown in figure 4.4. Construction of haplotypes allows definition of a candidate interval of 2.1cM between the flanking markers D9S1806 and D9S175, at the cytogenetic region 9ql3- q21 (figure 4.3). Recombinations in individuals 111-10 and IV-5 place GSM1 distal to D9S1806 and proximal to D9S175 respectively. No recombinations were observed between geniospasm and the markers D9S1822, D9S1837, D9S276 and D9S1876. Four family members (IV-1, IV-3, IV-7 and IV-16) had inherited the associated haplotype as a whole or in part but were unaffected. All four asymptomatic gene carriers were older than 13 years of age and would therefore be expected to have developed symptoms by this age given the early onset of geniospasm in most cases. 4.6 RESULTS OF LINKAGE ANALYSIS USING CHROMOSOME 9 MARKERS IN FAMILY 2 To determine whether hereditary geniospasm in family 2 is also linked to the GSM1 locus, markers spanning the GSM1 critical region on chromosome 9 were analysed in family 2. Pairwise LOD scores for family 2 are given in table 4.3. All markers yielded LOD scores < -2.0 from 0 = 0 to 0 = 0.01, thus excluding geniospasm in family 2 from the GSM1 locus. 89 Table 4.1: Two-point LOD scores between hereditary geniospasm in family 1 and microsatellite markers on chromosome 9. Penetrance of geniospasm set at 80%. Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D9S147E - o o -1.22 -0.20 0.54 0.83 0.94 0.74 0.35 D9S161 0.10 0.10 0.15 0.30 0.42 0.49 0.43 0.26 D9S43 — o o 0.42 1.41 2.04 2.18 1.99 1.46 0.72 D9S1791 — o o 0.47 1.47 2.10 2.25 2.05 1.52 0.77 D9S200 -8.02 0.90 1.85 2.31 2.31 1.92 1.32 0. 60 D9S166 — o o 1.43 2.40 2.94 2.99 2.60 1.88 0. 94 D9S1806 -15.28 1.77 2.74 3.26 3.28 2.83 2.05 1.02 D9S1822 4.70 4.69 4.65 4.43 4.10 3.30 2.30 1.12 D9S1837 5.24 5.23 5.18 4.93 4 .55 3.64 2.54 1.24 D9S1876 4.91 4.91 4.86 4.61 4 .25 3.39 2.35 1.14 D9S175 -2.86 0.83 1.81 2.40 2.50 2.22 1.62 0.80 D9S1834 -3.28 -0.41 0.62 1.37 1.66 1.67 1.31 0.66 D9S153 -4.31 -0.61 0.42 1.17 1.46 1.51 1.19 0.61 o n i —i i— D9S257 — o o -5.61 -2.62 i 0.32 0.88 0.83 0.44 90 Table 4.2: Two-point LOD scores between hereditary geniospasm in family 1 and microsatellite markers on chromosome 9. Penetrance of geniospasm set at 100%. Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D9S147E — OO -12.05 -6.11 -2.2 -0.76 0.27 0.46 0.26 D9S161 — OO -12.76 -6.82 -2.88 -1.39 -0.25 0.12 0.16 D9S43 — o o -10.06 -4.15 -0.37 0.89 1.54 1.34 0.71 D9S1791 — o o -9.96 -4 .05 -0.26 1.01 1.67 1.46 0.8 D9S200 -10.67 0.35 1.48 2.29 2.43 2.13 1.52 0.72 D9S166 — o o -6.7 -1.8 1.25 2.18 2.45 1.95 1.04 D9S1806 — o o -6.39 -1.5 1.53 2.44 2.65 2.09 1.1 D9S1822 -5.79 -1.28 1.61 3.27 3.6 3.27 2.42 1.23 D9S1837 — o o -0.68 2.22 3.84 4.12 3.67 2.7 1.38 D9S1876 — o o -1.04 1.86 3.48 3.77 3.36 2.45 1.23 D9S175 — o o -7.47 -2.55 0.56 1.56 1.97 1.62 0.86 D9S1834 -33.08 -13.86 -7.45 -2.27 -0.39 0.88 1.03 0.61 D9S153 —- o o -13.57 -6.63 -2.05 -0.36 0.8 0.94 0.57 D9S257 — oo -15.07 -9.01 -3.55 -1.35 0.26 0.64 0.42 91 Table 4.3: Two-point LOD scores between hereditary geniospasm in family 2 and microsatellite markers spanning GSM1 candidate interval. Penetrance set at 80%. Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D9S166 — oo -8.05 -5.01 -2.75 -1.71 -0.70 -0.22 -0.01 D9S1806 — oo -7.18 -4.16 -2.03 -1.13 -0.35 -0.03 0.06 D9S1822 -16.59 -7.31 -4.29 -2.15 -1.23 -0.41 -0.07 0.05 D9S1837 -11.12 -7.62 -4.59 -2.24 -1.25 -0.40 -0.05 0.05 D9S276 -11.27 -3.94 -1.96 -0.68 -0.22 0.11 0.20 0.15 D9S1876 -16.53 -6.91 -3.91 -1.85 -1.00 -0.29 -0.01 0.06 D9S175 — oo -6.45 -3.47 -1.44 -0.64 -0.03 0.15 0.14 92 Figure 4.3: Genetic map of the pericentromeric part of chromosome 9 showing selected microsatellite markers and the GSM1 candidate region. Sex-averaged genetic distances are from the Galton laboratory chromosome 9 map (Povey et al, 1997) and are given in centimorgans. D9S161 3.8 cM D9S43 3.4 cM D9S1791 5.8 cM D9S2Q0 4.1 cM D9S166 0.5 cM D9S1806 0.2 cM D9S1822 0.4 cM GSM1 lo c u s D9S1837/ D9S276 0.2 cM D9S1876 2.1 cM 1.3 cM D9S175 4 cM D9S1834 5.8 cM D9S153 15.5 cM V D2S257 93 Figure 4.4: Chromosome 9q haplotypes in family 1. The disease-associated haplotype is enclosed by boxes. Alleles in parentheses are those for which phase could not be determined with certainty. Dashes are used where the marker could not be typed. o c - T o c - r o c - r o c - T d - i T CM o o to * r o c m c o O IO S N (O - nc c co co co in m o- M -M- CO CM CM CM CO CM CM -M- in CM CO CO CM > o in in CO cm o c o c n i tj - n i o o n n co in in tj - t - co CO CO CO CO tj - o -M-co m CO CO Tj- CO CO I-'. CM -M- in CM CO CO CM Tj- ID CM CO CO CM CM ■M" CM t OO r CM CMin-M-CMCOCOi-COCO CM M i i c c co ■M" co co in in M n t CMCM Tt CO in CO CM CMinn-CMCOCO-r-COCO CM -M- in CM CO CO CMCO^tcOCOr^COCOCM CM-M-inCMCOCOCMCOCM CM CM CO CM CO co Tt ■ M -C OC MC M Tf co cm CO n h- in cm ^ c co co r^- n ■M- in n -M- in n co m in -M- in CO CO CO CO CO CO COCM CMT}- TJ- CM CO CM > CM-M-IOCMCOCOCMCOCM MM nCMCOCOCMCOCM C O C M C O C O C M C in CM'M’ xj- CM CM 00 CM CMO-tnCMCOCOCMCOCM CM CMTj-mCMCOCOCMCOCM ■ CMCO-M-cOCOr^COCOCM CM CM Tf in CO CO in CM h ' - ' M - c o T - t ^ . i n c o i n ' M - o^i' ccc- -in -cococo-M coi^-in'M CM CM -M- CO n CM i CM CM r-'-'M-coi-f'.iocoin'M- m CM CM't-M-CMOOCM'r-COCM CM-M-inCMCOCOCMCOCM CM-M-TtCMOOCM-r-COCM coco ICO CO -M- CO CO CO-M-COCOI^COCOCM. m -M- m . CM CM 00 CM O- cOTj-incMini^i-coco CM - n in in t T - j r co o- o- co - MC - OC MC M ■ M -C OC MC M ■M- CO CO CO CM CM C OC MC OC O O t M - CO h- CO CO CM Tt cm CM CM 00 CO CM CM o co co cm o co co n "S- in CO in CO CO CO CO CO CO CO CO CO CM CO CO CO CO CO CO cm in 4.7 DISCUSSION The data presented in this chapter clearly establish the existence of a locus for hereditary geniospasm on the proximal long arm of chromosome 9 in a large British kindred (family 1). The relatively high LOD score at the marker D9S1837 (Zmax= 5.24 at 0 = 0) is significantly in excess of the Z > 3.0 threshold considered to indicate genetic linkage. Haplotype data allow the GSM1 locus to be assigned to a relatively small candidate region spanning 2.1 cM, between the flanking markers D9S1806 and D9S175, in the cytogenetic region 9ql3-q21. The candidate gene positional cloning strategy presents difficulties in the case of hereditary geniospasm. Little is known of the pathophysiology of the disorder, indeed it is not even clear whether the disorder originates in the central nervous system or in peripheral tissues such as muscle. In the absence of a better understanding of the pathogenesis of geniospasm, selection of appropriate candidates from the genes known to map to the GSM1 locus is problematic. Of the genes assigned to 9ql3-q21 (Povey et al, 1997; and GDB) (table 4.4), none appear to be biologically promising candidates for GSM1. Other neurological disorders which map to the vicinity of the GSM1 locus include: Friedreich’s ataxia, situated proximal to GSM1 at 9ql3, close to marker D9S15 (Duclos et al, 1994; Campuzano et al, 1996); chorea-acanthocytosis which maps 5 cM distal to geniospasm between the markers GATA89al 1 and D9S1843 (Rubio et al, 1997); hereditary sensory neuropathy type I at 9q22.1-q22.3 (Nicholson et al, 1996), and two neurosensory non-syndromic deafness loci, DFNB7 and DFNB11, both at 9ql3-q21 (Jain et al, 1995; Scott et al, 1996). The DFNB11 critical region, between D9S15 and D9S927, overlaps the GSM1 region (Jain et al, 1995). The association of deafness with geniospasm in two families (Grossman, 1957; Alsager et al, 1991) is interesting in this context and raises the possibility of either a contiguous gene syndrome (in which a deletion disrupts two genes in close proximity), or allelic heterogeneity (different mutations in the same gene causing a range of phenotypic manifestations). However, deafness is believed to be due to otosclerosis in geniospasm families and is neurosensory in DFNB7/DFNB11 families. Furthermore, deafness does not segregate with geniospasm in the family reported by Grossman (seeming to originate in the spouse of a parent transmitting 95 geniospasm) and is inherited in a recessive fashion at the DFNB loci, while deafness in the geniospasm families appears dominantly inherited, suggesting that the association may be incidental in these families. The penetrance of hereditary geniospasm is not established with certainty but some authors have suggested that penetrance is almost complete (Goldsmith, 1927; Gordon et al, 1993; Soland et al, 1996). The finding of four asymptomatic gene carriers in family 1 (individuals IV-1, IV-3, IV-7 and IV-16) supports the contention that penetrance may be lower than previously suggested. Thus, the penetrance figure of 80% used in this study may be closer to the true value. However, use of a conservative linkage model in which penetrance is considered to be 100%, results in LOD scores which, while lower for most markers, remain significant for linkage of GSM1 to this region. It has been suggested that hereditary geniospasm may be a forme fruste of other inherited dyskinesias such as focal dystonia or essential tremor with which it is sometimes classified (Danek, 1993). One of the major genetic loci for generalised, childhood onset dystonia, DYT1, maps to chromosome 9q34 (Ozelius et al, 1989; Kramer et al, 1994) and is therefore genetically distinct from hereditary geniospasm (see chapter 5). Focal dystonia has been assigned to a locus on the short arm of chromosome 18 in one kindred with adult onset torticollis (Leube et al, 1996) indicating that geniospasm in family 1, and focal dystonia in the kindred described by Leube et al, are genetically distinct. Loci for essential tremor have recently been identified on chromosomes 2 and 3 (Higgins et al, 1997; Gulcher et al, 1997), thus distinguishing this disorder from geniospasm, at least in the families studied to date. Paroxysmal dystonic choreoathetosis (chapter 3), a disorder causing an episodic mixed movement disorder with no abnormality between attacks, has been mapped to loci on chromosomes 2q and lp in a number of families (Auburger et al, 1996; Fink et al, 1996; Fouad et al, 1996; Jarman et al, 1997; Hofele et al, 1997; Raskind et al, 1998). Both the clear clinical differences between paroxysmal dystonic choreoathetosis and hereditary geniospasm, and the genetic mapping data, indicate that the two conditions are not allelic. However, as focal dystonia, paroxysmal dystonic choreoathetosis, essential tremor and geniospasm (see below) may all be genetically heterogeneous, the mapping data currently available are not sufficient to fully exclude the possibility that geniospasm is allelic to one of these disorders. 96 The finding that a second, unrelated, British family with geniospasm is not linked to markers on the proximal long arm of chromosome 9 demonstrates that hereditary geniospasm is genetically heterogeneous and indicates the existence of at least one other locus capable of producing an identical clinical phenotype. LOD scores for markers in the GSM1 critical region in family 2 are more negative than -2.0 for all markers at the lower recombination fractions. These data combined with the close spacing of markers between D9S1806 and D9S175 are sufficient for exclusion of the geniospasm gene in family 2 from the entire GSM1 interval. The existence of a second geniospasm gene is perhaps surprising given the somewhat unusual and highly focal nature of the condition. The relatively small genetic interval to which the GSM1 locus has been mapped will facilitate attempts to identify the gene using a physical mapping approach as part of a positional cloning strategy. Molecular characterisation of GSM1 would be of biological interest and may provide insights into the pathogenesis of this unusual disorder. 4.8 FUTURE DIRECTIONS OF STUDY As discussed in the preceding section, molecular characterisation of the GSM1 gene is the ultimate aim of a positional cloning project. Genetic linkage analysis such as that described in this chapter is the first step in such an undertaking. The GSM1 candidate region is sufficiently small for physical mapping of the gene to be a realistic possibility. However, further refinement of the genetic localisation of GSM1 may be possible by studying other geniospasm families, some of whom may be linked to the GSM1 locus. Refinement of the GSM1 candidate interval would greatly facilitate future physical mapping of this gene. In addition to improving the mapping of GSM1, analysis of chromosome 9 markers in additional geniospasm families would help to determine the contribution of the GSM1 locus to geniospasm in these families. Additional geniospasm loci could then be mapped in non-GSM 1 families including family 2 from this study. 97 Table 4.4: Genes mapping to 9q 13-21 from the Genome Database (GDB) and Povey et al (1997). Symbol Name Cytogenetic GDB accession location number BTEB1 Basic transcription element 9ql3 GDB:141537 binding protein 1 PRKACG Protein kinase, cAMP- 9ql3 GDB:120719 dependent, catalytic, gamma ANX1 Annexin I (lipocortin I) 9ql2-q21.2 GDB: 120550 ALDH1 Aldehyde dehydrogenase 1, 9q21 GDB:119667 soluble GCNT1 Glucosaminyl (N-acetyl) 9q21 GDB:135991 transferase 1 GCNT2 Glucosaminyl (N-acetyl) 9q21 GDB:215361 transferase 2 ,1-branching enzyme ILRS Isoleucine-tRNA synthase 9q21 GDB:384085 COL15A1 Collagen type XV, alpha 1 9q21-q22 GDB:132578 polypeptide 98 CHAPTER FIVE LINKAGE ANALYSTS IN FAMILIES WITH PRIMARY TORSTON DYSTONIA 5.1 OUTLINE OF CHAPTER This chapter describes genetic linkage studies in three families with non-DYTl primary torsion dystonia (PTD). The chapter begins with a review of the clinical features of the dystonias and summarises the current understanding of the molecular genetic basis of dystonia. The clinical features of the three families studied and the linkage methods employed are described. The results of exclusion mapping of dystonia in these families at recently identified PTD loci (DYT6 and DYT7) are presented. The results of the on going genome-wide search for the dystonia gene in one of these families are given. The significance of these results are discussed in the context of current understanding of the molecular genetics of PTD, and future directions of study suggested. 5.2 AN INTRODUCTION TO PRIMARY TORSION DYSTONIA Dystonia is defined as a syndrome of abnormal involuntary movements characterised by sustained muscle contractions, frequently causing twisting and repetitive movements, or abnormal postures (Fahn et al, 1987). In the absence of an identified cause or other neurological abnormalities, this syndrome has been termed idiopathic torsion dystonia. However, with the mapping of several genes for idiopathic torsion dystonia, the term primary torsion dystonia (PTD) is now preferred (Fahn et al, 1997). The term secondary (or symptomatic) dystonia denotes dystonia occurring in the setting of another neurological disease, for example a degenerative disorder of the CNS or a structural brain lesion, or as a result of environmental exposure to drugs such as tardive dystonia caused by dopamine receptor antagonists. It has been suggested that the term secondary dystonia should be reserved for dystonia resulting from environmental exposure or local CNS lesions (Fahn et al, 1987). Fahn et al (1997) proposed that the classification of dystonia be expanded to classify degenerative disorders causing dystonia (such as Wilson’s disease or Huntington’s disease) under the heading of ‘heredodegenerative 99 dystonia’. These authors also suggested that presumed neurochemical disorders, with neurological features in addition to dystonia (such as dopa-responsive dystonia), be classified as ‘dystonia-plus syndromes’. In this chapter, however, the more traditional aetiological classification into primary (formerly idiopathic) and secondary dystonia will be adhered to. Instead, the primary dystonias will be sub-divided according to the evolving molecular genetic classification. Primary torsion dystonia is not a rare disorder. A study of the prevalence of dystonia in Rochester, Minnesota, using the records of the Mayo Clinic (Nutt et al, 1988), reported a prevalence in this community of 3.4 per 100,000 for generalised dystonia, and 29.4 per 100.000 for focal dystonias. Although the relatively small sample population in this study necessitated wide confidence intervals (0.4-12.4 per 100,000 for generalised dystonia; 17.2-47.9 per 100,000 for focal dystonias) the results are broadly in line with figures reported from the UK: 1.6 per 100,000 for generalised dystonia, and 12 per 100.000 for focal dystonias (Duffey et al, 1997). Because focal dystonia is often mild and asymptomatic, case ascertainment may be incomplete, which means that the figures for focal dystonia could be underestimates. The prevalence of dystonia in the Ashkenazi Jewish population has long been recognised to be particularly high. Early estimates suggested a prevalence of up to 1 per 38,000 among Jews in the USA and indicated that the prevalence among Jews may be up to five times higher than among non-Jews (Zeman and Dyken, 1967; Eldridge, 1970; Zilber et al, 1984). 5.2.1 Clinical features and inheritance of dystonia Dystonia may involve almost any muscle group, with a correspondingly wide range of clinical manifestations, encompassing equinovarus posturing of the foot with toe walking, hyperpronation of the arm, lordosis, scoliosis and torticollis. Writer’s cramp, spasmodic dysphonia, blepharospasm, and oromandibular dystonia (the last two occurring together as Meige’s syndrome), all fall within the rubric of dystonia. Dystonic spasms are often initially intermittent, appearing or intensifying with voluntary movement (which may lead to an erroneous diagnosis of hysteria), but dystonic postures may eventually become fixed. Electrophysiologically, dystonia is characterised by co-contraction of antagonist muscles and overflow of activity into extraneous muscles. 100 Table 5.1: Classification of dystonia by distribution Focal dystonia Single body part affected, e.g. neck (torticollis) Segmental dystonia Two or more contiguous body parts affected, e.g. one leg and trunk (segmental crural dystonia) Generalised dystonia Segmental crural plus any other body part Multifocal dystonia Two or more non-contiguous areas affected Hemidystonia Ipsilateral leg and arm affected 101 5.2.1.1 Primary torsion dystonia PTD is clinically heterogeneous, and may be classified according to age at onset, body part first affected, and distribution (table 5.1). Severity is largely determined by age at onset (Marsden and Harrison, 1974). The age at onset distribution of dystonia is bimodal with peaks at 9 and 45 years. There are clear clinical differences between patients with early-onset (<20 years) and late-onset (>20 years) PTD. Early-onset PTD usually begins in a limb, particularly the leg, and frequently progresses to generalised dystonia. Muscles of the head and neck are frequently spared. In contrast, late-onset PTD typically begins in the neck or head, for example with torticollis, and tends to remain focal in distribution. Segmental dystonia is of intermediate severity and may arise at any age (Marsden and Harrison, 1974; Greene et al, 1995). Early, limb-onset PTD is particularly prevalent amongst the Ashkenazi Jewish population where inheritance is autosomal dominant with approximately 30% penetrance (Bressman et al, 1989). In the UK, generalised, multifocal and segmental dystonia are estimated to be dominantly inherited in 85% of cases, with approximately 40% penetrance (Fletcher et al, 1990b). The remaining 15% of cases are thought to be non-genetic phenocopies. The genetic contribution to adult-onset focal dystonia is less clear. Most cases are apparently sporadic, but careful examination of first degree relatives of patients with focal dystonia reveal that between 8% and 25% have an affected, often asymptomatic, relative (Waddy et al, 1991; Bressman et al, 1995, Leube et al, 1997b). These family studies suggest the existence of one or more autosomal dominant genes with low penetrance. A small number of large families exist in which focal dystonia is inherited as a dominant trait with relatively high penetrance (Uitti and Maraganore, 1993; Bressman et al, 1996; Leube et al, 1996). It seems likely, however, that non-genetic causes account for a proportion of cases. 5.2.1.2 Secondary dystonias Dystonia may occur as a symptom of numerous diseases affecting the central nervous system, accounting for about one in three cases of dystonia (Fahn et al, 1987). This is a highly 102 Table 5.2: Selected genetic causes of secondary dystonia. DISEASE ADDITIONAL CLINICAL INHERI GENE/ LOCALISATION/ FEATURES TANCE METABOLIC DEFECT Wilson’s disease Psychiatric, hepatic, occular AR Copper-transporting ATPase manifestations. Parkinsonism, BG gene lucencies on CT Huntington’s disease Dementia, chorea AD HD gene (CAG expansion) Hallervorden-Spatz Progressive dystonia. Parkinsonism, AR 20pl2.3-pl3 (Taylor etal, 1996) disease cognitive impairment, retinopathy and optic atrophy may occur Neuroacanthocytosis Chorea, neuropathy, acanthocytosis AR Chorea-acanthocytosis: 9q21 (Rubio et al, 1997); abetalipoproteinaemia: triglyceride-transfer-protein gene (Narcisi et al, 1995) Machado-Joseph Cerebellar ataxia, spasticity, peripheral AD SCA3 gene (CAG expansion) disease neuropathy Niemann-Pick Supranuclear gaze palsy, ataxia, AR NPC1 gene (Carstea et al, 1997) disease, type C dementia, hepatosplenomegaly, foamy cells in marrow, sea-blue histiocytes on skin biopsy Ataxia-telangectasia Telangectasia, ataxia, occulomotor AR ATM gene apraxia, neuropathy, malignancies Late-onset neuronal Retinopathy seizures, dementia, ataxia, AR CLN3 gene (The International ceroid lipofuscinosis myoclonus Batten disease consortium, 1995) Leigh’s syndrome Vomiting, subacute brainstem syndrome, AR Pyruvate dehydrogenase deficiency lactic acidosis, BG lucencies on CT X-linked Pyruvate carboxylase deficiency Maternal Mitochondrial DNA mutations (DiMauro and De Vivo, 1996) Juvenile/ adult Dementia, psychiatric disorder, seizures, AR Arylsulphatase A deficiency metachromatic white matter dystrophy leucodystrophy Late-onset GMj and Dementia, ataxia, amyotrophy AR P-galactosidase deficiency GM2 gangliosidoses Hexosaminidase deficiency Lesch-Nyhan syndrome Mental retardation, spasticity, self- X-linked Hypoxanthine-guanine mutilation phosphoribosyl transferase deficiency Organic acidaemias Acidosis, episodic ataxia and AR Propionyl CoA carboxylase, encephalopathy, Methylmalonyl CoA mutase, neutropaenia/thrombocytopaenia, BG Glutaryl CoA dehydrogenase lucencies on CT deficiencies Mohr-Tranebj aerg deafness, cortical blindness, mental X-linked DDP gene (deafness/dystonia syndrome retardation peptide) (Jin et al, 1996) Leber’s hereditary Sub-acute visual loss, BG lucencies on Maternal Mitochondrial DNA mutations optic atrophy - CT dystonia BG: Basal ganglia 103 heterogeneous group of disorders, many of which have a genetic basis (table 5.2). Structural lesions of the basal ganglia, athetoid cerebral palsy, neuroleptic drug exposure and numerous metabolic, toxic and infective disorders may all cause dystonia (Fahn et al, 1987). In most cases the clinical presentation is not dominated by dystonia; other neurological features, such as cognitive impairment, seizures, visual loss, or pyramidal lesions, may be present. Occasionally, however, dystonia may be the sole clinical manifestation, although there are often clinical clues indicating another underlying disease. Hemidystonia is almost always due to a lesion of the contralateral basal ganglia or its connections. Rapid progression of symptoms, onset with dystonia at rest, or the presence of other neurological signs, all suggest secondary dystonia. 5.2.2 The molecular genetic basis of dystonia Over the last decade there have been considerable advances in the understanding of the genetics of dystonia. These have been driven by the need to take a positional cloning approach to understanding pathophysiology, as anatomical and neurochemical studies in PTD are generally unrevealing. A molecular genetic classification of the dystonias is now evolving, which complements the more traditional clinical classification, and allows phenotype-genotype correlations to be made (table 5.3). The Human Genome Organisation Nomenclature Committee have assigned the gene symbol ‘DYT’ to the primary dystonias, and there are currently seven loci: DYT 1-7. Contrary to established practice some clinically or genetically distinct dystonia syndromes have been ascribed a DYT designation without an identified genetic locus - for example DYT2 and DYT4 (see below). Recently it has been proposed that this classification be extended to 12 DYT designations (DYT 1-12) along clinical and genetic lines (Muller et al, 1998). However, this classification includes disorders such as the paroxysmal dystonias, and rapid-onset dystonia-parkinsonism, which are usually classified separately, and some of which do not have identified genetic loci. This scheme will not be used here in the interests of consistency. Each of the seven DYT categories is associated with a characteristic, but sometimes overlapping, spectrum of phenotypic manifestations. Some loci are of considerable clinical importance (such as DYT1 and DRD), while the contribution of others such as DYT4 , DYT 6, and DYT7, has not yet been fully evaluated. 104 Table 5.3: Genetic loci for the primary dystonias Symbol Disease Inheritance Gene/linkage DYT1 PTD- Usually early, limb-onset with progression. AD (30-40% torsinA Commoner in Jewish patients. penetrance) 3 bp deletion DYT2 Autosomal recessive PTD (Spanish Gypsies) AR — DYT3 Philippino dystonia-parkinsonism X-linked Xql3.1 DYT4 PTD with laryngeal involvement AD — DRD (DYT5) L-dopa-responsive dystonia AD GCH1 AR TH DYT6 Mixed phenotype PTD (Mennonites) AD 8p21-q22 DYT7 Late-onset, focal PTD AD 18p Key: PTD: Primary torsion dystonia AR: Autosomal recessive AD: Autosomal dominant GCH1: GTP cyclohydrolase 1 gene TH: Tyrosine hydroxylase gene 105 5.2.2.1 DYT1 Linkage studies in Jewish and non-Jewish families with PTD led to the identification of the DYT1 locus on chromosome 9q34 (Ozelius et al, 1989; Kramer et al, 1990). Strong linkage disequilibrium has been demonstrated between DYT1 and a 9q34 haplotype in both familial and sporadic Jewish cases, indicating a founder mutation estimated to have occurred among the Ashkenazim of Eastern Europe about 350 years ago (Ozelius et al, 1992; Risch et al, 1995). The DYT1 gene is enriched in this population as a result of genetic drift, with a gene frequency estimated to be as high as 1/2000 (Risch et al, 1995). Very recently the DYT1 gene was cloned and a unique 3 bp (GAG - glutamic acid) deletion identified in all chromosome 9q34-linked families, regardless of ethnic background and surrounding haplotype (Ozelius et al, 1997). The deletion results in loss of one of a pair of glutamic acid residues in a novel protein named torsinA. TorsinA is an ATP-binding protein with some similarity to the family of heat-shock proteins; its function in the nervous system and role in the pathogenesis of dystonia are not yet understood. DYT1 messenger RNA is expressed at very high levels in the dopaminergic cells of the substantia nigra pars compacta, a particularly interesting finding given the existing evidence implicating dopaminergic dysfunction in the pathogenesis of dystonia (Augood et al, 1998). The common 3 bp deletion appears to have arisen independently in different ethnic groups (Ozelius et al, 1997; Klein et al, 1998a), and to date no other mutations have been identified in the gene. A small proportion of patients from families too small for linkage analysis, but with the typical DYT1 phenotype, do not carry the deletion (Ozelius et al, 1997). It is not yet clear if these patients have novel mutations in torsinA, or whether other genes underlie dystonia in such cases. Recurrence of the same disease-associated mutation suggests either that the GAG deletion occurs at a mutational ‘hotspot’, or alternatively that only a single variation in the protein can give rise to the dystonia phenotype. Although detailed phenotype-genotype studies are still awaited, it is already clear that the DYT1 mutation produces a relatively homogeneous clinical phenotype in both Jews and non-Jews. Typically onset is early, usually in a limb, and there is spread to at least one other limb but rarely to cranio-cervical muscles (Bressman et al, 1994a; Ozelius et al, 1997). Patients with dystonia who carry the Jewish chromosome 9q34 haplotype (D Y Tl^), have a mean age at onset of 12.5 years, and 70% eventually develop 106 generalised or multifocal dystonia. However, a small proportion of patients with DYT1aj have a less severe phenotype, with later onset (up to the age of 44), and little or no progression of symptoms after long follow-up (Bressman et al, 1994a). In Europe, DYT1 mutations have been estimated to account for approximately 55% of families with generalised or segmental dystonia, although the confidence intervals on which this figure is based are wide (Warner et al, 1993). Mutation analysis of DYT 1 in UK patients is currently underway at the Institute of Neurology. It has previously been suggested that DYT1 mutations may be a common cause of late- onset focal dystonia. However, there is mounting evidence that early and late-onset PTD are genetically, as well as clinically, distinct. Amongst the Ashkenazim, D Y T l^ is rarely found in patients with adult-onset or purely focal dystonia (Bressman et al, 1994a; Gasser et al, 1996), and exclusion of the DYT1 locus in large non-Jewish families with dominantly inherited late-onset torticollis also supports this view (Bressman et al, 1996). In the study of Ozelius et al (1997), none of the probands from 76 families with focal or segmental dystonia affecting cranio-cervical muscles were found to have the 3 bp deletion in torsinA. However, DYT1 mutations may occasionally result in a mixed dystonia phenotype with involvement of both cranio-cervical and limb muscles. The GAG deletion has been found in approximately 10% of probands from families with a mixed dystonia phenotype (Ozelius et al, 1997). 5.2.2.2DYT2 Autosomal recessive PTD. The existence of autosomal recessive forms of PTD is controversial. Recessive inheritance of PTD has been proposed in a small number of consanguineous Spanish Gypsy kindreds (Gimenez-Roldan et al, 1988). However, as the parents of the affected probands were not clinically examined by the authors, this report must be considered unconfirmed. There is no clear evidence supporting autosomal recessive inheritance of PTD outside this population (Fletcher, 1990a). Many cases of PTD which were originally believed to be inherited as autosomal recessive traits have now been shown to be inherited in a dominant fashion with reduced penetrance. 107 5.2.2.3DYT3 The syndrome of dystonia-parkinsonism (Lubag) is an X-linked, neurodegenerative disorder confined to the Philippines. Onset is in early adult life with focal dystonia which subsequently becomes generalised and is often accompanied by parkinsonism unresponsive to L-dopa. Allelic association with markers at Xq 13.1 indicates the existence of a founder mutation in this population, thought to have originated on the island of Panay (Muller et al, 1994; Haberhausen et al, 1995). 5.2.2.4 DYT4 A large Australian family with dystonia with prominent laryngeal involvement has been assigned the gene symbol DYT4 after the demonstration that this dystonia phenotype does not map to chromosome 9q34 (Parker, 1985; Ahmad et al, 1993). Although laryngeal dystonia is not uncommon among patients with sporadic late-onset dystonia, the phenotype of dominantly inherited dystonia with onset in laryngeal muscles appears, to date, to be unique to this family. The DYT4 phenotype is described in section 5.4.3. The gene has not yet been assigned a genetic map position. 5.2.2.5 DRD (formerly DYT5) Dopa-responsive dystonia has been assigned the symbol DRD, and is discussed in chapter 6. 5.2.2.6DYT6 A locus for a mixed dystonia phenotype has recently been mapped to a 40 cM interval, spanning the peri-centromeric region of chromosome 8, in two German-American Mennonite families (Almasy et al, 1997). Although some family members have a phenotype similar to the DYT1 phenotype, others have only segmental or focal cranio- cervical dystonia after long follow-up. Spread to cranio-cervical muscles occurs in most patients and is the major cause of disability in many. The wide age-at-onset distribution (2-69 years), and overall tendency to involve limb and cranio-cervical muscles equally, are claimed as distinguishing features of the DYT6 phenotype. Both families share an identical haplotype over the 40 cM candidate interval, which suggests a common founder mutation despite the absence of known relationship between these families. 108 5.2.2.7DYT7 Only one locus for late-onset, purely focal dystonia has been identified to date. Leube et al (1996) described a German kindred with purely focal dystonia showing autosomal dominant inheritance and reported linkage to markers spanning an approximately 30 cM interval at the telomere of chromosome 18p (Leube et al, 1996). Dystonia in this kindred was largely confined to the cervical muscles resulting in torticollis in most individuals. Similarly affected, apparently sporadic cases of focal dystonia in Northern Germany were found to share a common haplotype over a 6 cM region of 18p (Auburger et al, 1997; Leube et al, 1997a), suggesting that dominant inheritance at low penetrance of a founder mutation in the DYT7 gene may be an important cause of focal PTD among a relatively isolated population of Northwest Germany. However, these findings have not been confirmed in another study of focal dystonia in Northern Germany and must therefore be interpreted with caution (Klein et al 1998b). The contribution of this locus to focal dystonia in other populations remains to be determined. The existence of a gene for focal dystonia on chromosome 18p is supported by the finding that patients with cytogenetic abnormalities of the terminal short arm of chromosome 18 (deletions or balanced translocations) have focal or segmental dystonia as part of a broader syndrome including mental retardation, dysmorphic appearance and short stature (Page et al, 1998). 5.3 AIMS OF THE STUDY The aims of the present study were twofold: firstly to evaluate the contribution of the DYT6 and DYT7 loci to families with non-DYTl PTD, and secondly to attempt to identify the DYT4 locus by means of a genome-wide linkage search. 5.3.1 Does dystonia in families with non-DYTl PTD map to the DYT6 or DYT7 loci? Until the DYT6 and DYT7 genes are cloned, evaluation of the role of these loci in PTD will depend upon allelic association studies in patients with apparently sporadic PTD such as that of Leube and co-workers (1997a), and linkage analysis in suitable families. However, large families with non-DYTl PTD suitable for linkage analysis are 109 comparatively rare. Only six such families have been reported (Parker, 1985; Uitti and Maraganore, 1993; Bressman et al, 1994b; Holmgren et al, 1995; Bressman et al, 1996; Bentivoglio et al, 1997), excluding the three families in which the DYT6 and DYT7 loci were mapped, although other unreported families may exist. Five of these non-DYTl families were available for further study. All had juvenile or adult-onset dystonia with a predominantly cranio-cervical, or a mixed, PTD phenotype (section 5.4); and in all the DYT1 locus had been previously excluded by means of linkage analysis (Ahmad et al, 1993; Bressman et al, 1994b; Bressman et al, 1996; Bentivoglio et al, 1997). Thus, in order to determine whether any of these families were linked to either locus linkage analysis was performed using microsatellite markers spanning the DYT6 and DYT7 loci. For two of these families (Bressman et al, 1994b and 1996), linkage analysis was performed by Dr T.G. Nygaard, Columbia University, New York, USA, and these results are therefore not discussed in this thesis. 5.3.2 Genetic mapping of the DYT4 locus An Australian family with PTD characterised by spasmodic dysphonia have been assigned the gene symbol DYT4 despite the absence of a genetic map locus for the disease gene (sections 5.2.2.4 and 5.4.3). The DYT1 locus has been previously excluded as the cause of dystonia in this family (Ahmad et al, 1993). The family was one of three families in which linkage to the DYT6 and DYT7 loci was studied as described in the previous section (5.3.1). In order to attempt to map the DYT4 locus, a genome-wide search for genetic linkage was undertaken using a panel of microsatellite markers. 5.4 PTD FAMILIES AND METHODS The clinical features of the three families studied have been described in detail previously (Parker, 1985; Uitti and Maraganore, 1993; Bentivoglio et al, 1997). Selected clinical details of the families studied and of the DYT6 and DYT7 families are presented in table 5.4. None of the families have any apparent Ashkenazi Jewish heritage. All patients were examined on-site by two neurologists (family 1: Drs Uitti and Maraganore; family 2: Drs Bentivoglio and Cassetta; family 3: Drs Waddy and Oley), and the examinations were videotaped. Videotapes were reviewed by a senior neurologist, experienced in the field of movement disorders (Professor AE Harding or Professor CD Marsden). A diagnosis of definite dystonia was made where unequivocal 110 sustained muscle contraction resulting in twisting or repetitive movements was witnessed (Fahn et al, 1987). Writer’s cramp was tested for in all patients. A diagnosis of definite or possible dystonia was made in deceased family members on the basis of history obtained from family members interviewed. A diagnosis of probable dystonia was made where a rapid jerky dystonic tremor was observed, when clinical features were subtle, or when not all examiners agreed on a diagnosis of definite dystonia. Causes of secondary dystonia were excluded with appropriate investigations. Exclusion of the DYT1 locus as a cause of dystonia in these families had been previously established by exclusion mapping using markers at chromosome 9q34 (Ahmad et al, 1993; Bressman et al, 1996; Bentivoglio et al, 1997). 5.4.1 Family 1 Family 1 is an American family of German origin first described by Uitti and Maraganore (1993). The pedigree is illustrated in figure 5.1. Of the 22 family members examined, five had definite cervical dystonia manifest as torticollis, retrocollis or anterocollis. The family included a pair of monozygotic twins who were considered as a single individual for the purposes of linkage analysis. Five family members had possible cervical dystonia including two obligate gene carriers. The two obligate gene carriers (individuals II-4 and 11-13) were considered to be definitely affected for the purposes of linkage analysis. The onset of dystonia was during adult life; mean age at onset was 35 years of age (range 18-49 years). After an average duration of symptoms of 19 years (range 7-38 years) dystonia remained focal in all affected, and possibly affected, family members. 5.4.2 Family 2 Family 2 is an Italian family, first described by Bentivoglio et al (1997). The pedigree is illustrated in figure 5.2. Among the 45 family members examined, eight had definite dystonia and six had probable dystonia. Three deceased family members had definite dystonia by history. Symptoms began in the cranio-cervical region in six of the definitely affected cases, and in the arms in the remaining two. After a mean disease duration of 41 years (range 5-66 years) symptoms have remained focal in one patient, and progressed to involve the cranio-cervical muscles, the arms or the trunk in the 111 others. Of the patients who showed progression, dystonia was segmental in four patients, and mild-generalised in two at the time of examination. Among the probable cases, progression of symptoms was less evident, with a focal or segmental distribution of dystonia in all cases. Of the 14 living definitely and probably affected family members, 12 had prominent involvement of cranio-cervical muscles on examination. 5.4.3 Family 3 Family 3 is an Australian family of English origin (Parker, 1985; Ahmad et al, 1993). The disease gene in this family has been assigned the symbol DYT4. The pedigree is illustrated in figure 5.3. At the latest assessment 11 family members were definitely affected by dystonia among the 39 examined. Dystonia began with spasmodic dysphonia in nine patients and with torticollis in two patients, both of whom later developed spasmodic dysphonia. Dystonia progressed in nine patients with definite dystonia, eventually becoming generalised in eight, and segmental in one. Two siblings have Wilson’s disease, both had dystonia starting with spasmodic dysphonia and were clinically indistinguishable from other family members affected by dystonia. However, as Wilson’s disease is a cause of secondary dystonia, it is not certain whether dystonia in these individuals is due to the DYT4 gene or the Wilson’s disease gene. The clinical similarity of the dystonia phenotype in these siblings to that observed in other family members favours DYT4 as the cause of dystonia. Moreover, dystonia has continued to progress in these individuals despite treatment with penicillamine. Exclusion of a gene for dystonia in the immediate vicinity of the Wilson’s disease locus in this family using chromosome 13 markers had previously been established (Ahmad et al, 1993). Age of onset of symptoms ranged from 13 to 37 years. 5.4.4 DNA analysis For family 1, DNA samples from 23 family members were genotyped for all markers. For family 2, DNA from 47 family members was available for study. A core set of 36 most informative individuals from family 2 were genotyped for all markers. For family 3, 43 DNA samples were available for study. A core set of DNA samples from 18 individuals were used for the initial genome search. For areas with a high LOD score in the genome screen, three further samples were genotyped in addition to the core 112 Figure 5.1: Primary torsion dystonia* family 1. Family members who had been examined and for whom DNA was available are indicated by asterisks. The index case is indicated by an arrow. Patients with possible dystonia are indicated by hatched symbols. Diamond shaped symbols indicate clinically unaffected siblings (number of siblings indicated by the number in the centre of the symbol). ------* o o o o O o o w 0 0 TO CO Figure 5.2: Primary torsion dystonia - family 2. Family members who had been examined, and for whom DNA was available, are indicated by asterisks. Individuals for whom DNA was available, but not included in the linkage study are indicated by asterisks in parentheses. Patients with possible dystonia are indicated by hatched symbols. > Figure 5.3: Primary torsion dystonia - family 3. The 18 members of the core pedigree used in the genome search are indicated by single asterisks (not in parentheses). The 36 family members typed for markers at the DYT6 and DYT7 loci include the core pedigree and 18 additional family members indicated by double asterisks. 2 X> x: > 2 T3 £ Q JZ C - T3 •6 .22 "O TD T3 XI x: H 3 £ o "t- * pa ■a o >» o t La . 2 at at 2 . La u u •** — (N < La s u a L 3 3 2 2 . - N-/ N-/ O NO NO O r~~ 3 8 - C'' f- «n H O C c/t*C S 5 •S a O ° XJO e 0> cJ a> 8 8 5.4.5 Microsatellite markers analysed The candidate regions containing the DYT6 and DYT7 loci are large (30-40 cM each). To fully exclude either region for each family required a marker density which varied with the size of the family and the number of recombinations with affected individuals in each region. For the DYT6 region, microsatellite markers between markers D8S1791 (8p21) and D8S1475 (8q22) were analysed. For the DYT7 candidate region, markers from 18pter to D18S843 were analysed. These markers are the flanking markers for each candidate region. Microsatellite markers were from the Genethon, CHLC and Utah maps (section 2.4.2). The number and average interlocus spacing of markers typed for each family are given in table 5.5. All map distances are according to the Marshfield Clinic integrated genetic maps at: http://www.marshmed.org/genetics/. Table 5.5: Summary of microsatellite maker numbers and density at DYT6/DYT7 loci. Fam ily Family 1 Family 2 Family 3 Number of markers typed: chromosome 8 14 10 10 chromosome 18 7 12 12 Average marker spacing (cM): chromosome 8 3.4 4.9 4.9 chromosome 18 7.8 4.2 4.2 117 5.4.6 Linkage analysis 5.4.6.1 Exclusion mapping o f the DYT6 and DYT7 loci In each family, pairwise LOD scores were calculated for dystonia and markers in the candidate regions using the FASTLINK 3.OP version of the MLINK linkage program (section 2.4.8). Multipoint analysis was performed using the VITESSE program (O'Connell and Weeks, 1995). Multipoint and pairwise analyses were performed with the assistance of Dr T.G. Nygaard, Columbia University, New York, USA. An autosomal dominant model of inheritance was used with a disease gene frequency of 0.0001. Only definitely affected family members were classified as ‘affected’, while unaffected and possibly affected family members were classified as phenotype ‘unknown’. This conservative, ‘affecteds only’, methodology was used for all analyses to avoid bias from the inclusion of possibly affected individuals or incorrect estimation of penetrance or age of onset. Obligate gene carriers were considered to be definitely affected for the purposes of linkage analysis. The affected identical twins in family 1 were considered as a single person for the purposes of linkage. Marker allele scoring was uniform across all families. 5.4.6.2 Mapping o f the DYT4 locus Pairwise LOD scores for dystonia and markers analysed in the genome search in family 3 were calculated using version 5.1 of the MLINK program (section 2.4.8). All calculations were performed twice; once with the Wilson’s disease patients defined as affected by dystonia, and again assigned as status unknown. An autosomal dominant model of inheritance was used with a disease allele frequency of 0.0001. Age-related penetrance values derived previously for this family were used: penetrance of 18, 55, 73, 91 and 100% by the ages of 20, 25, 30, 35 and 70 years respectively (Ahmad et al, 1993). 5.5 RESULTS OF LINKAGE ANALYSIS AT DYT6 AND DYT7 LOCI The results of pairwise LOD scores between dystonia and markers at the DYT6 (chromosome 8) and DYT7 (chromosome 18) loci are presented in appendix 3. These results show no evidence of linkage between dystonia and either locus in any of the three families studied. Recombinations are present between dystonia and all informative 118 markers in each family. In family 1, several markers produced small positive LOD scores in the DYT6 candidate region (D8S374: Zmax= 0.77 at 0 = 0; D8S543: Zmax = 0.71 at 0 = 0; D8S285: Zmax= 0.65 at 0 = 0; D8S1113: Zmax= 0.56 at 0 = 0 and D8S507 Zmax = 0-39 at 0 = 0 ). However, multipoint analysis indicates the presence of a recombinant affected individual across the entire span of the DYT6 locus in family 1. Formal exclusion of a disease gene from a genetic interval requires LOD scores ^ -2.0 across the entire interval. This was achieved by the use of multipoint linkage analysis. Multipoint analyses for all three families are presented in figures 5.4-5.9. These are based on overlapping 4-point analyses for most intervals, with 5-point and 6-point analyses employed where multipoint LOD scores were > -2.0. Only informative markers are considered for the purposes of multipoint analyses. These allow formal exclusion of dystonia from the entire span of the DYT6 and DYT7 loci for all families except for portions of the DYT6 locus in family 1. Multipoint LOD scores in family 1 rise to a maximum of -0.76 in the interval between D8S1797 and D8S286. Maximum multipoint scores rise to -1.2 to -1.5 at the midpoints of several other intervals in this family. 5.6 RESULTS OF GENOME-WIDE SEARCH FOR THE DYT4 LOCUS 137 microsatellite markers were typed in the genome-wide search for the DYT4 locus with an average spacing of 27.5 cM between markers. Two-point LOD scores are presented in Appendix 4. LOD scores for most informative markers are significantly negative, thus excluding the DYT4 locus from the immediate vicinity of these markers. The most positive LOD scores are for markers on chromosome 19. Twenty-one family members were genotyped for most chromosome 19 markers to increase the maximum LOD score obtainable. The maximum LOD score obtained is Z = 2.66 at 0 = 0 for D19S225, with individuals V-24 and V-25 with Wilson’s disease (WD) designated as at risk. This falls to Z = 0.53 at 0 = 0.2 with these individuals designated as affected. Since it is possible that only one of the siblings with WD has inherited the DYT4 gene in this family, a third LOD score calculation was performed for chromosome 19 markers, in which both WD siblings were excluded from the analysis. Under these conditions Zmax = 2.46 at 0 = 0 for D19S225. Additional chromosome 19 markers flanking D19S225 were therefore studied. LOD scores are negative at 0 = 0 - 0.1 for the three markers telomeric to D19S225 as there are obligate recombinations in individuals V-11,V-13 and VI-1. 119 Figure 5.4: Multipoint LOD score graph for selected markers at the DYT6 locus in family 1. Based on overlapping 4-point analyses using informative markers. o o VO o' o CO Figure 5.5: Multipoint LOD score graph for selected markers at the DYT7 locus in family 1. Based on overlapping 4-point analyses using informative markers. o o LO O o CO d o CD d o o o © CO IO CD - - -- co -- o (0 00 CO CO CO 00 CM IO oo IO CN - T <0 Q 0 0 CO CO oo cn IO CN Figure 5.6: Multipoint LOD score graph for selected markers at the DYT6 locus in family 2. Based on overlapping 4-point analyses using informative markers. o o co o o '<• d o CN o o o c o o o o o d o W o 00 o p to o t a co CN o r» t a CO O I CO O I IO t a CO CN CO CO CO CO o c Q IO o c CO !=» o c CN co cn CO IO Figure 5.7: Multipoint LOD score graph for selected markers at the DYT7 locus in family 2. Based on overlapping 4-point analyses using informative markers. O O io d O o CD d CM o o O o CO d o o 2 o o V CO ID CD 2 CO . L 0) 00 IO CM oo CO 00 00 IO in CM oo co co oo CO CO co oo cn IO CM CO Figure 5.8: Multipoint LOD score graph for selected markers at the DYT6 locus in family 3. Based on overlapping 4-point analyses using informative markers. Multipoint LOD scores were calculated with the Wilson's disease siblings coded as affected and phenotype unknown. o to co o CM o o d o o o o o o o o CM o 2 o ° —1 J ° _ o w cd o o cd o cd o o CO o CO 00 CO o o> O I 00 r- r*- oo 10 co 0 0 to 0> oo CO CM IO o o 00 CM D18S59 D18S1368 D18S1098 I D18S52 D18S452 D18S843 D18S54 Figure 5.10: Chromosome 19 haplotypes in selected members of family 3. The haplotype segregating among several affected members of the family is enclosed by a box. Where phase cannot be determined, alleles are in parentheses. Identity of family members is denoted by pedigree numbers from the main pedigree in figure 5.3. The siblings with Wilson's disease are indicated by hatched symbols. t T S f T T on N M T T N t ' O n O IcPCNJi-CNJ^Tl-LO^r i-incNii-cjT-mco T-rJ-Tfri-CNJCM-^CNJ n on on on O n O n i-CMCMCM-M-^CMCM 'M-lOCOCMCM-'fri-LO o n - r in co ^ t iLOCM'- -OCO C i-lO M 'i-C M C O -i-L CO CO CM ^ ■ n o o ■m- lo ^ lo - i in n ^ m - - CM CM T- ■M- -M- ID -sf ICM 1— LO CM •'tl ■'t ■'t M O M M i "M- in CM CO CM -M- CM CM T- -M- -M- LO -M- LO CM n CM in M ^ -— M M CM -I—■M- CM ^ CM Cvi t o l o c CM N ^- N C CNJ CNI■Nf - CM ^1- i LO t t - - M -3- CM - i M - - n M- CM in M- M- t 1 1 CM CO CO- CO M I i CO CM in -I- - CM - - t o i D M i -M- -M- in ID C\J CM CNJ - 1 t 'M- ) CD U) - o i 'M- ■M"CO M C L , LO LO Tj- Tf -M- , CO LO CM t i O CM - t — OM 't O L lO'M’ (N vo LOD scores are positive at all recombination fractions to 0 = 0.4 for the adjacent markers D19S422 and D19S417, and for the marker D19S414. No LOD scores are > 3.0, the critical threshold for accepting linkage. Haplotypes were constructed manually for chromosome 19 markers and these are illustrated in figure 5.10. There remain 66 gaps of > 20 cM in which no markers have yet been analysed. Unfortunately, during the course of this study, the DNA stocks available for several of the affected members of the pedigree were exhausted or virtually exhausted. The family live in a remote part of Queensland, Australia, and it has not yet been possible for local clinicians to obtain fresh DNA samples to complete the genome search. 5.7 DISCUSSION 5.7.1 Exclusion of the DYT6 and DYT7 loci in three PTD families Three large unrelated families with non-DYTl dystonia were studied for evidence of linkage to the DYT6 or DYT7 genetic loci. The data presented do not support linkage of dystonia to either genetic locus in any family. Formal exclusion of a disease from a genetic locus, however, requires the stringent criterion of Z = ^ -2.0 for the entire interval. Thus, although the pairwise LOD score at 0 = 0 meets this criterion for most informative markers studied, the area of exclusion surrounding each marker is typically relatively small, as Z is less negative than -2.0 at 0 = 0.05 (equivalent to approximately 5 cM) for most markers, while the candidate intervals considered are large. Multipoint linkage analysis is an efficient means of extracting additional quantitative information from a dataset of adjacent markers. The technique is particularly useful for disease exclusion mapping, to extend the genetic distance over which formal exclusion is achieved. Multipoint analysis allows formal exclusion of the DYT6 and DYT7 loci in families 2 and 3, and exclusion of the DYT7 locus in family 1. Exclusion of the DYT6 locus in family 1 is complete for parts of the genetic interval, but falls short of the threshold for exclusion in several areas with a maximum multipoint LOD score of -0.76 in the interval between D8S1797 and D8S286. This inconclusive result is a consequence of the presence of only a single recombinant affected individual across the region in this family, which has fewer members than either family 2 or family 3. Although not 127 formally excluded, the negative multipoint LOD scores across the interval make it unlikely that dystonia in family 1 is linked to the DYT6 locus. The affecteds-only method of analysis employed in this study is a conservative methodology which reduces bias from inclusion of possibly affected individuals and eliminates the need to make estimates of penetrance. However, the technique also reduces the informativeness of a pedigree by excluding at-risk individuals, thus tending to make some LOD scores less negative. Furthermore, the use of the Marshfield Clinic integrated genetic map (at: http://www.marshmed.org/genetics/) rather than the Genethon map, inflates map distances between markers and makes exclusion of the entire interval between markers more difficult. Analysis of the dataset for family 3 excluding the siblings with WD (V-24, V-25) resulted in pairwise LOD scores not significantly different from those calculated with these individuals designated as affected (appendix 3). The entire span of both loci are excluded for family 3 even after exclusion of these siblings (data not shown). The three dystonia families included in this study had important phenotypic similarities to both the DYT6 and DYT7 kindreds (table 5.4). Considered individually each family conforms broadly to either the DYT6 mixed phenotype (families 2 and 3), or the DYT7 focal dystonia phenotype (family 1). Considering all three families and the DYT6 and DYT7 families together, there are features in common between all families, including a juvenile or adult onset of symptoms in most individuals, and prominent involvement of cranio-cervical muscles (20 out of 22 or 91% in the original DYT6 and DYT7 kindreds, and 22 out of 24 or 92% in the three PTD families). These clinical features are in contrast to the childhood onset and cranio-cervical sparing seen in DYT1 dystonia. The finding that dystonia in the three families studied is genetically distinct from the DYT6 and DYT7 loci indicates the existence of at least one other as yet unmapped gene (and possibly as many as three separate genes) for PTD. Thus, the phenotypes of adult- onset focal dystonia and mixed phenotype dystonia are genetically heterogeneous, and may result from mutations of as yet unidentified gene(s) as well as from DYT6 and DYT7 mutations. Such a finding is in contrast to the apparent genetic homogeneity of the ‘DYT1 phenotype’ of early, limb-onset dystonia, the great majority of which is accounted for by mutations of torsinA (Bressman et al, 1994a; Ozelius et al, 1997). 128 The contribution of the DYT6 and DYT7 loci to families and singleton patients with focal and mixed phenotype dystonia remains to be fully evaluated. The presence of allelic association at D18S1098 in patients with torticollis in a rural region of Northwest Germany suggests that this locus is an important cause of dystonia, at least in this community. The Mennonite community in which the DYT6 locus was described is known to be genetically isolated, and it may be that this locus is not of widespread importance outside the Mennonite community. 5.7.2 Genome-wide search for the DYT4 locus After analysis of 137 microsatellite markers in family 3, approximately one third of the genome has been excluded as the site of the DYT4 locus, but definitive linkage has not yet been established. The positive LOD scores obtained for some markers on chromosome 19 are of interest. However, the maximum LOD score does not exceed 3.0, the minimum threshold accepted for evidence of linkage between a marker and a disease. The segregation of some chromosome 19 markers with dystonia in this family may be a chance finding. Such a finding may be expected by chance when numerous observations are made (markers studied). The chromosome 19 haplotypes (figure 5.10) demonstrate a shared maternal haplotype in the interval D19S225 to D19S223 for all four affected offspring of individual IV-21. The five at-risk offspring and the WD siblings have not inherited this haplotype (with the exception of the D19S422 to D19S412 segment in individual VI-24), but the haplotype does not appear to segregate with dystonia in individual VI-1. Absence of segregation with dystonia in VI-1 may exclude this region as the site of the DYT4 locus. However, phase cannot be determined for markers D19S414 and D19S225 in VI-1 as both parents share the same genotype. Thus it remains theoretically possible that the DYT4 locus may be in the interval between D19S414 and D19S416 if one assumes a crossover between markers D19S225 and D19S416 in the meiosis between V-15 and her affected son. The data currently available are therefore inconclusive and the area cannot be entirely excluded as the site of the DYT4 locus. Analysis of further markers in the interval between D19S414 and D19S416 to determine phase in VI-1 would be of interest when fresh DNA samples are available, as would genotyping all additional at-risk family members in this pedigree for chromosome 19 markers. Furthermore, as microsatellite markers were analysed using a radioactive labelling 129 technique, with manual scoring of autoradiographs, which is known to be less accurate than fluorescence based techniques (Schwengel et al, 1994), rescoring of chromosome 19 markers to exclude genotyping errors would be of value. It is not possible to reconstruct chromosome 19 haplotypes for the affected family members V-2, V-l 1 and V-13. Nevertheless, all three individuals share the same allele at D19S255 as the other affected family members. Five alleles were observed for D19S255 in this family and an equal frequency of 0.2 assumed for all alleles. This assumption is a simplification, however, which does not allow for the fact that allele frequencies in a population are rarely equal. Since V-l 1, V-13 and particularly V-2 are separated from the core pedigree by several meioses, allele frequency estimations will significantly affect the final LOD score. Thus, these individuals may share the common allele by chance if it occurs frequently in the population, but this observation is less likely by chance with a rare allele. Accurate determination of allele frequencies, either by standardisation with the CEPH (Centre d’Etudes du Polmorphisme Humaine) scoring system (for which allele frequencies are available for the Caucasian population) or ideally by genotyping of controls from the Australian population, would improve the accuracy of the LOD scores generated. If the shared allele has a population frequency less than 0.2, the LOD score will be pushed higher for this marker. The estimated maximum LOD score obtainable for this family is Z = 4.09 at 0 = 0 with known genotypes for the 36 samples used in the DYT6/DYT7 analyses, and assuming a marker with five alleles of equal frequency. This estimated LOD score assumes a ‘perfect fit’ (all affecteds share one allele, and no at-risk individuals have this allele) and assumes that the WD siblings are coded as affected. With the WD siblings excluded from the analysis the maximum LOD score falls to Z= 3.48. These estimated LOD scores are a theoretical maximum given these assumptions, but it is possible that some clinically unaffected family members could be non-manifesting gene carriers. Thus, the actual maximum LOD score achievable in this family may be lower than the figures above and could fall short of the threshold Z = 3.0 required for linkage. The penetrance figures derived previously for this family (Ahmad et al, 1993) are based on ascertainment of cases from this family alone as the phenotype is unique, and are therefore susceptible to bias. 130 5.8 FUTURE DIRECTIONS OF STUDY 5.8.1 Non-DYTl PTD families Large families with non-DYTl dystonia such as those described in this chapter are rare and provide a unique opportunity to identify unmapped PTD genes. Identification of these genes will depend on initial linkage mapping using genome-wide searches, followed by a concerted physical, or positional candidate, mapping strategy. The completion of the genome search for the DYT4 locus in family 3 should be the first objective. This will require collection of fresh DNA samples for those individuals for whom stocks are exhausted in order to fill the gaps on several chromosomes where no microsatellite markers have been analysed. In order to increase the power to detect linkage and exclude regions of the genome, future analyses should be performed on a larger sample set consisting of DNA from 36 family members. In addition, typing of additional chromosome 19 markers surrounding D19S225, determination of population allele frequencies for these markers, and retyping of chromosome 19 markers using fluorescence-based methods should be performed. Genome-wide screening for linkage in family 2 is also underway at the Institute of Neurology and would also be feasible in family 1, particularly if re-evaluation of the pedigree resulted in the identification of additional affected family members. Allelic association studies at D18S1098 in British patients with torticollis would be of interest to determine whether a founder mutation common to the British and German populations contributes to this common form of dystonia. However, definitive mutation analysis of these genes in patients with dystonia will have to await the identification of the genes themselves. Attempts to evaluate the importance of environmental factors in the causation of dystonia, particularly the interaction of environmental factors with the ‘disease-susceptibility’ geneotype will be of importance in the future. 5.8.2 Screening for DYT1 mutations in British patients Mutation analysis of torsinA is now underway in British and French kindreds to evaluate the prevalence of mutations (in particular the GAG deletion) in families with a classical early, limb-onset phenotype. This will be accomplished by screening for the 131 GAG deletion using restriction enzyme digests, and by DNA sequencing of the gene for other mutations in those patients found to be negative for the deletion. Previous linkage analysis using chromosome 9q34 markers in families with generalised and segmental dystonia have shown evidence for linkage in some families (Warner et al, 1993). However, most of these families are relatively small, so linkage to the DYT1 locus cannot be confirmed or excluded in individual families. Consideration of all families together suggests heterogeneity with evidence for linkage to the DYT1 locus in an estimated 55% of British and French families. Detailed correlation of genotype and phenotype will become possible for carriers of the 3 bp deletion in British families. It will be of great interest to determine whether the 3 bp deletion is the only mutation in torsinA in British patients with a DYT1 phenotype. Once those families with DYT1 mutations have been identified, the next logical step in the genetic dissection of these dystonias will be the identification of other genetic loci causing generalised and segmental dystonia in the sub-set of patients negative for DYT1 mutations. These studies are ongoing at the Institute of Neurology, and are not described here as the identification of the DYT1 gene came after the completion of the work described in this thesis. 132 CHAPTER STX MUTATION ANALYSTS OF GTP CYCLOHYDROLASE T IN PATIENTS WITH DYSTONIA RESPONSIVE TO ANTICHOLINERGIC DRUGS 6.1 OUTLINE OF CHAPTER This chapter describes a study involving mutation analysis of the GTP cyclohydrolase I gene (GCH1) in a group of seven patients diagnosed with primary torsion dystonia, all of whom had shown an excellent response to anticholinergic drug therapy. The role of GCH1 in dopa-responsive dystonia and the rationale of the present study are described. The results of GCH1 sequencing and restriction enzyme analysis of patients and controls are described, and the significance of the findings discussed. 6.2 INTRODUCTION 6.2.1 Dopa-responsive dystonia Dopa-responsive dystonia (DRD) occupies a unique place amongst the dystonias as the condition may be very effectively treated, and its molecular pathogenesis is better understood than that of any other dystonia. Segawa (Segawa et al, 1971 and 1976) provided the first detailed clinical description of DRD, and described the distinctive clinical phenotype and the dramatic, sustained, therapeutic response to low doses of L- dopa. Patients typically present in childhood with dystonia involving the lower limb which progresses to become generalised unless treated. Diurnal variation of symptoms with improvement of dystonia after sleep is a characteristic feature of DRD. It is now recognised that the DRD phenotype may encompass a number of atypical presentations, including parkinsonism, spastic paraplegia and a presentation mimicking athetoid cerebral palsy (Nygaard et al, 1988; Bandmann et al, 1996). Most cases are inherited as an autosomal dominant trait with reduced penetrance, and females outnumber males by approximately three to one. Linkage between the DRD locus and markers on chromosome 14q (Nygaard et al, 1993) led to the identification of heterozygote mutations in GCH1 in several cases of typical DRD (Ichinose et al, 1994). GCH1 133 enzyme activity in affected individuals was shown to be reduced to 2 - 20% of normal levels (Ichinose et al, 1994). The enzyme GCH1 catalyses the initial and rate-limiting step of tetrahydrobiopterin synthesis (figure 6.1). Tetrahydrobiopterin is an essential co- factor for three aromatic amino acid mono-oxygenases, tyrosine hydroxylase, tryptophan hydroxylase and phenylalanine hydroxylase, as well as for nitric oxide synthase. Tyrosine hydroxylase is in turn the rate-limiting enzyme in the dopamine synthesis pathway. Dopamine deficiency in the CNS is thought to be central to the development of dystonia in DRD (Williams, 1995) as L-dopa therapy alleviates dystonia in this condition, and treatment with dopamine receptor antagonists (such as antiemetics and neuroleptics) may cause both acute and tardive dystonia. Rare autosomal recessive inheritance of a DRD phenotype has also been described in patients with homozygous mutations in other genes coding for enzymes in the dopamine synthesis, or tetrahydrobiopterin synthesis pathways. Ludecke et al (1995) reported two brothers with homozygous missense mutations in the tyrosine hydroxylase gene associated with reduced activity of the enzyme (Knappskog et al, 1995). Hanihara et al (1997) reported two sisters from a consanguineous marriage with an atypical form of DRD characterised by severe dopa-responsive dystonia with diurnal variation, mental retardation, and seizures. Both sisters had homozygous mutations in gene coding for the enzyme 6-pyruvoyl-tetrahydropterin synthase, which catalyses the second step in tetrahydrobiopterin synthesis. 6.2.2 Anticholinergic-responsive dystonia Anticholinergic drug therapy is widely recognised to be of benefit in some patients with primary torsion dystonia. Benzhexol hydrochloride (Artane) taken at high dosage (up to 120 mg daily) has been shown to result in significant improvement of symptoms in a proportion of patients with dystonia, but the response is variable and often poorly sustained (Fahn, 1983). However there have been a number of reports of dramatic and sustained improvement in patients with torsion dystonia treated with anticholinergic drugs at low dosage (Comer, 1952; Bums, 1959; Mucklow and Metz, 1964; Nygaard et al, 1991; Harwood et al, 1994). In 1952 Comer reported two siblings with severe generalised dystonia, both of whom had shown a remarkable response to benzhexol therapy at a dose of 7 to 8 mg daily 134 Figure 6.1: The tetrahydrobiopterin synthesis pathway and its relationship to the synthesis of dopamine. GTP GTP cyclohydrolase I dihydroneopterin triphosphate 6-pyruvoyl tetrahydrobiopterin synthase \ ' 6-pyruvoyl tetrahydrobiopterin 2-ketoreductase t sepiapterin sepiapterin reductase t tetrahydrobiopterin dihydrobiopterin tyrosine hydroxylase tyrosine ------► dopa dopamine 135 (Comer, 1952). Dystonia in a paternal uncle indicating autosomal dominant inheritance with reduced penetrance, as well as the clinical description, including prominent diurnal variation in symptoms, suggest that these patients may have had DRD. Subsequently Nygaard et al provided an update on one of the siblings, whose response to low dose benzhexol had been maintained for 30 years, and described another patient with a similar response (Nygaard et al, 1991). Both had been switched to L-dopa with good results, supporting the belief that these patients had DRD. Several other cases of generalised dystonia with marked response to low-dose anticholinergic therapy have been described and have led to suggestions that DRD underlies many such cases. 6.3 AIMS AND DESIGN OF THE STUDY This study was designed to test the hypothesis that the phenotype of anticholinergic- responsive dystonia is a variant of DRD resulting from GCH1 mutations. Patients were selected from two sources: firstly from a previous survey of 107 British patients with generalised, multifocal and segmental dystonia (Fletcher et al, 1990b) fulfilling the agreed criteria for primary torsion dystonia (Fahn et al, 1987); secondly from the genetics database at the National Hospital for Neurology and Neurosurgery. Patients were examined and classified as definitely affected, unaffected, or possibly affected, as described in section 5.4, with the exception that rapid, jerky dystonic tremor was considered sufficient for a diagnosis of definite dystonia. Eighty-four patients were identified whose response to treatment with anticholinergic chugs was known. Of these, seven patients were selected for study, on the basis of a marked response to the anticholinergic drug benzhexol at doses ranging from 15-120mg daily, such that dystonia was completely or virtually abolished. Patients with a less marked response to anticholinergic drugs were not included. GCH1 was sequenced (as described in section 2.5) in these seven cases to identify mutations. Sequence variants identified by sequencing may be pathogenic mutations or neutral polymorphisms. A pathogenic mutation is more likely where: the sequence change identified changes a codon to one coding for a different amino acid or a termination codon (non-synonymous), the amino acid is replaced by another with a different side chain type (non-conservative change), and the amino acid is highly conserved through evolution suggesting an important functional or structural role. Pathogenic mutations 136 should segregate with the disease within families, and should not be found in control subjects in a dominant disease. Therefore, 105 control samples were analysed for the presence of sequence variants identified by DNA sequencing. Loss or gain of restriction enzyme cutting sites resulting from the sequence variant were used to screen controls. 6.4 PATIENTS AND CONTROLS The pedigrees of the seven index cases studied are given in figure 6.2. No patients had Jewish ancestry. All had been investigated for causes of secondary dystonia, with negative results. The clinical characteristics of patients studied are summarised in table 6.1. Dystonia started in the lower limb in all seven patients. The average age at onset was 8.7 years (range 5 to 20 years). Dystonia progressed to become generalised in five patients, and was segmental and multifocal in one patient each. After treatment with benzhexol, dystonia was either entirely absent or minimal on examination in all patients. Five patients had a family history compatible with autosomal dominant inheritance with reduced penetrance. The mother of patient 5, the brother of patient 1, and the mother, brother and daughter of patient 6, all suffered from PTD. In two families (1 and 3), other family members were asymptomatic but on examination were classified as affected on the basis of writer’s cramp, dystonic posturing of the outstretched hands, and marked postural tremor. The brother of patient 1 had been treated with benzhexol with almost complete resolution of dystonia (dose not known). No other affected family members had received treatment with anticholinergic drugs. Two patients (4 and 5) had received a trial of L-Dopa treatment with no improvement in symptoms, but were included in the study because it was not known if the dose administered had been sufficient to exclude a clinical response. The 105 control subjects were made up of the spouses of patients with Huntington’s disease and other neurological disorders from the database of the Institute of Neurology, London. None suffered from dystonia. All controls were Caucasian, as were the patients studied. 137 Figure 6.2: Pedigrees of index patients with anticholinergic-responsive dystonia. Individuals with dystonia are denoted by filled symbols and index cases studied are indicated with arrows. GCH1 genotypes are given for families 2 and 7. Wt = wild type allele. Family 1 Patient 1 Family 2 Family 3 a Pro23Leu/wt Lys224Stop/wt o Patient 2 wt/wt Pro23Leu/Lys224Stop Patient 3 Family 5 Family 4 Patient 5 Patient 4 Family 6 Family 7 wt/wt Aspl 15Asn/wt Patient 5 Patient 7 Aspl 15Asn/wt wt/wt Aspll5Asn/wt 138 Table 6.1: Summary of clinical features of index cases studied and GCH1 mutations identified. *3 >* ffl N O & £ 5 a o s O © © U-< O b E Z Z N ( . i t o 00 O B o o k VO z z O O Three heterozygote point mutations of the GCH1 gene were identified in two patients (table 6.2). The presence of mutations was confirmed by sequencing of the antisense DNA strand, and by use of restriction enzyme digests to confirm genotype. All three base changes identified were non-synonymous. The GCH1 amino acid residues altered by the mutations were highly conserved between species. In patient 2, an A-»T transversion in exon 6 (figure 6.3) changes a lysine residue at position 224 to a premature stop codon (Lys224Stop). A C-»T transition (figure 6.4) causing a proline to leucine amino acid change at codon 23 in exon 1 (Pro23Leu) was also identified in this patient. These mutations create newyl/w-/and Taq-I cleavage sites respectively. Restriction site analysis confirmed the presence of mutations in the patient, and revealed the codon 23 mutation to be present in one of 210 control chromosomes screened (in a male). The codon 224 mutation was not found in controls. Parental genotypes for patient 2 were determined by restriction site analysis, which revealed the Lys224Stop mutation to be present in the mother, and the Pro23Leu variant in the father. Both parents were asymptomatic and normal when examined (in the afternoon) by the author and another neurologist on separate occasions. The patient’s sister carried neither mutation. A G-»A substitution in the last nucleotide of exon 1 (figure 6.5) in patient 7 results in a missense aspartic acid to asparagine amino acid change at residue 115 (Aspl 15Asn). This mutation abolishes a Dde-I restriction site which was used to confirm the presence of the mutation in the patient; the mutation was not found in any of 210 control chromosomes. This patient was a sporadic case with no family history of dystonia. However, the patient’s father and brother were also found to have the mutation but both were asymptomatic, and on careful examination by the author (in the afternoon), manifested no dystonia or parkinsonism. An unaffected sister did not carry the mutation. No sequence variants were identified in the GCH1 coding region in the remaining 5 patients. 140 Table 6.2: Summary of GCH1 mutations identified and resulting restriction site changes. CONTROL PATIEN EXON CODON PROTEIN RESTRICTION SITE CHROMOSOME T CHANGE ALTERATIO CREATED/ABOLISHE MUTATIONS N D 2 6 AAA->TAA Lys224Stop Alu-I + 0/210 2 1 CCC->CTC Pro23Leu Taq-I + 1/210 7 1 GAT->AAT Aspll5Asn* Dde-I - 0/210 +/- indicates creation or abolition of restriction site respectively * Disrupts -1 position of exon 1-intron 1 splice donor site. 141 Figure 6.5: Sequencing elecropherogram showing a G —> A transition in exon 1 of GCH in patient 7. The control sample is illustrated in the lower panel. The upper panel shows both G (black) and A (green) peaks at position 343 in the patient (marked by the arrow) indicating a heterozygote mutation. I 1200 - 1200- 144 Figure 6.4: Sequencing elecropherogram showing C —> T transition in exon 1 of GCH in patient 2. The control sample is illustrated in the lower panel. The upper panel shows both C (blue) and T (red) peaks at position 68 in the patient (marked by the arrow) indicating a heterozygote mutation. i 1200 - 1200- 143 Figure 6.3: Sequencing elecropherogram showing an A —» T transversion in exon 6 of GCH1 in patient 2. The control sample is illustrated in the lower panel. The upper panel shows both A (green) and T (red) peaks at position 670 in the patient (marked by the arrow) indicating a heterozygote mutation. I 6.6 DISCUSSION Three mutations in the GCH1 gene were identified in two out of seven patients with highly anticholinergic-responsive dystonia. These findings indicate that these two patients have a form of DRD. Patient 2 is a compound heterozygote with Lys224Stop inherited from her mother and Pro23Leu inherited from her father. The nonsense mutation found in exon 6 creating a stop codon would be predicted to result in premature truncation of the GCH1 protein with loss of the C-terminal 10% of the protein. The Pro23Leu mutation in the paternal GCH1 allele was found in one of 210 control chromosomes examined, raising the possibility that Pro23Leu may be a rare variant in the population of no pathological significance. However, several lines of evidence suggest that the mutation is pathogenic. First, Pro23Leu changes a highly conserved amino acid residue of the gene. Second, the mutation results in a non-conservative amino acid change. Third, the control subject in whom the mutation was found was male and is therefore more likely than a female to be a non-manifesting carrier of DRD. The estimated penetrance of DRD is only 30% and is sex-dependent with affected females outnumbering males by about 3:1. There are several examples, including the family of patient 7 described here, of the same mutation causing DRD in females, but resulting in no phenotype in male members of the same family (Ichinose et al, 1994). The probability of the control subject with Pro23Leu manifesting a DRD phenotype if the mutation were pathogenic is therefore reduced by virtue of his gender. The fourth and final line of evidence is that the same Pro23Leu mutation present in patient 2 (CCC-»CTC), has also been identified as a putative pathogenic mutation in two other patients with DRD since the completion of this study (de la Fuente-Femandez, 1997; Muller et al, 1998). The family of patient 2 is of British ancestry, while the other patients with Pro23Leu are Spanish and German, making it unlikely that the three families are related. The identification of this mutation in three DRD families is strongly in favour of a pathogenic role of the mutation. However, the possibility that Pro23Leu may be a neutral variant in the population, and not responsible for DRD in the other patients reported, cannot be entirely ruled out. It should be noted that most of the putative GCH1 mutations previously reported have not been screened- for in control populations (Bandmann et al, 1996; Furukawa et al, 1996; de la Fuente- Femandez, 1997). 145 The compound heterozygote status of patient 2 is of interest. Homozygote missense mutations of GCH1 result in undetectable enzyme activity (Ichinose et al, 1995a), and patients typically present in infancy with hyperphenylalaninaemia, severely retarded development, abnormal muscle tone and convulsions. The parents of patient 2 are both asymptomatic gene carriers with no manifestations of DRD. This is the first observation of compound heterozygote mutations of GCH1 in a patient with dystonia, and therefore raises the possibility that dystonia due to GCH1 deficiency may be inherited as an autosomal recessive trait in a minority of cases. It may be that neither mutation alone is sufficient to result in a DRD phenotype (particularly in a male) but that a mutation in each allele is sufficient to reduce GCH1 activity below a critical threshold required to cause DRD. Very recently, three other patients with compound heterozygote GCH1 mutations have been described (Furukawa et al, 1998; Bezin et al, 1998), confirming the belief that GCH1-deficient dystonia may be inherited as an autosomal recessive trait. These patients had particularly severe dystonia and developmental motor delay, consistent with a phenotype of intermediate severity between GCH1-deficient hyperphenylalaninaemia and DRD. In one of these cases both parents were healthy carriers, and in the others the mothers and their female relatives had dystonia. The phenotype of patient 2 in this study is less severe than that of the other compound heterozygotes subsequently described, an observation which may mitigate against the contention that the Pro23Leu is pathogenic. Interestingly, the Aspl 15Asn mutation in patient 7 may disrupt a splice donor site as well as changing the amino acid at residue 115. G—»A mutations at the -1 position of 5’ splice sites have been shown to disrupt RNA splicing in a number of other diseases, including: acute intermittent porphyria, Ehers-Danlos syndrome, Tay-Sachs disease, epidermolysis bullosa, and guanidinoacetate methyltransferase deficiency (Grandchamp et al, 1989; Weil et al, 1989; Akli et al, 1990; Gardella et al, 1996; Stockier et al, 1996). Guanine is the majority nucleotide at the last base position of exons, and mutations of this nucleotide may disrupt the splice donor site, resulting in exon skipping. Splice site mutations have also been described in DRD but these have been within introns in all cases reported to date (Hirano et al, 1995; Furukawa et al, 1996; Muller et al, 1998; 146 Steinberger et al, 1998). Future analysis of RNA splicing would therefore be of interest in this patient All three mutations identified in this study were novel mutations in GCH1. Until recently, all mutations identified in this gene in patients with DRD have been unique to the family in which they were identified. This observation, together with the identification of new mutations in sporadic cases of DRD (Furukawa et al, 1998), suggest that there is a high mutation rate in this gene. However, the Pro23Leu mutation recently identified in two additional patients with DRD, is identical to the mutation in patient 2 (CCC —» CTC). This is the first example of a single mutation in different families with DRD. If Pro23Leu is pathogenic, this finding suggests that the highly conserved proline residue at position 23 is either part of a functionally important domain, and/or that it represents an area particularly susceptible to mutation (a mutation ‘hotspot’). Identification of Pro23Leu in one of 210 control chromosomes favours the latter explanation. It is worth noting also that a different mutation in codon 224, causing a conservative amino acid change (Lys224Arg, AAA —» AGA), has been identified in two other patients with unusual presentations of DRD (Bandmann et al, 1996; Furukawa et al, 1998). Five patients had no identifiable mutation within the coding region of the GCH1 gene. This does not exclude the possibility that these individuals have a variant of DRD. Up to half of all typical DRD patients have no detectable abnormality in the coding region of GCH1 (Ichinose et al, 1995b; Bandmann et al, 1996). Mutations outside the coding region of GCH1 - for example in the gene promoter, or mutations in a regulatory gene, or gene coding for another enzyme of the tetrahydrobiopterin synthesis pathway - may be responsible for DRD in such cases. However, it may be that the patients without GCH1 mutations suffer from PTD rather than dystonia caused by GCH1 deficiency. It has been suggested that patients with PTD respond more slowly to anticholinergic drugs than patients with DRD as well as showing a generally poorer response and needing higher doses (Nygaard et al, 1991). The higher dose of benzhexol required to produce a response in patients without identified mutations, and the absence of diurnal variation of symptoms suggest that PTD may be the cause of dystonia in these patients. This contention is supported by recent laboratory work performed by the author and Dr E.M. Valente, since the completion of the work described in this thesis. This has shown 147 that patients 1 and 5 have the 3 bp deletion in the torsinA gene, confirming at a molecular level the belief that some of these patients have DYT1-related PTD. The phenotype of patients with and without GCH1 mutations was broadly similar in this study. The major clinical clue to the diagnosis of DRD in patient 2 was the presence of diurnal fluctuation of symptoms. However, absence of diurnal variation does not exclude the diagnosis, as exemplified by patient 7. The difficulty in distinguishing between PTD and DRD on the basis of clinical features alone makes a therapeutic trial of L-dopa mandatory in all cases of early-onset dystonia. Not all patients with clinically definite DRD respond to anticholinergic drugs (Harwood et al, 1994). Nygaard et al (1991) found that 17 patients with DRD had previously tried an anticholinergic drug with varying degrees of success. Overall, most patients with DRD derived some benefit from anticholinergics, but in the majority L-dopa therapy was judged superior in terms of both efficacy and adverse effects. Anticholinergic therapy is not a substitute for L-dopa in DRD. In summary, three mutations were identified in two patients with anticholinergic- responsive dystonia, suggesting that a variant of DRD underlies dystonia in a minority of patients with apparent PTD. These findings highlight the difficulty of distinguishing clinically between PTD and DRD. A therapeutic trial of L-dopa is therefore essential in all cases of dystonia with onset in childhood or early adult life, even in the absence of any diurnal fluctuation of symptoms or family history. A favourable response to anticholinergic drugs in patients with what appears to be PTD should alert clinicians to the possibility that such patients have an L-dopa responsive dystonia. 148 CHAPTER SEVEN CONCLUSIONS AND FUTURE DIRECTIONS OF STUDY 7.1 GENERAL CONCLUSIONS Over the last decade there have been considerable advances in molecular genetics, which have resulted in the discovery of genes responsible for numerous monogenic diseases. Rapid advances in technology and genetic resources, such as increasingly detailed genetic maps, have made possible the identification of novel genes and disease- associated mutations by means of the strategies of functional cloning and positional cloning. These techniques have been reviewed in chapter 1. The aims of this study were to apply the initial steps of a positional cloning strategy - linkage analysis using anonymous genetic markers - to locate the genes for three dominantly inherited movement disorders: paroxysmal dystonic choreoathetosis, hereditary geniospasm, and non-DYTl primary torsion dystonia. Subsequent candidate gene analysis was to be performed where possible. In addition, direct DNA sequencing was used to attempt to identify GCH1 mutations in patients with anticholinergic-responsive dystonia. 7.1.1 Mapping a gene for paroxysmal dystonic choreoathetosis A large British family with paroxysmal dystonic choreoathetosis were traced from two index cases. The clinical phenotype was studied in detail and compared with that of previously reported families, and a list of diagnostic criteria for PDC proposed. A genome-wide search using microsatellite markers resulted in the identification of a locus on the long arm of chromosome 2 in this family. Further detailed genetic mapping and haplotype analysis made it possible to assign the gene to a 3.8 cM genetic interval at chromosome 2q36-q37 (maximum LOD score 8.7). This represents an improvement in the localisation of the FPD1 gene beyond that reported by other groups working concurrently. The mapping data from other families with classical PDC is consistent with a locus on the distal long arm of chromosome 2, confirming genetic homogeneity in all kindreds thus far reported. In accordance with the positional cloning paradigm, genetic databases were searched for putative candidate genes in the FPD1 interval. 149 Three candidate genes were identified, and two of these studied by using intragenic polymorphisms to perform further linkage analysis. Absence of segregation of the polymorphism in the delta subunit of the nicotinic acetylcholine receptor gene with PDC allowed this gene to be excluded as a candidate. The neuronal chloride/bicarbonate anion exchanger (SLC4A3), however, segregates with the disease in the ‘B’ family (LOD score 6.08). These mapping data, together with the known physiological importance of SLC4A3 in maintaining neuronal homeostasis, suggest that this gene warrants further study. 7.1.2 Mapping a gene for hereditary geniospasm In the course of this study two large families with hereditary geniospasm were identified and interviewed. A genome-wide search for linkage in the larger of these families identified linkage to a 2.1 cM interval at chromosome 9ql3-21 (maximum LOD score 5.24). The second, clinically indistinguishable, family was not linked to this locus, providing evidence for genetic locus heterogeneity in this condition. No genes which may be considered obvious candidates for GSM1 are known to map to this region. 7.1.3 Linkage studies in primary torsion dystonia Three families with primary torsion dystonia were studied. All had an ‘intermediate’ phenotype between that of early, limb-onset dystonia, and late-onset focal dystonia. The DYT1 locus had previously been excluded by linkage analysis in the families. Recently identified loci for non-DYTl dystonia on chromosome 8 (DYT6), and chromosome 18 (DYT7), were studied. Makers spanning the DYT6 and DYT7 loci were analysed in these families with negative results. These results indicate that non-DYTl dystonia is genetically heterogeneous, and that one or more genes for PTD remain to be discovered. A genome-wide search was partially completed in the largest of the families, with identification of a region on chromosome 19 with positive LOD scores, but no definitive linkage has yet been established. 7.1.4 GCH1 mutation analysis in patients with anticholinergic-responsive dystonia In view of the observation that some patients with apparent primary torsion dystonia are highly responsive to anticholinergic drugs, mutation analysis of GCH1 was undertaken 150 to test the hypothesis that a proportion of such patients have a form of GCH1-deficient dystonia. Seven patients were identified who had been diagnosed with PTD and who showed an excellent or complete response to anticholinergics. Sequencing of all six exons of GCH1 in these patients revealed three novel point mutation in two patients, one of whom was a compound heterozygote. These findings expand the range of recognised clinical presentations of GCH1 mutations, and serve to emphasise the importance of a therapeutic trial of L-dopa in all patients with early-onset dystonia. 7.2 FUTURE DIRECTIONS OF STUDY Physical mapping within the FPD1 and GSM1 candidate regions was not undertaken as part of this study as it is a technically demanding, and time-consuming process, currently beyond the capabilities of many laboratories. However, the determination of the entire human genome sequence with the completion of the Human Genome Project is predicted to result in a diminishing need for physical mapping as the candidate gene approach becomes increasingly applicable. The technology required for high- throughput, reliable mutation analysis of candidate genes is evolving rapidly. Current techniques include semi-automated DNA sequencing, single-strand conformation polymorphism analysis or mismatch cleavage techniques (Strachan and Read, 1996). There is also the promise of very rapid DNA sequencing using the chip-based hybridisation technology now being developed (Goffeau, 1997). Sequencing of the SLC4A3 candidate gene is already underway in the United States, in patients with PDC. As new genes in the FPD1 and GSM1 candidate regions are identified, it will be possible to screen for mutations with ever-increasing ease. Concurrent work to refine the genetic mapping of these disorders in additional families, using linkage analysis or linkage disequilibrium, may narrow the candidate regions for the FPD1 and GSM1 genes. Similarly, it may be possible to map new loci for primary torsion dystonia, although success will be determined by the availability of suitable families. Study of the pathophysiology of movement disorders using physiological, pharmacological and functional imaging techniques will complement genetic analysis, and may improve selection of candidate genes as understanding of these disorders increases. 151 At the level of mutation analysis of known disease genes such as GCH1, or DYT1, several challenges remain. Approximately half of all patients with clinically typical DRD have no identifiable mutation in the coding region of the GCH1 gene. Mutations affecting splicing, promoter or regulatory regions of the gene may underlie a proportion of these cases. Alternatively, genes regulating GCH1 activity, or other genes affecting dopamine synthesis or neurotransmission may be involved. In addition, the genetic and environmental factors which determine the highly variable penetrance of DRD and PTD have yet to be studied. Molecular genetics holds great promise for the future in the fields of diagnosis, understanding of pathophysiology, and ultimately for therapeutic advances in treating human disease. The present era of disease gene discovery is not an end in itself but a new beginning for medicine. 152 APPENDIX ONE RESULTS OF TWO-POINT LINKAGE ANALYSIS IN THE ‘ft’ FAMILY Two-point LOD scores for 44 microsatellite markers analysed in the ‘B’ family are presented here (see chapter 3). These were analysed as part of a genome-wide search for linkage of the locus for paroxysmal dystonic choreoathetosis in this family. Parameters used for linkage analysis are given in section 3.4. Markers are arranged in chromosome order and are ordered from p-ter to q-ter. The map positions of markers are given in Reed et al (1994). Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D1S255 -17.17 -9.22 -5.41 -2.59 -1.4 -0.38 0 0.09 D1S197 -19.52 -14.02 -8.65 -4.26 -2.4 -0.81 -0.17 0 . 04 D1S209 -6.05 -3.37 -1.44 -0.17 0.24 0.43 0.34 0.17 D2S142 -17.13 -7.66 -4.35 -1.73 -0.63 0.26 0.49 0.38 D4S405 — o o -20.44 -12.96 -6.84 -4.26 -1.91 -0.79 -0.22 D4S250 — o o -12.96 -9.37 -5.11 -3.11 -1.3 -0.52 -0.17 D4S430 — o o -14.5 -8.67 -4 .54 -2.82 -1.27 -0.52 -0.14 D4S408 -15.43 -11.12 -6.45 -3.02 -1.61 -0.41 0.05 0.15 D5S416 — 0 0 -12.66 -8.15 -4.35 -2.66 -1.1 -0.4 -0.09 D5S426 — 0 0 -4 -2.08 -0.74 -0.24 0.11 0.17 0.11 D5S428 -21.78 -18.09 -15.02 -10.38 -6.95 -3.51 -1.7 -0.63 D5S421 -17.19 -11.77 -8.28 -5.18 -3.52 -1.62 -0.77 -0.33 D5S400 — o o -14.63 -9. 67 -5.4 -3.4 -1.55 -0.66 -0.2 D7S484 — 0 0 -12.89 -9.14 -5.86 -3.93 -1.97 -0.95 -0.34 D7S502 -16.3 -8.99 -4.15 -0.92 0.23 0.92 0.86 0.48 D7S524 — o o -21.93 -15.6 -9.3 -6.06 -2.91 -1.32 -0.44 153 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D7S486 — OO -14.55 -9.72 -6.32 -4.48 -2.21 -1.1 -0.46 D8S552 — oo -12.74 -6.95 -2.95 -1.4 -0.19 0.18 0.18 D8S283 — oo -6.59 -4.44 -2.12 -1.03 -0.14 0.14 0.15 D8S260 — oo -12.78 -9.58 -6.34 -4 .38 -2.17 -1 -0.34 D8S257 -8.98 -7.9 -6.2 -4.28 -3.29 -1.63 -0.7 -0.22 D8S284 -17 .75 -12.8 -8.8 -4.85 -2.76 -0.97 -0.27 -0.03 D9S176 — 00 -22.51 -15.92 -9.54 -6.34 -3.18 -1.53 -0.57 D9S179 -13.81 -10.65 -7.75 -4.98 -3.48 -1.94 -1.05 -0.45 D10S537 -13. 6 -10.74 -7.69 -4.63 -3.11 -1.62 -0.83 -0.34 D11S925 -8.26 -5.85 -4.54 -2.7 -1.79 -0.9 -0.44 -0.16 D11S907 -8.66 -6.37 -4.47 -3.1 -2.09 -0.9 -0.38 -0.16 D12S99 -9.65 -6.69 -4.01 -2.02 -1.24 -0.58 -0.3 -0.13 D12S358 -10.04 -6.79 -4.03 -2.03 -1.24 -0.56 -0.26 -0.1 D12S62 -9.37 -7.03 -4.36 -2.27 -1.37 -0.55 -0.22 -0.09 D12S368 — oo -18.31 -12.24 -6.86 -4.36 -1.96 -0.76 -0.17 D12S95 — oo -7 -4.08 -2.02 -1.18 -0.47 -0.17 -0.03 D12S97 -21.51 -11.13 -6.26 -2.96 -1.7 -0.67 -0.26 -0.08 D13S153 -22.44 -15.01 -10.28 -6.04 -3.86 -1.75 -0.72 -0.21 D13S170 -18.58 -13.98 -10.21 -6.48 -4.24 -2.04 -0.99 -0.41 D13S158 -21.06 -15.89 -11.43 -6.55 -4.17 -1.93 -0.84 -0.27 D14S50 — OO -12.61 -7.47 -3.48 -1.91 -0.65 -0.2 -0.06 D14S75 -19.7 -15.7 -11.43 -7.35 -5.14 -2.71 -1.37 -0.53 D14S63 -18 .78 -12.14 -7.45 -3.96 -2.46 -1.09 -0.45 -0.11 D14S68 -16.58 -11.7 -7.92 -5.08 -3.39 -1.41 -0.49 -0.08 154 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D18S62 -6.48 -5.47 -4.47 -2.85 -1.75 -0.79 -0.38 -0.17 \— i— i—I CD 1 1 1 D18S71 -26.55 -16.75 -6.24 -3.99 -1.83 -0.77 -0.22 D18S57 — oo -5.44 -2.5 -0.57 0.11 0.52 0.51 0.32 D18S61 -17 .81 -12.46 -8.55 -5.15 -3.29 -1.57 -0.75 -0.29 155 APPENDIX TWO RESULTS OF TWO-POINT LINKAGE ANALYSIS IN HEREDITARY GENIOSPASM - FAMILY 1 Two-point LOD scores for 104 microsatellite markers analysed in hereditary geniospasm family 1 (see chapter 4). These were analysed as part of a genome-wide search for the locus for hereditary geniospasm. Parameters used for linkage analysis are given in section 4.3.3. Markers are arranged in chromosome order and are ordered from p-ter to q-ter. The map positions of markers are given in Reed et al (1994). Distances between chromosome 9 markers are illustrated in figure 4.3. CHROMOSOME 1 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D1S207 -10.16 -4.79 -2.83 -1.42 -0.83 -0.32 -0.11 -0.04 D1S252 — OO -6.24 -3.28 -1.23 -0.43 0.17 0.27 0.15 D1S484 — oo -6.06 -3.33 -1.21 -0.32 0.37 0.48 0.26 D1S196 — 00 -5.73 -2.79 -0.79 -0.04 0.45 0.46 0.25 D1S416 -6.06 -2.74 -0.82 0.37 0.69 0.7 0.44 0.15 D1S238 -3.22 -0.5 0.46 0.99 1.08 0.93 0.61 0.23 D1S413 — o o -3.59 -1.62 -0.34 0.09 0.3 0.24 0.1 D1S249 — oo -10.94 -7.93 -5.77 -4 .51 -2.54 -1.3 -0.5 D1S229 — 00 -9.28 -5.36 -2.6 -1.44 -0.4 0.02 0.13 156 CHROMOSOME 2 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D2S131 — OO -9.43 -7.3 -4 .77 -3.11 -1.47 -0.64 -0.19 D2S139 — 0 0 -4.7 -3.54 -1.87 -1.03 -0.3 -0.03 0.04 D2S160 -11.77 -9.42 -7.36 -4.33 -2.64 -1.12 -0.42 -0.09 D2S142 -7.14 -6. 64 -5.59 -3.33 -2.13 -0.98 -0.4 -0.11 DS326 — o o -10.09 -7.16 -4.43 -2.91 -1.38 -0. 61 -0.19 D2S72 — o o -7.36 -4.36 -2.2 -1.29 -0.47 -0.12 0.01 D2S157 — o o -6.93 -3.94 -1.85 -0.99 -0.26 0.03 0.09 D2S126 — o o -5.9 -4.8 -2.77 -1.6 -0.54 -0.08 0.07 CHROMOSOME 3 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D3S1279 — o o -5.44 -3.44 -1.93 -1.23 -0.56 -0.26 -0.11 D3S11 — o o -10.73 -6.74 -3.83 -2.52 -1.24 -0.56 -0.18 D3S1303 -8.9 -7.71 -5.42 -3.19 -2.13 -1.05 -0.48 -0.16 D3S1297 -5.01 -5 -4.79 -3.35 -2.37 -1.33 -0.73 -0.31 D3S1314 — o o -9.65 -5.96 -3.18 -1.99 -0.88 -0.34 -0.08 D3S1265 — o o -9.1 -6.2 -4 .22 -3.04 -1.52 -0.7 -0.23 CHROMOSOME 4 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D4S403 -9.02 -7.04 -5.96 -3.57 -2.18 -0.94 -0.38 -0.1 D4S418 — 0 0 -6.62 -3.65 -1.64 -0.87 -0.26 -0.05 0.01 D4S398 -7.06 -5.76 -4.75 -2 . 99 -1.89 -0.89 -0.42 -0.17 D4S250 -6.54 -5.5 -4 .53 -3.22 -2.07 -0.9 -0.33 -0.06 157 CHROMOSOME 5 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D5S268 — oo -7.46 -5.42 -3.65 -2.52 -1.29 -0.65 -0.28 D5S426 —>00 -10.64 -7.33 -4.05 -2.53 -1.13 -0.46 -0.12 D5S407 — oo -8.63 -6.31 -3.71 -2.44 -1.2 -0.57 -0.2 D5S428 — oo -2.35 -1.36 -0.68 -0.4 -0.16 -0.06 -0.01 D5S409 — oo -5.6 -3.45 -1.74 -1.01 -0.39 -0.14 -0.03 D5S421 — oo -9.17 -6.04 -3.55 -2.35 -1.12 -0.47 -0.13 D5S210 — 00 -6.62 -5.59 -4.15 -2.73 -1.29 -0.55 -0.16 D5S410 -3.57 -3.56 -3.52 -3.23 -2.59 -1.55 -0.87 -0.38 D5S400 — oo -8.56 -5.62 -3.77 -3.1 -1.99 -1.06 -0.43 D5S408 — oo -4.76 -2.77 -1.37 -0.8 -0.33 -0.14 -0.04 CHROMOSOME 6 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 O O CM D6S260 — oo -9.24 -5.57 -2.78 -1.57 -0.48 1 0.11 D6S273 — oo -5.74 -2.8 -0.86 -0.15 0.35 0.42 0.29 — oo i D6S291 -7.02 -3.95 -1.75 o 00 -0.2 -0.01 0.01 D6S262 — oo -9.15 -5.21 -2.43 -1.29 -0.29 0.09 0.16 o o •^r D6S308 -6.31 -3.88 -1. 94 -0.58 I 0.35 0.41 0.28 D6S314 — o o -9.16 -5.21 -2.43 -1.29 -0.29 0.09 0.16 o i— — o o l D6S305 -7.26 -3.98 -1.38 i 0.25 0.35 0.22 D6S264 — o o -5. 92 -2.98 -1.05 -0.35 0.13 0.23 0.16 D6S281 0.84 0.84 0.83 0.79 0.72 0.58 0.4 0.21 158 CHROMOSOME 7 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D7S531 — OO -10.41 -7.12 -4.35 -2.99 -1.57 -0.78 -0.29 D7S513 * o o -10.51 -7.59 -5.19 -3.7 -2.01 -1.05 -0.43 D7S629 — 0 0 -8.6 -5.71 -3.81 -2.87 -1.55 -0.77 -0.3 D7S519 — o o -4.4 -2.44 -1.07 -0.53 -0.11 -0.01 -0.02 D7S502 * 0 0 -6.96 -3.99 -1.87 -0.98 -0.23 0.01 0.03 D7S524 — o o -13.14 -8.23 -4.63 -3.03 -1.46 -0. 64 -0.2 D7S486 -3.45 -2.06 -1.09 -0.43 -0.19 0 0.05 0.04 D7S684 * o o -7.73 -5.52 -3.51 -2.44 -1.29 -0. 65 -0.24 D7S550 — o o -7.65 -4.68 -2.48 -1.5 -0.6 -0.23 -0.09 CHROMOSOME 8 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D8S504 0.4 0.4 0.39 0.34 0.29 0.19 0.11 0.05 D8S552 -7.44 -6.11 -4 .37 -2.33 -1.43 -0.6 -0.21 -0.03 0 0 - r CO i— <— * OO 1 I D8S283 -6.39 i i -0.59 -0.03 0.1 0.07 D8S260 -6.32 -3.24 -1.31 -0.05 0.37 0.56 0.44 0.2 D8S257 0.38 0.38 0.36 0.32 0.27 0.17 0.1 0.04 CM o o D8S198 * oo -9.18 -6.2 -3.81 I -1.17 -0.56 -0.23 OO < ■'3’ — 1 — 0 0 i 1 D8S284 -6.9 -4.93 -3.64 -3.13 o OO OO -0.31 * 1 1 t— ►Cs. OO o -J D8S272 -6.58 -5.19 -4.22 -3.44 -2.66 OO -0.32 159 CHROMOSOME 9 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D9S147E — OO -1.22 -0.20 0.54 0.83 0.94 0.74 0.35 D9S161 0.10 0.10 0.15 0.30 0.42 0.49 0.43 0.26 D9S43 — 0 0 0.42 1.41 2.04 2.18 1.99 1.46 0.72 D9S1791 — o o 0.47 1.47 2.10 2.25 2.05 1.52 0.77 D9S200 -8.02 0.90 1.85 2.31 2.31 1.92 1.32 0.60 D9S166 — o o 1.43 2.40 2.94 2.99 2.60 1.88 0.94 D9S1806 -15.28 1.77 2.74 3.26 3.28 2.83 2.05 1.02 D9S1822 4.70 4.69 4.65 4.43 4.10 3.30 2.30 1.12 D9S1837 5.24 5.23 5.18 4.93 4 .55 3.64 2.54 1.24 D9S1876 4.91 4.91 4.86 4 . 61 4 .25 3.39 2.35 1.14 D9S175 -2.86 0.83 1.81 2.40 2.50 2.22 1.62 0.80 D9S1834 -3.28 -0.41 0.62 1.37 1.66 1.67 1.31 0.66 D9S153 -4 .31 -0. 61 0.42 1.17 1.46 1.51 1.19 0.61 D9S257 — o o -5.61 -2.62 -0.51 0.32 0.88 0.83 0.44 CHROMOSOME 11 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D11S922 — OO -11.94 -7.2 -3.72 -2.26 -0.93 -0.34 -0.07 D11S904 — o o -8.65 -6.6 -4.69 -3.35 -1.79 -0.89 -0.33 D11S907 — o o -1.72 -0.74 -0.14 0.05 0.13 0.1 0.05 D11S901 -6.02 -4 . 64 -3.68 -3.11 -2.8 -1.77 -0.95 -0.39 D11S925 —>00 -11.98 -7.12 -3.64 -2.19 -0.89 -0.33 -0.09 D11S968 — 00 -8.28 -4.32 -1.61 -0.57 0.22 0.41 0.31 160 CHROMOSOME 13 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D13S192 — OO -8.13 -5.14 -2.93 -1.92 -0.91 -0.38 -0.1 D13S170 -0.26 -0.25 -0.25 -0.22 -0.19 -0.15 -0.1 -0.06 CHROMOSOME 15 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D15S128 — 00 -4.7 -2.76 -1.3 -0.67 -0.14 0.02 0.02 D15S118 -4.57 -2.69 -1.64 -0.76 -0.33 0.03 0.11 0.06 D15S117 -3.67 -2.64 -1.65 -0.85 -0.48 -0.17 -0.08 -0.04 D15S114 — oo -6.35 -3.42 -1.43 -0.66 -0.05 0.15 0.16 CHROMOSOME 17 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 o CO \—1KO i—1 I—1 O CM o o i—i 1 1 1 1 i >£>- D17S513 — oo -4.96 -3 o 00 D17S784 0.16 0.16 0.18 0.25 0.3 0.32 0.27 0.17 CHROMOSOME 18 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 O l —i i— O 1 — o o i o D18S59 -9.8 -5.81 -2.91 -1.65 o u > 0.11 D18S62 -3.61 -3.61 -3.58 -2.57 -1.68 -0.8 -0.35 -0.11 D18S71 — o o -6.86 -3.89 -1.78 -0.9 -0.16 0.1 0.13 D18S57 — o o -7.56 -4 .57 -2 .38 -1.42 -0.54 -0.15 0.01 D18S61 — o o -6.13 -3.18 -1.14 -0.35 0.22 0.29 0.14 161 CHROMOSOME 19 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D19S221 — OO -10.67 -7.01 -4.16 -2.82 -1.44 -0.7 -0.26 D19S49 - 1 1 . 1 1 -7.95 -4.82 -2.31 -1.31 -0.48 -0.12 0.01 D19S225 — o o -8.75 -5.83 -3.86 -2.89 -1.56 -0.77 -0.28 D19S220 -8.46 -7.94 -5.17 -2.9 -1.84 -0.81 -0.34 -0.16 CHROMOSOME 20 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D20S112 — o o -6.1 -5 -3.44 -2.44 -1.35 -0.72 -0.29 D20S200 — o o -11.55 -8.2 -4.9 -3.29 -1.66 -0.79 -0.27 D20S171 -11.17 -8.81 -6.5 -3.34 -1.96 -0.75 -0.22 0 D20S93 — o o -6.93 -4 . 96 -3.32 -2.31 -1.25 -0.66 -0.28 162 APPENDIX THREE RESULTS OF TWO-POINT LINKAGE ANALYSIS AT THE DYT6 AND DYT7 LOCI Two-point LOD scores for microsatellite markers spanning the DYT6 and DYT7 loci on chromosomes 8 and 18 respectively in three PTD families. For family 3, a second set of LOD scores calculated with the WD siblings (V-24 and V-25) coded as unknown is given in the second row below each marker. Clinical descriptions of families, and parameters used for linkage analysis are given in section 5.4. Markers are ordered from p-ter to q-ter. Map distances in centimorgans (cM) are given for each marker to the marker below. All map distances are according to the Marshfield Clinic integrated chromosome 8 map at: http://www.marshmed.org/genetics/. FAMILY 1, DYT6 LOCUS. Recombination values (0) Marker 0 0.01 0.05 0.1 0.2 0.3 0.4 i D8S1791(6.6cM) -3.15 -2.63 -1.43 o 00 00 -0.39 -0.15 -0.04 O CM CO O O CD 1 1 i D8S538 (OcM) -2.8 -1.1 -0.46 o o 0 D8S601 (2.2 cM) -3.1 -1.39 -0.72 -0.44 -0.19 -0.08 -0.02 o o o o LO i D8S509 (5.9cM) -0.05 1 -0.05 -0.05 -0.03 -0.01 D8S166 (2cM) -2.8 -0.81 -0.18 0.02 0.12 0.09 0.03 D8S285 (2.6cM) 0.65 0.63 0.57 0.48 0.31 0.16 0.04 D8S374 (2.3 cM) 0.77 0.75 0.67 0.57 0.37 0.18 0.05 D8S507 (2cM) 0.39 0.38 0.33 0.27 0.17 0.08 0.02 D8S1113 (2cM) 0.56 0.54 0.47 0.39 0.24 0.11 0.02 o o T - 1 D8S260 (3 . 4cM) -2.8 -0.85 -0.23 1 0.1 0.08 0.03 o i—i oo D8S1797(7 cM) -2.8 1 -0.19 0.02 0.12 0.09 0.03 D8S543(5.8cM) 0.71 0.69 0.62 0.53 0.34 0.17 0.04 o o o X x —1—1 D8S279 (4.3cM) -0.81 -0.16 1 -0.07 l -0.03 -0.01 D8S286 -2.8 -0.81 -0.19 0.02 0.12 0.09 0.03 163 FAMILY 1, DYT7 LOCUS Recombination values (0) Marker 0 0.01 0.05 0.1 0.2 0.3 0.4 O O CM 1 i o D18S59 (16.lcM) -3 -2.48 -1.29 -0.76 -0.3 i-1 o CO ■^r D18S1368(7.5cM) -3.13 -2.56 -1.35 -0.81 I -0.13 -0.03 i -j D18S52(12.8cM) -3.1 -1.39 o N> -0.44 -0.19 -0.08 -0.02 D18S62(OcM) 0.04 0.04 0.04 0.03 0.02 0.01 0 o o ■^r D18S452 (7.3) -6.19 -2.78 -1.44 -0.89 -0.39 -0.15 1 o D18S1163(3.4cM) -3.05 -1.36 -0.7 1 -0.21 -0.09 -0.02 o ■^r I i o o D18S843 -2.7 -1.38 -0.72 -0.19 oo -0.02 FAMILY 2, DYT6 LOCUS. Recombination values (0) Marker 0 0.01 0.05 0.1 0.2 0.3 0.4 D8S1791(6.6cM) -5.48 -3.06 -1.81 -1.3 -0.79 -0.45 -0.19 D8S538(2.2 cM) -2.44 -0.94 -0.36 -0.2 -0.15 -0.15 -0.09 D8S509 (7.9cM) -2.85 -2.57 -1.53 -0.99 -0.49 -0.24 -0.08 D8S285(4.9cM) -2.1 -1.94 -1.07 -0.59 -0.18 -0.03 0.02 D8S507(4cM) -5.61 -2.67 -1.4 -0.93 -0.55 -0.34 -0.15 O O CM D8S260(3.4cM) -5.48 -2.61 -1.28 -0.73 -0.27 -0.09 1 D8S1797(7cM) -5.68 -3.03 -1.65 -1.02 -0.44 -0.17 -0.04 D8S543(5.8cM) -5.38 -2.01 -0.75 -0.31 -0.02 0.02 0.01 CO o o O i—1CM 1 D8S279(4.3cM) -2.78 -2.42 -1.32 -0.79 -0.33 1 o CO D8S286 -2.53 -1.23 -0.63 I -0.29 -0.21 -0.11 164 FAMILY 2, DYT7 LOCUS. Recombination values (0) Marker 0 0.01 0.05 0.1 0.2 0.3 0.4 D18S59 (12.9cM) -5.48 -3.06 -1.81 -1.3 -0.79 -0.45 -0.19 D18S1140(2cM) 0.25 0.24 0.02 0.16 0.09 0.04 0.01 D18S818 (1.2cM) -5.37 -3.15 -1.79 -1.15 -0.55 -0.26 -0.09 D18S1368(4cM) 0.14 0.13 0.11 0.09 0.05 0.02 0.01 D18S1098(1.2cM) 0.09 0.09 0.07 0.06 0.03 0.01 0 D18S481(1.3cM) 0.27 0.26 0.22 0.18 0.1 0.05 0.02 D18S54(lcM) -2.44 -0.94 -0.36 -0.2 -0.15 -0.15 -0.09 D18S52 (12.8cM) -2.85 -2.57 -1.53 -0.99 -0.49 -0.24 -0.08 D18S62 (OcM) -2.1 -1.94 -1.07 -0.59 -0.18 -0.03 0.02 D18S452 (7.3cM) -5.61 -2.67 -1.4 -0.93 -0.55 -0.34 -0.15 D18S1163(3.4cM) -5.48 -2.61 -1.28 -0.73 -0.27 -0.09 -0.02 D18S843 -2.38 -0.86 -0.27 -0.08 0.02 0.04 0.03 165 FAMILY 3, DYT6 LOCUS. Recombination values (0) Marker 0 0 . 01 0 . 05 0.1 0.2 0.3 0 .4 D8S1791(6.6cM) -5. 84 -4 .05 -2 .99 -2.07 -0.98 -0.42 -0 .13 -5. 61 -3. 73 -2 .09 -1.34 -0.69 -0.36 -0 .14 D8S538(2.2cM) -1. 71 -1. 73 - 1 ..7 -1.35 -0.64 -0.26 -0 .07 -2. 31 -2. 01 -1 .03 -0.6 -0.29 -0.18 -0 .1 D8S509(7.9cM) -2. 59 -0. 95 -0 .33 -0.12 0.02 0.06 0 . 05 -2. 59 - 0. 95 -0 .33 -0.12 0.02 0.06 0 . 05 D8S285(4.9cM) -4 .22 -2. 63 -1 .87 -1.21 -0.41 -0.07 0 . 04 -4 . 82 -1. 99 -0 .78 -0.36 -0.09 -0.01 0 . 01 D8S507(4cM) -1. 68 - 0. 38 0 . 15 0.28 0.28 0.19 0. 06 -2. 28 - 0 . 97 -0 .38 -0.18 -0.04 0.01 0. 01 D8S260(3.4cM) -4 . 68 -3. 01 -1 .89 -1.23 -0.6 -0.31 -0 .13 -5. 28 -3. 35 -1 .69 -0.97 -0.4 -0.16 -0 .05 D8S1797(7cM) -2. 21 -2 .1 -1 .53 -0.94 -0.33 -0.08 0 . 01 -2. 81 -2. 66 -1 .61 -0.92 -0.34 -0.11 -0 .02 D8S543(5.8cM) -1 .6 -1. 28 -0 .27 0.16 0.39 0.35 0 .2 -2 .2 -1. 87 - c .8 -0.32 0.04 0.13 0.1 D8S279(4.3cM) -6 .1 -3. 29 -1 .49 -0.8 -0.3 -0.14 -0 .07 -6. 25 -3. 53 -1 .81 -1.07 -0.45 -0.19 -0 .07 D8S286 -5 .5 -2 .2 -0 .91 -0.43 -0.07 0.05 0 . 05 -5 .5 -2. 19 -0 .91 -0.43 -0.06 0.05 0 . 05 166 FAMILY 3, DYT7 LOCUS. Recombination values (0) Marker 0 0.01 0.05 0.1 0.2 0.3 0.4 D18S59(12.9cM) -4.33 -2.73 -1.94 -1.29 -0.51 -0.16 -0.01 o o 0 0 -4.93 -3.18 -1.71 -0.97 -0.33 1 0 D18S1140(3.2cM) 0.24 0.22 0.16 0.09 -0.03 -0.07 -0.05 0.24 0.22 0.16 0.09 -0.03 -0.07 -0.05 D18S1368 (4cM) -2.1 -0.53 0.05 0.21 0.23 0.16 0.08 -2.1 -0.53 0.05 0.21 0.23 0.16 0.08 D18S1098(1.2 cM) -1.6 -1.6 -1.25 -0.71 -0.19 0.01 0.05 -2.2 -1.81 -0.75 -0.32 0 0.08 0.07 i D18S481(1.3cM) -0.12 -0.12 -0.11 o o -0.06 -0.03 -0.01 o o CO -0.12 -0.12 -0.11 -0.09 -0.06 1 -0.01 D18S54(lcM) -1.42 -1.31 -0. 61 -0.23 0 0 -0.04 -1.7 -0.51 0.05 0.2 0.19 0.08 -0.02 D18S52 (12.8 cM) -4.05 -2.45 -1.73 -1.16 -0.48 -0.17 -0.03 -4.65 -2.92 -1.54 -0.86 -0.31 -0.1 -0.02 D18S62(OcM) -0.07 -0.08 -0.1 -0.11 -0.09 -0.05 -0.02 -0.06 -0.07 -0.1 -0.11 -0.09 -0.05 -0.02 D18S452(7.3cM) -7.69 -4.0 -2.63 -1.82 -0.78 -0.25 -0.03 -8.09 -3.44 -1.53 -0.8 -0.25 -0.06 0.01 D18S1163(3.4cM) -0.04 -0.05 -0.05 -0.05 -0.04 -0.02 -0.01 -0.04 -0.05 -0.05 -0.05 -0.04 -0.02 -0.01 D18S843 -1.42 -1.22 -0.7 -0.36 -0.06 0.04 0.05 -1.73 -1.38 -0.73 -0.37 -0.07 0.04 0.05 167 APPENDIX FOUR RESULTS OF GENOME-WIDE SEARCH FOR THE DYT4 LOCUS The two-point LOD scores for 137 microsatellite markers analysed in the Australian PTD family (family 3, chapter 5) are presented here. These were analysed as part of a genome-wide search for the DYT4 locus in this family. Two sets of LOD scores were calculated for each marker: with individuals V-24 and V-25 with Wilson’s disease and dystonia designated as at-risk for PTD in the upper row for each marker, and with these individuals designated as affected by PTD in the lower row. For selected markers on chromosome 19, a third analysis was performed, with these individuals excluded. These results are presented in the third row of scores for chromosome 19 markers. Parameters used for linkage analysis are given in section 5.4. Markers are arranged in chromosome order and are ordered from p-ter to q-ter. The markers indicated by asterisks were analysed by Dr T. Warner and Mr L. Williams. Genetic distances in centimorgans (cM) between markers are indicated in the chromosome 19 panel. Map positions and distances are from the Genethon map (Dib et al, 1996). CHROMOSOME 1 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D1S243* -3.13 -3.13 -3.09 -2.24 -1.48 -0.72 -0.43 -0.12 CO o i— 1 -3.13 i -3.09 -2.07 -1.43 -0.75 -0.37 -0.09 00 CN oo D1S214* — oo -7.55 -4 . 99 1 -1.91 -0.93 -0.42 -0.14 — oo -6.57 -4.66 -3.24 -2.25 -1.08 -0.48 -0.15 D1S288* -4 .05 -4.04 -3.66 -2.23 -1.46 -0.73 -0.34 -0.12 — 00 -6.45 -4.02 -2.03 -1.24 -0.57 -0.27 -0.1 D1S234 -6.55 -3.93 -2 -0.68 -0.19 0.12 0.13 0.03 -7.23 -5.64 -4.60 -2.94 -1.86 -0.86 -0.38 -0.13 o o D1S255 -3.03 -1.99 -1.02 -0.32 i 0.13 0.12 0.04 -6.95 -5.91 -4 .54 -2.6 -1.66 -0.75 -0.29 -0.07 168 CHROMOSOME 1 (CONTINUED) Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D1S2 0 9* -3.37 -2.11 -1.12 -0.43 -0.16 0.04 0.09 0.06 — OO -4.39 -2.42 -1.06 -0.53 -0.11 0.03 0.05 D1S216 -7.34 -4.08 -2.11 -0.78 -0.27 0.09 0.15 0.09 -7 .34 -6.01 -4.92 -2.77 -1.64 -0.62 -0.17 0.01 D1S207* 1.34 1.33 1.32 1.23 1.10 0.80 0.47 0.15 — o o -4.39 -2.40 -1.07 -0.57 -0.19 -0.06 -0.01 D1S435 — > 0 0 -5.31 -3.30 -1.84 -1.18 -0.54 -0.21 -0.05 — > 0 0 -7.58 -4.60 -2.47 -1.56 -0.69 -0.27 -0.06 D1S248 -7.19 -4.50 -2.5 -1.20 -0.66 -0.20 -0.01 0.04 — 0 0 -6.76 -3.84 -1.83 -1.04 -0.36 0.07 0.02 D1S303* -0.07 -0.07 -0.06 -0.04 -0.02 -0.01 0 0 -0.34 -0.34 -0.32 -0.27 -0.20 -0.11 -0.04 -0.01 D1S196* 0.18 0.17 0.17 0.14 0.10 0.05 0.02 0.00 0.18 0.17 0.17 0.14 0.10 0.05 0.02 0.00 D1S238* -4 .18 -2.62 -1.62 -0.89 -0.54 -0.26 -0.10 -0.02 -3.03 -3.02 -2.67 -1.52 -0.94 -0.41 -0.16 -0.03 D1S249* -3.56 -2.61 -1.66 -0.95 -0.63 -0.35 -0.19 -0.09 -3.88 -3.03 -2.68 -1.52 -0.95 -0.42 -0.16 -0.06 D1S229* 0.17 0.17 0.16 0.13 0.10 0.05 0.01 0.00 0.18 0.18 0.17 0.14 0.11 0.06 0.02 0.00 D1S304 * -7.10 -4 .72 -2.76 -1.37 -0.80 -0.28 -0.05 0.03 -7.10 -5.79 -3.99 -2.01 -1.18 -0.44 -0.11 -0.01 169 CHROMOSOME 2 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 CO o 00 D2S205 i -1.9 -0.93 -0.27 -0.02 0.12 0.11 0.05 — OO -7.02 -4.06 -2.03 -1.21 -0.49 -0.17 -0.02 D2S131 -7.29 -4.57 -2.62 -1.29 -0.76 -0.31 -0.11 -0.02 -6.72 -4.36 -2.43 -1.11 -0.6 -0.2 -0.05 -0 o i—i in D2S177* -3.11 -2.07 -1.1 -0.42 i 0.02 0.04 0.01 — o o -7.47 -4.53 -2.43 -1.52 -0.67 -0.26 -0.06 CM O i — 1 D2S139* -0.19 -0.19 -0.18 1 -0.06 0.02 0.06 0.05 0.96 0.96 0.95 0.90 0.82 0.62 0.39 0.15 O i—1 D2S160 -3.8 -2.38 -1.4 -0.79 -0.44 1 -0.07 -0.02 -3.8 -2.38 -1.4 -0.79 -0.44 -0.19 -0.07 -0.02 D2S114* — o o -2.01 -1.02 -0.35 -0.11 0.41 0.05 0.02 o o i — — o o i -4.29 -2.33 -0.99 -0.47 -0.12 i 0.01 D2S141 -7.05 -4.87 -2.92 -1.5 -0.88 -0.31 -0.06 0.03 -5.89 -4.58 -3.54 -2.1 -1.25 -0.47 -0.12 0.01 D2S152 -3.19 -0.96 0.01 0.58 0.71 0.65 0.43 0.17 -3.19 -2.92 -1.28 -0.05 0.33 0.49 0.37 0.16 m 00 o CM ~ r <— i i D2S155 -3.67 -3.65 -2.86 i -1.03 -0.53 -0.11 — o o -5.65 -4.1 -2.21 -1.41 -0.69 -0.33 -0.12 D2S159 -4.96 -3.92 -1.69 -1.02 -1.02 -0.41 -0.14 -0.03 o m \ — i -3.81 -2.77 -1.79 -1.03 -0.68 -0.33 i -0.04 D2S125 -2.67 -1.63 -0. 67 -0.03 0.19 0.28 0.20 0.04 o 1 00 — — o o 1 -3.92 -1.98 -0.66 1 0.13 0.14 0.03 170 CHROMOSOME 3 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D3S1279* 0.45 0.45 0.44 0.39 0.34 0.25 0.16 0.08 0.45 0.45 0.44 0.40 0.35 0.25 0.16 0.08 D3S1263 -7.22 -5.31 -3.42 -1.94 -1.23 -0.55 -0.21 -0.05 -6.64 -5.22 -3.89 -2.05 -1.21 -0.48 -0.16 -0.03 D3S1277* -3.58 -3.06 -2.68 -1.52 -0.95 -0.42 -0.16 -0.08 -3.58 -3.06 -2.68 -1.52 -0.95 -0.42 -0.16 -0.08 D3S1300 — OO -6.63 -3.68 -1.69 -0.91 -0.27 -0.03 0.03 — o o i CO U) -4.97 -2.32 -1.29 -0.43 -0.09 0.02 D3S12 61* -3.38 -2.34 -1.38 -0.72 -0.5 -0.25 -0.15 -0.08 -2.23 -1.19 -0.25 0.31 0.46 0.43 0.27 0.07 D3S1271* -3.7 -2.28 -1.29 -0.59 -0.29 -0.04 0.051 0.05 -3.7 -2.27 -1.06 -0.4 -0.13 0.06 0.09 0.06 D3S1303* -6.87 -5.04 -3.11 -1.71 -1.1 -0.54 -0.26 -0.1 O CO CM — OO -7.32 -4.41 -2.34 -1.48 -0.7 1 -0.12 D3S1279 0.50 0.50 0.49 0.44 0.38 0.26 0.17 0.08 0.50 0.50 0.49 0.44 0.38 0.26 0.17 0.08 D3S1282 -6.81 -3.53 -1.59 -0.31 0.13 0.38 0.33 0.15 —> o o -9.24 -5.3 -2.59 -1.5 -0.56 -0.15 0.01 D3S1262* 0.47 0.47 0.47 0.43 0.38 0.26 0.13 0.03 0.47 0.47 0.47 0.43 0.38 0.26 0.13 0.03 D3S1265 — o o -7.01 -4.31 -2.3 -1.47 -0.71 -0.34 -0.12 — o o -8.87 -5.62 -2.93 -1.84 -0.87 -0.4 -0.14 171 CHROMOSOME 4 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D4S412* -3.9 -3.84 -2.76 -1.41 -0.85 -0.35 -0.13 -0.02 — OO -4 .72 -2.74 -1.36 -0.8 -0.32 -0.11 -0.02 D4S403 — o o -3.35 -1.4 -0.16 0.24 0.43 0.34 0.15 — o o -8.07 -4.81 -2.34 -1.32 -0.46 -0.11 0.02 D4S418* — o o -1.26 -0.28 0.33 0.51 0.51 0.35 0.14 — o o -5.24 -3.81 -1.95 -1.13 -0.41 -0.1 -0.02 D4S398* -7.05 -3.94 -2.01 -0.72 -0.27 0.02 0.04 -0.01 o o i — 1 -7.05 -5. 61 -3.29 -1.36 -0. 64 -0.33 i -0.02 o CO o D4S395* -3.64 -2.89 -1.96 -1.2 -0.82 -0.39 -0.15 1 — 0 0 -1.83 -0.84 -0.19 0.02 0.12 0.07 0.02 D4S414* -4 .33 -3.28 -2.3 -1.51 -1.11 -0.63 -0.33 -0.12 -3.75 -2.71 -1.73 -1 -0.67 -0.36 -0.2 -0.09 D4S407 -3.45 -2.13 -1.15 -0.47 -0.21 -0.01 0.02 0.01 — OO -4 . 93 -2.94 -1.55 -0. 97 -0.43 -0.17 -0.04 i o D4S422* -3.71 -2.47 -1.5 -0.82 -0.52 -0.24 o CD -0.02 o t—1 00 O O CD 1 -3.66 -2.42 -1.44 -0.75 -0.45 1 -0.01 i D4S413* -6.87 -4.79 -2.9 -1.52 -0.93 -0.4 -0.15 o o u> CM CM "3* -6.29 1 -2.34 -1.01 -0.5 -0.11 0 0.01 D1S415* -5.12 -4.04 -2.81 -1.61 -1.03 -0.46 -0.19 -0.05 -5.09 -3.75 -2.05 -0.73 -0.22 0.14 0.176 0.07 172 CHROMOSOME 5 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D5S392 -7.36 -5.83 -3.67 -1.66 -0.86 -0.21 0.01 0.05 -7.36 -5.89 -4.33 -2.33 -1.43 -0.57 -0.16 0 D5S419 -7.24 -6.2 -5.18 -3.27 -2.12 -1.01 -0.46 -0.15 -7 .24 -6.2 -5.19 -3.39 -2.22 -1.08 -0.49 -0.16 D5S407* -3.13 -2.68 -1.83 -1.05 -0.68 -0.3 -0.12 -0.03 -1.98 -1.53 -0.71 -0.09 0.12 0.2 0.14 0.04 D5S428 — o o -2.66 -1.83 -1.04 -0.67 -0.31 -0.12 -0.03 — o o -4.92 -3.1 -1.67 -1.05 -0.46 -0.18 -0.04 D5S410* -0.15 -0.15 -0.15 -0.16 -0.13 -0.07 -0.03 -0.01 -0.15 -0.15 -0.15 -0.16 -0.16 -0.17 -0.05 -0.01 D5S400 -7.05 -5.61 -3.28 -1.36 -0.65 -0.16 -0.03 -0.03 -7.05 -5.73 -4.76 -3.47 -2.24 -1.05 -0.48 -0.16 D5S408* -3.61 -3.6 -3.36 -2.18 -1.48 -0.75 -0.36 -0.13 — o o -6.14 -4.29 -2.95 -1.45 -0.68 -0.31 -0.11 173 CHROMOSOME 6 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D6S309* -7.18 -4.35 -2.39 -1.06 -0.55 -0.17 -0.05 -0.01 — OO -7.69 -4.69 -2.52 -1.57 -0.67 -0.25 -0.06 D6S276* -6.77 -4.38 -2.44 -1.1 -0.57 -0.16 -0.03 -0.01 -7 .22 -5. 97 -4.49 -2.52 -1.61 -0.74 -0.31 -0.09 D6S282* — oo -3.73 -1.77 -0.52 -0.1 0.13 0.11 0.03 oo CO — oo 1 -5.13 -2.66 -1.62 -0.7 -0.27 -0.06 D6S286* -3.15 -2.38 -1.46 -0.8 -0.54 -0.31 -0.19 -0.09 -3.15 -2.38 -1.46 -0.8 -0.54 -0.31 -0.19 -0.09 00 CO 1—1 D6S262* -6.94 -5.12 -3.23 1 -1.23 -0.64 -0.32 -0.12 -6. 94 -5.12 -3.23 -1.83 -1.23 -0.64 -0.32 -0.12 D6S305 — oo -7.02 -4.2 -2.16 -1.31 -0.56 -0.21 -0.05 VO CO CM — oo 1 -3.86 -1.83 -1.02 -0.36 -0.1 -0.02 D6S281 -2.96 -2.94 -2.27 -1.04 -0.54 -0.15 -0.02 -0.01 -5.82 -3.71 -1.8 -0.53 -0.09 0.17 0.17 0.06 174 CHROMOSOME 7 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D7S531 -3.24 -1.01 -0.03 0.54 0. 67 0.62 0.41 0.16 -3.24 -2.97 -1.33 -0.1 0.29 0.46 0.35 0.15 D7S493 -7.05 -4.47 -2.54 -1.21 -0.7 -0.29 -0.14 -0.07 o t — I -7 .05 -4.47 -2.54 -1.21 -0.7 -0.29 i -0.07 D7S484 — o o -6.74 -3.81 -1.79 -0.99 -0.36 -0.13 -0.06 — o o -10.4 -7.16 -3.99 -2.55 -1.21 -0.54 -0.17 D7S519* -3.7 -2.32 -1.34 -0. 66 -0.39 -0.16 -0.06 -0.01 o o LO -3.5 -2.01 -1.04 -0.42 -0.2 1 -0.01 0 D7S524 1.03 1.03 1.02 0.96 0.88 0.67 0.41 0.14 o ro i—i -3.6 -3.57 -2.65 -1.33 -0.78 i -0.11 -0.02 D7S518 * -7.36 -6.28 -4.36 -2.32 -1.45 -0.66 -0.3 -0.11 O UD — OO -8.29 -4.36 -2.32 -1.45 1 -0.3 -0.11 D7S483 0.15 0.15 0.14 0.12 0.09 0.05 0.02 0 0.15 0.15 0.14 0.12 0.09 0.05 0.02 0 175 CHROMOSOME 8 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 O O CM D8S264 -2.9 -1.69 -0.71 1 0.24 0.36 0.28 0.1 O CM i —i -3.54 -3.54 -3.49 -2.29 -1.4 -0.58 1 -0.04 D8S282 0.96 0.96 0. 94 0.88 0.78 0.55 0.3 0.08 o 0 1 — OO i -4.54 -2.75 -1.41 -0.87 o CO 00 -0.15 o VO 1 1 o — — 1 1 D8S260 1 1 -3.99 -2.08 -0.8 -0.35 0.02 0.01 O o 1 "3* i—i — 1 — oo 1 1 -4.86 -2.9 -1.52 -0.15 -0.03 D8S257* -3.59 -2.06 -1.08 -0.42 -0.17 0.02 0.07 0.06 -3.59 -2.06 -1.08 -0.42 -0.17 0.02 0.07 0.06 o - r D8S272* 1 -2.94 -1.92 -1.09 -0.69 -0.3 -0.12 -0.03 o co * 1 — OO i -5.23 -3.22 -1.72 i— o -0.18 -0.04 CHROMOSOME 9 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D9S178 -7.16 -5 -3.09 -1.72 -1.14 -0.59 -0.29 -0.11 -7.16 -5 -3.09 -1.72 -1.14 -0.59 -0.29 -0.11 i D9S161* -7.23 -4.99 -3.08 -1.71 -1.13 o cn CD -0.29 -0.11 -7.23 -4.99 -3.08 -1.71 -1.13 -0.59 -0.29 -0.11 D9S152 * 0.87 0.87 0.85 0.76 0.65 0.45 0.26 0.11 O 1 O CO i—1 CD —1 — OO 1 1 -2.05 -0.4 0.02 0.07 0.06 D9S160 — oo -4.59 -2.62 -1.29 -0.76 -0.33 -0.14 -0.05 — > 0 0 -3.44 -1.5 -0.26 0.14 0.33 0.25 0.08 176 CHROMOSOME 10 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D10S249* -3.24 -2.47 -1.54 -0.88 -0.6 -0.35 -0.2 -0.09 -3.24 -2.47 -1.54 -0.88 -0.6 -0.35 -0.2 -0.09 D10S191* -0.03 -0.28 -0.27 -0.23 -0.18 -0.1 -0.05 -0.01 -0.03 -0.28 -0.27 -0.23 -0.18 -0.1 -0.05 -0.01 D10S201 -3.84 -2.52 -1.52 -0.77 -0.44 -0.15 -0.04 -0.05 -6.8 -4.62 -2.7 -1.34 -0.79 -0.33 -0.12 -0.03 D10S190 -3.87 -2.38 -1.46 -0.81 -0.54 -0.31 -0.19 -0.12 CM CO oo -3.87 1 -1.46 -0.81 -0.54 -0.31 -0.19 -0.12 D10S212* -4.57 -4 .27 -2.59 -1.26 -0.73 -0.29 -0.1 -0.02 -4 -3.7 -2.05 -0.78 -0.32 -0.02 0.036 0.018 CHROMOSOME 11 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D11S929 -6.81 -5.77 -4.73 -3.18 -2.13 -1.05 -0.48 -0.16 — OO -7.46 -4.53 -2.43 -1.54 -0.72 -0.33 -0.12 D11S904* 0.21 0.21 0.21 0.18 0.15 0.08 0.02 0 1.37 1.36 1.34 1.22 1.07 0.75 0.43 0.13 D11S916 -3.1 -2.33 -1.41 -0.75 -0.5 -0.28 -0.17 -0.08 -3.1 -2.33 -1.41 -0.75 -0.5 -0.28 -0.17 -0.08 D11S898* 0.31 0.31 0.3 0.26 0.21 0.12 0.05 0.01 0.31 0.31 0.3 0.26 0.21 0.12 0.05 0.01 D11S910 0.25 0.25 0.25 0.24 0.21 0.13 0.05 0 CM o CO -3.55 1 -1.05 -0.4 -0.16 -0.02 -0.01 -0.02 177 CHROMOSOME 12 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D12S94 -3.49 -2.22 -1.23 -0.51 -0.21 0.013 0.052 0.022 — CO i - j o -7.23 -4.77 -2.61 -1.63 -0.72 -0.28 o i o LO — i D12S87* -3.7 -2.04 i -0.38 -0.13 0.03 0.05 0.03 -7.45 -4 .73 -2.77 -1.39 -0.82 -0.31 -0.1 -0.01 O CM CM D12S82* — o o -2.14 -1.14 -0.47 1 -0.05 -0.01 -0.01 O CM i — — o o 1 -4.43 -2.44 -1.1 -0.59 i -0.07 -0.02 D12S8 6* -7.12 -4.39 -2.44 -1.1 -0.59 -0.19 -0.04 -0 — o o -6.67 -3.73 -1.74 -0. 97 -0.34 -0.1 -0.02 CHROMOSOME 13 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 o o CO D13S175 -2.32 -2.32 -2.16 -1.23 -0.73 -0.3 -0.12 1 — OO -3.37 -1.45 -0.22 -0.18 -0.34 -0.25 -0.08 LO CD i—1 1 D13S170* -3.92 -2.72 -0.55 -0.16 0.08 0.09 0.04 -2.77 -1.59 -0.63 -0.01 0.17 0.21 0.13 0.04 CO OO CD D13S173* 1 -2.19 -1.21 -0.55 -0.3 -0.11 -0.04 -0.01 -3.86 -2.19 -1.21 -0.55 -0.3 -0.11 -0.04 -0.01 178 CHROMOSOME 14 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D14S72* -3.13 -2.36 -1.44 -0.78 -0.52 -0.29 -0.17 -0.09 -3.13 -2.36 -1.44 -0.78 -0.52 -0.29 -0.17 -0.09 D14S80 0.25 0.24 0.24 0.21 0.17 0.11 0.07 0.03 0.25 0.24 0.24 0.21 0.17 0.11 0.07 0.03 D14S75* 0.24 0.24 0.23 0.19 0.15 0.08 0.03 0 0.24 0.24 0.23 0.2 0.15 0.08 0.03 0 D14S63 -4.45 -2.96 -1.95 -1.17 -0.78 -0.34 -0.15 -0.03 — OO -5.25 -3.25 -1.8 -1.16 -0.53 -0.21 -0.05 D14S81* 1.13 1.13 1.12 1.05 0.95 0.71 0.42 0.13 — o o -4 .27 -2.56 -1.24 -0.71 -0.28 -0.11 -0.03 CHROMOSOME 15 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D15S128 — o o -4.96 -3 -1.61 -1.04 -0.52 -0.27 -0.11 — OO - 1 . . 2 2 -4.29 -2.25 -1.41 -0.68 -0.33 -0.12 D15S121 — o o -2.03 -1.05 -0.43 -0.21 -0.07 -0.02 0 — o o -5.08 -3.76 -1.98 -1.19 -0.5 -0.19 -0.04 D15S205 -3.06 -1.88 -0.91 -0.25 -0.01 0.119 0.09 0.02 -3.61 -3.5 -2.19 -0.88 -0.39 -0.04 0.03 0.006 o o i i — D15S120* -2.89 -2.17 -1.2 -0.54 -0.29 -0.08 i 0 -2.89 -2.17 -1.2 -0.54 -0.29 -0.08 -0.01 0 179 CHROMOSOME 16 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D16S423* -3.63 -3.59 -2.70 -1.38 -0.82 -0.33 -0.12 -0.02 -3.63 -3.59 -2.68 -1.37 -0.83 -0.35 -0.14 -0.03 D16S420* -7.03 -4.06 -2.11 -0.61 -0.34 -0.01 0.05 0.02 — o o -6.34 -3.4 -1.44 -0.71 -0.17 0 0.01 D16S413 0.278 0.277 0.269 0.233 0.192 0.124 0.07 0.031 0.278 0.277 0.269 0.233 0.192 0.124 0.07 0.03 CHROMOSOME 17 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D17S786 — > 0 0 -7.45 -4.69 -2.62 -1.72 -0.86 -0.41 -0.14 — OO -9.59 -5.97 -3.25 -2.1 -1.02 -0.47 -0.16 D17S783* -4.08 -3.94 -2.53 -1.2 -0.68 -0.26 -0.12 -0.06 — o o -6.06 -4.74 -2.82 -1.83 -0.91 -0.42 -0.14 D17S798 0.14 0.14 0.13 0.11 0.08 0.03 0.01 0 0.14 0.14 0.13 0.11 0.08 0.03 0.01 0 D17S794* -7.33 -5.11 -3.15 -1.67 -1.02 -0.43 -0.16 -0.03 — o o -6.85 -3.92 -1.92 -1.14 -0.47 -0.18 -0.04 D17S784* -6.56 -3.75 -1.8 -0.53 -0.08 0.182 0.173 0.068 — o o -7.46 -4.91 -2.6 -1.58 -0.66 -0.25 -0.05 180 CHROMOSOME 18 Recombina tio n v< alues (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 o 0 — OO i 1 D18S59* -4.91 -2.94 -1.54 -0.95 o ho -0.16 CM oo CM 1—1 0 o LO 1 1 — oo -7.19 -1.33 -0.57 -0.22 1 o o t—1 D18S57* -2.69 -1.56 -0.61 1 0.17 0.22 0.15 0 . 05 o -sT r" 0 o I — oo -4.94 -3.19 -1.73 -1.07 -0.18 1 o o CO CM i D18S70* 1 -2.2 -1.24 -0.56 -0.28 0.05 0.05 o CM o CO 1 1 i o hO -2.2 -1.24 -0.56 oo 0.05 0 . 05 CHROMOSOME 19 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 o 0 CO D19S209* -3.75 -3.7 -2.75 -1.43 -0.87 -0.39 -0.19 1 ( 3 4 cM) — O O -5.73 -4.43 -2.6 -1.71 -0.85 -0.4 -0.14 D19S410 -5.95 -3.59 -1.67 -0.39 0.06 0.32 0.27 0.09 (8.2cM) — 00 -5.88 -2.97 -1.03 -0.32 0.16 0.21 0.08 D19S414 0.92 0.91 0.89 0.8 0.68 0.44 0.22 0.06 (2.7cM) 1.77 1.76 1.73 1.58 1.38 0.99 0.59 0.21 1.17 1.16 1.14 1.02 0.87 0.58 0.31 0.09 D19S225 2.66 2.65 2.6 2.35 2.04 1.44 0.87 0.33 (2.2cM) — O O -2.94 -0.99 0.17 0.49 0.53 0.35 0.13 2.46 2.46 2.4 2.17 1.88 1.32 0.79 0.29 D19S416 -8.95 -5.62 -2.76 -0.86 -0.19 0.23 0.25 0.1 (lcM) — oo -11.21 -6.35 -3.04 -1.75 -0.68 -0.25 -0.08 -9.14 -5.82 -2.95 -1.04 -0.35 0.11 0.17 0.06 181 CHROMOSOME 19 (CONTINUED) Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D19S213 -1.94 -0.9 0.03 0.6 0.73 0.66 0.44 0.15 i (2.8 cM) -7.43 -5.93 -3.54 -1.58 o 00 U) -0.25 -0.07 -0.04 o o Os] 1 — 1 1 -1.09 -0.16 0.42 0.57 0.54 0.36 0.11 D19S224 -5.13 -2.93 -1.04 0.13 0.46 0.52 0.33 0.08 o o i— OO i (1.6cM) — -5.73 -2.84 -0.96 -0.32 0.06 0.08 i -5.23 -3.03 -1.14 0.04 0.38 0.46 0.29 0.07 D19S422 2.22 2.21 2.16 1.96 1.7 1.19 0.69 0.23 (OcM) — o o -0.16 0.79 1.26 1.27 1 0.62 0.22 2.24 2.24 2.19 1.98 1.71 1.2 0.7 0.24 D19S417 0.69 0.69 0.69 0.68 0.62 0.54 0.36 0.05 (1.lcM) 0.83 0.82 0.81 0.74 0.65 0.46 0.25 0.07 D19S223 -5.22 -3.05 -1.16 0.06 0.43 0.55 0.4 0.15 (3.5cM) -10.63 -5.42 -2.53 -0.64 0 0.37 0.33 0.13 -5.2 -3.02 -1.13 0.08 0.45 0.56 0.41 0.15 D19S408 -3.44 -2.53 -1.59 -0.92 -0.64 -0.37 -0.21 -0.1 (2.8cM) -3.44 -2.53 -1.59 -0.92 -0. 64 -0.37 -0.21 -0.1 -3.44 -2.53 -1.59 -0.92 -0.64 -0.37 -0.21 -0.1 D19S412 — o o -6.26 -3.36 -1.41 -0.7 -0.18 -0.04 -0.01 (3 5 cM) — o o -5.41 -2.52 -0. 64 0 0.35 0.3 0.1 — o o -6.01 -3.11 -1.19 -0.51 -0.05 0.04 0.01 D19S210* 0.39 0.39 0.38 0.33 0.26 0.19 0.1 0.04 0.39 0.39 0.38 0.33 0.26 0.19 0.1 0.04 182 CHROMOSOME 20 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 CM LO ! — 1 i 1 D20S117 o -3.43 -1.62 -1.1 -0.52 -0.22 -0.07 O O CM 1 i -3.71 -2.39 -1.4 o CO -0.37 -0.12 -0.05 CO CM i t-H — 1 l i i o N> D20S111 -3.24 -3.24 -3.21 -0.62 00 -0.11 CO CM 1 -3.24 -3.25 -2.69 -1.71 -0.77 -0.34 -0.12 CHROMOSOME 21 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D21S265 — O O -4.91 -2.92 -1.51 -0.9 -0.36 -0.13 -0.03 — O O -7.76 -4.74 -2.5 -1.53 -0.65 -0.24 -0.05 D21S268* -4.36 -3.71 -2.75 -1.43 -0.87 -0.39 -0.18 -0.08 — o o -5.73 -4.43 -2.62 -1.71 -0.85 -0.42 -0.2 D21S1259 -3.77 -2.3 -1.32 -0.63 -0.34 -0.1 -0 0.023 -3.11 -1.62 -0.66 -0.07 0.10 0.15 0.10 0.04 CHROMOSOME 22 Recombination values (0) Marker 0 0.001 0.01 0.05 0.1 0.2 0.3 0.4 D22S420* 1.425 1.43 1.40 1.31 1.19 0.88 0.54 0.18 o 00 -3.31 -3.27 -2.28 -0.99 1 -0.11 -0.01 0 D22S446 -3.87 -2.18 -1.18 -0.49 -0.21 -0 0.03 0.01 -3.87 -3.77 -2.46 -1.12 -0.58 -0.16 -0.03 0 r- CM <— 1 D22S275 -7.19 -5 -3.09 i -1.14 -0.59 -0.29 -0.11 -7.19 -5 -3.09 -1.72 -1.14 -0.59 -0.29 -0.11 D22S277* -6.85 -5.38 -3.58 -2.12 -1.44 -0.74 -0.36 -0.13 183 REFERENCES Ahmad F, Davis MB, Waddy HM, Oley CA, Marsden CD, Harding AE. Evidence for locus heterogeneity in autosomal dominant torsion dystonia. Genomics 1993; 15:9-12. Akli S, Chelly J, Mezard C, Gandy S, Kahn A, Poenaru L. A 'G' to 'A' mutation at position -1 of a 5’ splice site in a late infantile form of Tay-Sachs disease. J. Biol. Chem. 1990; 265:7324-30. Albin RL, Young AB, Penney JB. The functional anatomy of basal ganglia disorders. Trends Neurosci 1989; 12:366-75. Almasy L, Bressman SB, Raymond D, Kramer PL, Greene PE, Heiman A, et al. Idiopathic torsion dystonia linked to chromosome 8 in two Mennonite families. Ann. Neurol. 1997; 42:670-3. Alper SL. The band 3-related anion exchanger (AE) gene family. Annu. Rev. Physiol. 1991;53:549-64. Alsager DE, Bowen P, Bamforth JS. Trembling chin - a report of this inheritable dominant character in a four-generation Canadian family. Clin. Genet. 1991; 40:186-9. Antoch MP, Song EJ, Chang AM, Vitatema MH, Zhao Y, Wilsbacher LD, et al. Functional identification of the mouse circadian Clock gene by transgenic BAC rescue. Cell 1997; 89:655-67. Auburger G, Ratzlaff T, Lunkes A, Nelles HW, Leube B, Birikofski F, et al. A gene for autosomal dominant paroxysmal choreoathetosis/spasticity (CSE) maps to the vicinity of a potassium channel gene cluster on chromosome lp, probably within 2 cM between D1S443 and D1 SI97. Genomics 1996; 31:90-4. 184 Auburger G, Leube B, Hendgen T, Goecke T, Kessler K, Benecke R. Focal adult onset idiopathic torsion dystonia is frequently caused by chromosome 18p (DYT7) mutations. J. Neurol 1997; 244:S13 (Abstract) Augood SJ, Penney JB, Friberg IK, Breakefield XO, Young AB, Ozelius LJ, et al. Expression of the early-onset torsion dystonia gene (DYT1) in human brain. Ann. Neurol. 1998; 43:669-73. Bain PG, Findley LJ, Thompson PD, Gresty MA, Rothwell JC, Harding AE, et al. A study of hereditary essential tremor. Brain 1994; 117:805-24. Bandmann O, Nygaard TG, Surtees R, Marsden CD, Wood NW, Harding AE. Dopa- responsive dystonia in British patients: new mutations of the GTP-cyclohydrolase I gene and evidence for genetic heterogeneity. Hum. Mol. Genet. 1996; 5:403-6. Bandmann O, Marsden CD, Wood NW. Genetic aspects of Parkinson's disease. Mov. Disord. 1998; 13:203-11. Baraitser M. The genetics of neurological disorders. 3rd edition. Oxford: Oxford University Press 1997. Becker H. Erbliches Kinnzittem in einer westspanischen Familie. Arch. Psychiatr. Nervenkr. 1982; 231:509-18. Beeson D, Jeremiah S, West LF, Povey S, Newsom-Davis J. Assignment of the human nicotinic acetylcholine receptor genes: the alpha and delta subunit genes to chromosome 2 and the beta subunit gene to chromosome 17. Ann. Hum. Genet. 1990; 54:199-208. Bell JI, Lathrop M. Multiple loci for multiple sclerosis. Nat. Genet. 1996; 13:377-8. Bentivoglio AR, Del Grosso N, Albanese A, Cassetta E, Tonali P, Frontali M. Non- DYT1 dystonia in a large Italian family. J. Neurol. Neurosurg. Psychiatry 1997; 62:357-60. 185 Bezin L, Nygaard TG, Neville JD, Shen H, Levine RA. Reduced lymphoblast neopterin detects GTP cyclohydrolase dysfunction in dopa-responsive dystonia. Neurology 1998; 50:1021-7. Bhatia KP, Marsden CD. The behavioural and motor consequences of focal lesions of the basal ganglia in man. Brain 1994; 117:859-76. Bhatia KP, Soland VL, Bhatt MH, Quinn NP, Marsden CD. Paroxysmal exercise- induced dystonia: eight new sporadic cases and a review of the literature. Mov. Disord. 1997; 12:1007-12. Biary N, Koller WC. Essential tongue tremor. Mov. Disord. 1987; 2:25-9. Bird TD,Bennett RL. Why do DNA testing? Practical and ethical implications of new neurogenetic tests. Ann. Neurol. 1995; 38:141-6. Blaw ME, Leroy RF, Steinberg JB, Herman J. Hereditary quivering chin and REM behavioral disorder. Ann. Neurol. 1989; 26:471 Botstein D, White RL, Skolnick M, Davis RW. Construction of a genetic linkage map in man using restriction fragment length polymorphisms. Am. J. Hum. Genet. 1980; 32:314-31. Bressman SB, Fahn S, Burke RE. Paroxysmal non-kinesigenic dystonia. In: Fahn S, Marsden CD, Caine DB (Eds.). Advances in Neurology, volume 50. Dystonia 2. New York: Raven Press 1988:403-13. Bressman SB, de Leon D, Brin MF, Risch N, Burke RE, Greene PE, et al. Idiopathic dystonia among Ashkenazi Jews: evidence for autosomal dominant inheritance. Ann. Neurol. 1989; 26:612-20. 186 Bressman SB, de Leon D, Kramer PL, Ozelius LJ, Brin MF, Greene PE, et al. Dystonia in Ashkenazi Jews: clinical characterization of a founder mutation. Ann. Neurol. 1994a; 36:771-7. Bressman SB, Heiman GA, Nygaard TG, Ozelius LJ, Hunt AL, Brin MF, et al. A study of idiopathic torsion dystonia in a non-Jewish family: evidence for genetic heterogeneity. Neurology 1994b; 44:283-7. Bressman S, Rassnick H, Almasy L, de Leon D, Bentivoglio AR, Petzinger G, et al. Inheritance of late-onset idiopathic torsion dystonia (ITD). Neurology 1995; 45:A457 (Abstract) Bressman SB, Warner TT, Almasy L, Uitti RJ, Greene PE, Heiman GA, et al. Exclusion of the DYT 1 locus in familial torticollis. Ann Neurol 1996; 40:681-4. Browne DL, Gancher ST, Nutt JG, Brunt ER, Smith EA, Kramer P, et al. Episodic ataxia/myokymia syndrome is associated with point mutations in the human potassium channel gene, KCNA1. Nat. Genet. 1994; 8:136-40. Bums CLC. The treatment of torsion spasm in children with trihexyphenidyl (Artane). The Medical Press 1959; 241:148-9. Byme E, White O, Cook M. Familial dystonic choreoathetosis with myokymia; a sleep responsive disorder. J. Neurol. Neurosurg. Psychiatry 1991; 54:1090-2. Campuzano V, Montermini L, Molto MD, Pianese L, Cossee M, Cavalcanti F, et al. Friereich's ataxia: autosomal recessive disease caused by an intronic GAA triplet repeat expansion. Science 1996; 271:1423-7. Carstea ED, Morris JA, Coleman KG, Loftus SK, Zhang D, Cummings C, et al. Niemann-Pick Cl disease: homology to mediators of cholesterol homeostasis. Science 1997; 277:228-31. 187 Church D, Stoler C, Rutter J, Murrell J, Trofatter J, Buckler A. Isolation of genes from complex sources of mammalian genomic DNA using exon amplification. Nat. Genet. 1994; 6:98-105. Collins FS. Positional cloning: let's not call it reverse anymore. Nat. Genet. 1992; 1:3-6. Collins FS. Positional cloning moves from perditional to traditional. Nat. Genet. 1995; 9:347-50. Comings DE. Powerful new tools in genetics. Am. J. Hum. Genet. 1979; 31:647-8. Comer BD. Dystonia musculorum deformans in siblings. Treated with Artane (trihexyphenidyl). Proceedings of the Royal Society of Medicine 1952; 45:451-2. Cottingham RW, Idury RM, Schaffer AA. Faster sequential genetic linkage computations. Am. J. Hum. Genet. 1993; 53:252-63. Cox DR, Myers RM. A map to the future. Nat. Genet. 1996; 12:117-8. Danek A. Gams M, Gamer C. Erbliches kinnmuskelzittem ('geniospasmus'). Akt. Neurol. 1991; 18:124-7. Danek A. Geniospasm: hereditary chin trembling. Mov. Disord. 1993; 8:335-8. Demirkiran M, Jankovic J. Paroxysmal dyskinesias: clinical features and classification. Ann. Neurol. 1995; 38:571-9. Destee A, Cassim F, Defebvre L, Guieu JD. Hereditary trembling chin or hereditary chin myoclonus? J. Neurol. Neurosurg. Psychiatry. 1997; 63:804-7. Dib C, Faure S, Fizames C, Samson D, Drouot N, Vignal A, et al. A comprehensive genetic map of the human genome based on 5,264 microsatellites. Nature 1996; 380:152-4. 188 DiMauro S, De Vivo DC. Genetic heterogeneity in Leigh syndrome. Ann. Neurol. 1996; 40:5-7. Duclos F, Rodios F, Wrogemann K, Mandel J, Koenig M. The Friedreich ataxia region: characterization of two novel genes and reduction of the critical region to 300kb. Hum. Mol. Genet. 1994; 3:909-14. Duffey P, Butler AG, Barnes M. The epidemiology of the primary dystonias in the North of England. Mov. Disord. 1997; 12:18 (Abstract). Eldridge R. The torsion dystonias: literature review on clinical and genetic studies. Neurology 1970; 20:1-78. Fahn S. High dosage anticholinergic therapy in dystonia. Neurology 1983; 33:1255-61. Fahn S, Marsden CD, Caine DB. Classification and investigation of dystonia. In: Marsden CD, Fahn S (Eds.). Movement Disorders 2. London: Butterworths 1987:332- 58. Fahn S. The paroxysmal dyskinesias. In: Marsden CD, and Fahn S (Eds). Movement disorders 3. Oxford: Butterworth-Heinemann 1994:310-45. Fahn S, Bressman S, Marsden CD. Classification of dystonia. Mov. Disord. 1997; 12:3- 4 (Abstract). Ferrari E, Fredholm BB, Morelli M, Popoli P, Fuxe K. Adenosine-dopamine receptor- receptor interactions as an integrative mechanism in the basal ganglia. Trends Neurosci 1997; 20:482-7. Ferre S, Fuxe K, Von Euler G, Johansson B, Fredholm BB. Adenosine-dopamine interactions in the brain. Neuroscience 1992; 51:501-12. 189 Ferre S, O'Connor WT, Fuxe K, Ungerstedt U. The striopallidal neuron: a main locus for adenosine-dopamine interactions in the brain. J. Neurosci 1993; 13:5402-6. Fields C, Adams MD, White O, Venter JC. How many genes in the human genome? Nat. Genet. 1994; 7:345-6. Fink JK, Rainier S, Wilkowski J, Jones SM, Kume A, Hedera P, et al. Paroxysmal dystonic choreoathetosis: tight linkage to chromosome 2q. Am. J. Hum. Genet. 1996; 59:140-5. Fink JK, Hedera P, Mathay JG, Albin R. Paroxysmal dystonic choreoathetosis linked to chromosome 2q: clinical analysis and proposed pathophysiology. Neurology 1997; 49:177-83. Fletcher NA. The genetics of idiopathic torsion dystonia. J. Med. Genet. 1990a; 27:409-12. Fletcher NA, Harding AE, Marsden CD. A genetic study of idiopathic torsion dystonia in the United Kingdom. Brain 1990b; 113:379-95. Forssman H. Hereditary disorder characterized by attacks of muscular contractions, induced by alcohol amongst other factors. Acta Med. Scand. 1961; 170:517-33. Fouad GT, Servidei S, Durcan S, Bertini E, Ptacek LJ. A gene for familial paroxysmal dyskinesia (FPD1) maps to chromosome 2q. Am. J. Hum. Genet. 1996; 59:135-9. Frey E. Ein streng dominant erbliches kinnmuskelzittem. Dtsch. Z. Nervenheilk 1930; 115:9-26. de la Fuente-Femandez R. Mutations in GTP cyclohydrolase I gene and vitiligo. Lancet 1997; 350:640 190 Furukawa Y, Shimadzu M, Rajput AH, Shimizu Y, Tagawa T, Mori H, et al. GTP- cyclohydrolase I gene mutations in hereditary progressive and dopa-responsive dystonia. Ann. Neurol. 1996; 39:609-17. Furukawa Y, Kish SJ, Bebin EM, Jacobson RD, Fryburg JS, Wilson WG, et al. Dystonia with motor delay in compound heterozygotes for GTP-cyclohydrolase I gene mutations. Ann. Neurol. 1998; 44:10-16. Ganner H,Vonbun A. Erbliches Kinnzittem in einer Tiroler Talschaft. Arch. Rassenbiol. 1935;29:121-32. Gardella R, Belletti L, Zoppi N, Marini D, Barlati S, Colombi M. Identification of two splicing mutations in the collagen type VII gene (COL7A1) of a patient affected by the localista variant of recessive dystrophic epidermolysis bullosa. Am. J. Hum. Genet. 1996; 59:292-300. Gasser T, Bove CM, Ozelius LJ, Hallett M, Chamess ME, Hochberg FH, et al. Haplotype analysis at the DYT1 locus in Ashkenazi Jewish patients with occupational hand dystonia. Mov. Disord. 1996; 11:163-6. Gimenez-Roldan S, Delgado G, Marin M, Villanueva JA, Mateo D. Hereditary torsion dystonia in Gypsies. In: Fahn S, Marsden CD, Caine DB (Eds). Dystonia 2. London: Butterworths 1987: 73-81. Gitschier J, Wood WI, Goralka TM, Wion KL, Chen EY, Eaton DH, et al. Characterization of the human factor VIII gene. Nature 1984; 312:326-30. Goffeau A. DNA technology: molecular fish on chips. Nature 1997: 385;202-3. Goldsmith JB. The inheritance of'facial spasm'. J. Hered. 1927; 18:185-7. Goodenough DJ, Fariello RG, Annis BL, Chun RWN. Familial and acquired paroxysmal dyskinesias. Arch. Neurol. 1978; 35:827-31. 191 Gordon K, Cadera W, Hinton G. Successful treatment of hereditary trembling chin with botulinum toxin. J. Child Neurol. 1993; 8:154-6. Grandchamp B, Picat C, de Rooij F, Beaumont C, Wilson P, Deybach JC, et al. A point mutation G-»A in exon 12 of the porphobilinogen deaminase gene results in exon skipping and is responsible for acute intermittent porphyria. Nucleic Acids Res. 1989; 17:6637-49. Greene P, Kang UJ, Fahn S. Spread of symptoms in idiopathic torsion dystonia. Mov. Disord. 1995; 10:143-52. Grinstein S. Intracellular chloride concentration: determinants and consequences. Prog. Clin. Biol. Res. 1987; 254:31-43. Grossman BJ. Trembling of the chin - an inheritable dominant character. Pediatrics 1957; 19:453-5. Gruol DL, Barker JL, Mae Huang LY, MacDonald JF, Smith TG. Hydrogen ions have multiple effects on the excitability of cultured mammalian neurons. Brain Res. 1980;183:247-52. Gulcher JR, Jonsson P, Kong A, Kristjansson K, Frigge ML, Karason A, et al. Mapping of a familial essential tremor gene, FET1, to chromosome 3ql3. Nat. Genet. 1997; 17:84-7. Gusella J, Wexler NS, Conneally PM, Naylor SL, Anderson SA, Tanzi RE, et al. A polymorphic DNA marker genetically linked to Huntigton's disease. Nature 1983; 306:234-8. Gunel M, Lifton RP. Counting strokes. Nat. Genet. 1996; 13:384-5. 192 Haberhausen G, Schmitt I, Kohler A, Peters U, Rider S, Chelly J, et al. Assignment of the dystonia-parkinsonism syndrome locus, DYT3, to a small region within a 1.8-Mb YAC contig of Xql3.1. Am. J. Hum. Genet. 1995; 57:644-50. Haldane JBS. The combination of linkage values and the calculation of distances between the loci of linked factors. J. Genet. 1919; 8:299-309. Hanihara T, Inoue K, Kawanishi C, Sugiyama N, Miyakawa T, Onishi H, et al. 6- Pyruvoyl-tetrahydropterin synthase deficiency with generalised dystonia and diurnal fluctuation of symptoms: a clinical and molecular study. Mov. Disord. 1997; 12:408-11. Harding AE. The DNA laboratory and neurological practice. J. Neurol. Neurosurg. Psychiatry 1993; 56:229-33. Harding AE. Movement disorders. In: Walton J (Ed.). Brain's diseases of the nervous system. 10th edition. Oxford: Oxford University Press 1993:393-425. Harwood G, Hierons R, Fletcher NA, Marsden CD. Lessons from a remarkable family with dopa-responsive dystonia. J. Neurol. Neurosurg. Psychiatry 1994; 57:460-3. Higgins JJ, Pho LT, Nee LE. A gene (ETM) for essential tremor maps to chromosome 2p22-p25. Mov. Disord. 1997; 12:859-64. Hirano M, Tamaru Y, Nagai Y, Ito H, Imai T, Ueno S. Exon skipping caused by a base substitution at a splice site in the GTP cyclohydrolase I gene in a Japanese family with hereditary progressive dystonia dopa responsive dystonia. Biochem. Biophys. Res. Commun. 1995; 213:645-51. Hirata K, Katayama S, Saito T, Ichihashi K, Mukai T, Katayama M, et al. Paroxysmal kinesigenic choreoathetosis with abnormal electroencephalogram during attacks. Epilepsia 1991; 32:492-4. 193 Hofele K, Benecke R, Auburger G. The gene locus FPD1 of the dystonic Mount-Reback type of autosomal dominant paroxysmal choreoathetosis. Neurology 1997; 49:1252-6. Holmgren G, Ozelius L, Forsgren L, Almay BGN, Holmberg M, Kramer P, Fahn S, et al. Adult onset idiopathic torsion dystonia is excluded from the DYT1 region (9q34) in a Swedish family. J. Neurol. Neurosurg. Psychiatry 1995; 59:178-81. Hubble JP, Busenbark KL, Koller WC. Essential tremor. Clinical Neuropharmacology 1989; 12:453-82. Hudgins RL, Corbin KB. An uncommon seizure disorder: familial paroxysmal choreoathetosis. Brain 1966; 89:199-204. Hudson AJ, Ebers GC, Bulman DE. The skeletal muscle sodium and chloride channel diseases. Brain 1995; 118:547-63. Huntington's disease collaborative research group. A novel gene containing a trinucleotide repeat that is expanded and unstable on Huntington's disease chromosomes. Cell 1993; 72:971-83. Ichinose H, Ohye T, Takahashi E, Seki N, Hori T, Segawa M, et al. Hereditary progressive dystonia with marked diurnal fluctuation caused by mutations in the GTP cyclohydrolase I gene. Nat. Genet. 1994; 8:236-42. Ichinose H, Ohye T, Matsuda Y, Hori T, Blau N, Burlina A, et al. Characterization of mouse and human GTP cyclohydrolase I genes. Mutations in patients with GTP cyclohydrolase I deficiency. J. Biol. Chem. 1995; 270:10062-71. Ichinose H, Ohye T, Segawa M, Nomura Y, Endo K, Tanaka H, et al. GTP cyclohydrolase I gene in hereditary progressive dystonia with marked diurnal fluctuation. Neurosci. Lett. 1995; 196:5-8. 194 The International Batten disease consortium. Isolation of a novel gene underlying Batten disease, CLN3. Cell 1995; 82:949-57. Jain PK, Fukushima K, Deshmukh D, Ramesh A, Thomas E, Lalwani AK, et al. A human recessive neurosensory nonsyndromic hearing impairment locus is a potential homologue of the murine deafness (DN) locus. Hum. Mol. Genet. 1995; 4:2391-4. Jarman PR, Davis MB, Hodgson SV, Marsden CD, Wood NW. Paroxysmal dystonic choreoathetosis: genetic linkage studies in a British family. Brain 1997; 120:2125-30. Jedynak CP, Bonnet AM, Agid Y. Tremor and idiopathic dystonia. Mov. Disord. 1991; 6:230-6. Jin H, May M, Tranebjaerg L, Kendall E, Fontan G, Jackson J, et al. A novel X-linked gene, DDP, shows mutations in families with deafness (DFN-1), dystonia, mental deficiency and blindness. Nat. Genet. 1996; 14:177-80. Johnson LF, Kinsboume M, Renuart AW. Hereditary chin-trembling with nocturnal myoclonus and tongue-biting in dizygous twins. Dev. Med. Child Neurol. 1971; 13:726- 9. Jordan E, Collins FS. A march of genetic maps. Nature 1996; 380:111-2. Kaplan JC, Fontaine B. Neuromuscular disorders: gene location. Neuromusc. Disord. 1997; 7:I-XI. Kato M, Araki S. Paroxysmal kinesigenic choreoathetosis: report of a case relieved by carbamezepine. Arch. Neurol. 1969; 20:508-13. Kato M, Kaga M, Araki S, Kuroiwa Y. Kinesigenic seizure: report of six cases. Clin. Neurol. (Tokyo). 1970; 10:299-304. 195 Kerem B, Rommens JM, Buchanan JA, Markiewicz D, Cox TK, Chakravarti A, et al. Identification of the cystic fibrosis gene: genetic analysis. Science 1989; 245:1073-80. Kertesz A. Paroxysmal kinesigenic choreoathetosis. An entity within the paroxysmal choreoathetosis syndrome. Description of 10 cases including 1 autopsied. Neurology 1967; 17:680-90. Kitada T, Asakawa S, Hattori N, Matsumine H, Yamamura Y, Minoshima S, et al. Mutations in the parkin gene cause autosomal recessive juvenile parkinsonism. Nature 1998;392:605-8. Klein C, Brin MF, de Leon D, Limborska SA, Ivanova-Smolenskaya IA, Bressman SB, et al. De novo mutations (GAG deletion) in the DYT1 gene in two non-Jewish patients with early-onset dystonia. Hum. Mol. Genet. 1998a; 7:1133-6. Klein C, Vierpegge P, Hagenah J, Risch NJ, Breakefield XO, Ozelius LJ. A genetic study of 72 patients with idiopathic focal dystonia in Northern Germany. Neurology 1998b; 50 (suppl 4):A116 (Abstract). Knappskog PM, Flatmark T, Mallet J, Ludecke B, Bartholome K. Recessively inherited L-DOPA-responsive dystonia caused by a point mutation (Q381K) in the tyrosine hydroxylase gene. Hum. Mol. Genet. 1995; 4:1209-12. Kopito RR, Lee BS, Simmons DM, Lindsey AE, Morgans CW, Schneider K. Regulation of intracellular pH by a neuronal homolog of the erythrocyte anion exchanger. Cell 1989;59:927-37. Kosambi DD. The estimation of map distances from recombination values. Ann. Eugen 1944; 12:172-5. Kramer PL, de Leon D, Ozelius L, Risch N, Bressman SB, Brin MF, et al. Dystonia gene in Ashkenazi Jewish population is located on chromosome 9q32-34. Ann. Neurol. 1990; 27:114-20. 196 Kramer PL, Heiman GA, Gasser T, Ozelius LJ, de Leon D, Brin MF, et al. The DYT1 gene on 9q34 is responsible for most cases of early limb-onset idiopathic torsion dystonia in non-Jews. Am. J. Hum. Genet. 1994; 55:468-75. Kruger R, Kuhn W, Muller T, Woitalla D, Graeber M, Kosel S, et al. Ala30Pro mutation in the gene encoding a-synuclein in Parkinson's disease. Nat. Genet. 1998; 18:106-8. Kruglyak L. The use of a genetic map of biallelic markers in linkage studies. Nat. Genet. 1997; 17:21-4. Kurlan R, Behr J, Medved L, Shoulson I. Familial paroxysmal dystonic choreoathetosis: a family study. Mov. Disord. 1987; 2:187-92. Lance JW. Familial paroxysmal dystonic choreoathetosis and its differentiation from related syndromes. Ann. Neurol. 1977; 2:285-93. Landa BL, Reynolds JE, Beeson D, Diehl SR. Dinucleotide repeat polymorphism at the CHRND locus. Hum. Mol. Genet. 1994; 3:1445 Lander E, Kruglyak L. Genetic dissection of complex traits: guidelines for interpreting and reporting linkage results. Nat. Genet. 1995; 11:241-7. Lathrop GM, Lalouel JM. Efficient computations in multilocus linkage analysis. Am. J. Hum. Genet. 1988; 42:498-505. Laurance BM, Matthews WB, Diggle JH. Hereditary quivering of the chin. Arch. Dis. Child 1968;43:249-51. Le Coniat M, Choi DS, Maroteaux L, Launay JM, Berger R. The 5-HT2B gene maps to 2q36.3-2q37.1. Genomics 1996; 32:172-3. 197 Leube B, Rudnicki D, Ratzlaff T, Kessler KR, Benecke R, Auburger G. Idiopathic torsion dystonia: assignment of a gene to chromosome 18p in a German family with adult onset, autosomal dominant inheritance and purely focal distribution. Hum. Mol. Genet. 1996; 5:1673-7. Leube B, Hendgen T, Kessler KR, Knapp M, Benecke R, Auburger G. Sporadic focal dystonia in Northwest Germany: molecular basis on chromosome 18p. Ann. Neurol. 1997a; 42:111-4. Leube B, Kessler KR, Goecke T, Auburger G, Benecke R. Frequency of familial inheritance among 488 index patients with idiopathic focal dystonia and clinical variability in a large family. Mov. Disord. 1997b; 12:1000-6. Ludecke B, Dwomiczak B, Bartholome K. A point mutation in the tyrosine hydroxylase gene associated with Segawa's syndrome. Hum. Genet. 1995; 95:123-5. MacMillan JC, Harper PS. Clinical genetics in neurological disease. J. Neurol. Neurosurg. Psychiatry 1994; 57:7-15. Mansfield DC, Brown AF, Green DK, Carothers AD, Morris SW, Evans HJ, et al. Automation of genetic linkage analysis using fluorescent microsatellite markers. Genomics 1994; 24:225-33. Marsden CD, Harrison MJG. Idiopathic torsion dystonia (dystonia musculorum deformans). A review of 42 patients. Brain 1974; 97:793-810. Marsden CD. Movement disorders. In: Weatherall DJ, Ledingham JGG, Warrell D.A (Eds). Oxford textbook of medicine. 3rd edition. Oxford: Oxford University Press 1996:3998-4022. Massaro D. Vingt-six cas de geniospasme en cinq generations. Revue Neurologique 1894; 2:534. 198 McKusick VA. Mendelian inheritance in man. 11th edition. Baltimore: Johns Hopkins University Press 1994. Mount LA, Reback S. Familial paroxysmal choreoathetosis. Arch. Neurol. Psychiatry 1940; 44:841-7. Mucklow ES, Metz HT. Remarkable response to benzhexol hydrochloride (Artane) in children with an unusual progressive striatal syndrome. Dev. Med. Child Neurol. 1964; 6:598-605. Muller U, Haberhausen G, Wagner T, Fairweather ND, Chelly J, Monaco AP. DXS106 and DXS559 flank the X-linked dystonia-parkinsonism syndrome locus (DYT3). Genomics 1994; 23:114-7. Muller U, Steinberger D, Nemeth AH. Clinical and molecular genetics of primary dystonias. Neurogenetics 1998; 1:165-77. Narcisi TME, Shoulders CC, Chester SA, Read J, Brett DJ, Harrison GB, et al. Mutations of the microsomal triglyceride-transfer-protein gene in abetalipoproteinaemia. Am. J. Hum. Genet. 1995; 57:1298-310. Nardocci N, Lamperti E, Rumi V, Angelini L. Typical and atypical forms of paroxysmal choreoathetosis. Dev. Med. Child Neurol. 1989; 31:670-4. Nehlig A, Daval JL, Debry G. Caffeine and the central nervous system: mechanisms of action, biochemical, metabolic and psychostimulant effects. Brain Res Revs 1992; 17:139-70. Nicholson GA, Dawkins JL, Blair IP, Kennerson ML, Gordon MJ, Cherryson JN, et al. The gene for hereditary sensory neuropathy type I (HSN-I) maps to chromosome 9q22.1-22.3. Nat. Genet. 1996; 13:101-4. 199 Nutt JG, Muenter MD, Aronson A, Kurland LT, Melton LJ. Epidemiology of focal and generalized dystonia in Rochester, Minnesota. Mov. Disord. 1988; 3:188-94. Nygaard TG, Marsden CD, Duvoisin RC. Dopa-responsive dystonia. Advances in Neurology 1988; 50:377-84. Nygaard TG, Marsden CD, Fahn S. Dopa-responsive dystonia: long-term treatment response and prognosis. Neurology 1991; 41:174-81. Nygaard TG, Wilhelmsen KC, Risch NJ, Brown DL, Trugman JM, Gilliam TC, et al. Linkage mapping of dopa-responsive dystonia (DRD) to chromosome 14q. Nat. Genet. 1993; 5:386-91. O'Connell JR, Weeks DE. The VITESSE algorithm for rapid exact multilocus linkage analysis via genotype set-recoding and fuzzy inheritance. Nat. Genet. 1995; 11:402-8. Ophoff RA, Terwindt GM, Vergouwe MN, van Eijk R, Oefner PJ, Hoffman SMG, et al. Familial hemiplegic migraine and episodic ataxia type-2 are caused by mutations in the Ca2+gene CACNL1A4. Cell 1996; 87:543-52. Ott J. Analysis of human genetic linkage. Baltimore: Johns Hopkins University Press 1991. Ott J. Complex traits on the map. Nature 1996; 379:772-3. Ozelius L, Kramer PL, Moskowitz CB, Kwiatkowski DJ, Brin M, Bressman SB, et al. Human gene for torsion dystonia located on chromosome 9q32-q34. Neuron 1989; 2:1427-34. Ozelius LJ, Kramer PL, de Leon D, Risch N, Bressman SB, Schuback DE, et al. Strong allelic association between the torsion dystonia gene (DYT1) and loci on chromosome 9q34 in Ashkenazi Jews. Am. J. Hum. Genet. 1992; 50:619-28. 200 Ozelius LJ, Hewett JW, Page CE, Bressman SB, Kramer PL, Shalish C, et al. The early- onset torsion dystonia gene (DYT1) encodes an ATP-binding protein. Nat. Genet. 1997; 17:40-8. Page CE, Klein C, Ozelius LJ, Gordon MF, Le Witt PA, Brin MF. Genetic analysis of three patients with dystonia and deletion in chromosome 18p. Neurology 1998; 50:A427 (Abstract). Parker N. Hereditary whispering dysphonia. J. Neurol. Neurosurg. Psychiatry 1985; 48:218-24. Pasteris NG, Trask BJ, Sheldon S, Gorski JL. Discordant phenotype of two overlapping deletions involving the PAX3 gene in chromosome 2q35. Hum. Mol. Genet. 1993; 2:953-9. Perez-Boija C, Tassinari CA, Swanson AG. Paroxysmal choreoathetosis and seizures induced by movement (reflex epilepsy). Epilepsia 1967; 8:260-70. Perlmutter JS,.Raichle ME. Pure hemidystonia with basal ganglion abnormalities on positron emission tomography. Ann. Neurol. 1984; 15:228-33. Plant GT, Williams AC, Earl CJ, Marsden CD. Familial paroxysmal dystonia induced by exercise. J. Neurol. Neurosurg. Psychiatry 1984; 47:275-9. Polymeropoulos MH, Lavedan C, Leroy E, Ide SE, Dehejia A, Dutra A, et al. Mutation in the a-synuclein gene identified in families with Parkinson's disease. Science 1997; 276:2045-7. Porkka-Heiskanen T, Strecker RE, Thakkar M, Bjorkum AA, Greene RW, McCarley RW. Adenosine: a mediator of the sleep-inducing effects of prolonged wakefulness. Science 1997; 276:1265-8. 201 Povey S, Attwood J, Chadwick B, Frezal J, Haines JL, Knowles M, et al. Report on the fifth international workshop on chromosome 9. Ann. Hum. Genet. 1997; 61:183-206. Ptacek LJ, George AL, Jr., Griggs RC, Tawil R, Kallen RG, Barchi RL, et al. Identification of a mutation in the gene causing hyperkalemic periodic paralysis. Cell 1991;67:1021-7. Ptacek LJ, Tawil R, Griggs RC, Engel AG, Layzer RB, Kwiecinski H, et al. Dihydropyridine receptor mutations cause hypokalemic periodic paralysis. Cell 1994; 77:863-8. Ptacek LJ. Channelopathies: ion channel disorders of muscle as a paradigm for paroxysmal disorders of the nervous system. Neuromusc. Disord. 1997; 7:250-5. Raskind WH, Bolin T, Wolff J, Fink J, Matsushita M, Litt M, et al. Further localization of a gene for paroxysmal dystonic choreoathetosis to a 5 cM region on chromosome 2q34. Hum. Genet. 1998; 102:93-7. Reed PW, Davies JL, Copeman JB, Bennett ST, Palmer SM, Pritchard LE, et al. Chromosome-specific microsatellite sets for fluorescence-based, semi-automated genome mapping. Nat. Genet. 1994; 7:390-5. Reiner A, Albin RL, Anderson KD, D'Amato CJ, Penney JB. Differential loss of striatal projection neurones in Huntington's disease. Proc. Natl. Acad. Sci. U.S.A. 1988; 85:5733-7. Richards RN, Barnett HJ. Paroxysmal dystonic choreoathetosis. A family study and review of the literature. Neurology 1968; 18:461-9. Risch N, de Leon D, Ozelius L, Kramer P, Almasy L, Singer B, et al. Genetic analysis of idiopathic torsion dystonia in Ashkenazi Jews and their recent descent from a small founder population. Nat. Genet. 1995; 9:152-9. 202 Robson KJ, Chandra T, MacGillivray RT, Woo SL. Polysome immunoprecipitation of phenylalanine hydroxylase mRNA from rat liver and cloning of its cDNA. Proc. Natl. Acad. Sci. U.S.A. 1982; 79:4701-5. Rommens JM, Iannuzzi MC, Kerem B, Drumm ML, Melmer G, Dean M, et al. Identification of the cystic fibrosis gene: chromosome walking and jumping. Science 1989; 245:1059-65. Rosenberg RN. Autosomal dominant cerebellar phenotypes: the genotype has settled the issue. Neurology 1995; 45:1-5. Rouleau GA, Merel P, Lutchman M, Sanson M, Zucman J, Marineau C, et al. Alteration in a new gene encoding a putative membrane-organizing protein causes neuro fibromatosis type 2. Nature 1993; 363:515-21. Royer Pokora B, Kunkel LM, Monaco AP, Goff SC, Newburger PE, Baehner RL, et al. Cloning the gene for an inherited human disorder - chronic granulomatous disease - on the basis of its chromosomal location. Nature 1986; 322:32-8. Rubio JP, Danek A, Stone C, Chalmers R, Wood N, Verellen C, et al. Chorea- acanthocytosis: genetic linkage to chromosome 9q21. Am. J. Hum. Genet. 1997; 61:899-908. Sanger F, Nicklen S, Coulson AR. DNA sequencing with chain-terminating inhibitors. Proc. Natl. Acad. Sci. U.S.A. 1977; 74:5463-7. Schaffer AA, Gupta SK, Shriram K, Cottingham RW. Avoiding recomputation in linkage analysis. Human Heredity 1994; 44:225-37. Scheffer IE, Bhatia KP, Lopes Cendes I, Fish DR, Marsden CD, Andermann E, et al. Autosomal dominant nocturnal frontal lobe epilepsy. A distinctive clinical disorder. Brain 1995; 118:61-73. 203 Scherer SS, Chance PF. Myelin genes: getting the dosage right. Nat. Genet. 1995; 11:226-8. Schloesser DT, Ward TN, Williamson PD. Familial paroxysmal dystonic choreoathetosis revisited. Mov. Disord. 1996; 11:317-20. Schuler GD, Boguski MS, Stewart EA, Stein LD, Gyapay G, Rice K, et al. A gene map of the human genome. Science 1996; 274:540-6. Schwengel DA, Jedlicka AE, Nanthakumar EA, Weber JA, Levitt RC. Comparison of fluorescence-based semi-automated genotyping of multiple microsatellite loci with autoradiographic techniques. Genomics 1994; 22:46-54. Scott D, Carmi R, Elbedour K, Yosefsberg S, Stone E, Sheffield VC. An autosomal recessive non-syndromic hearing loss locus identified by DNA pooling using two inbred bedouin kindreds. Am. J. Hum. Genet. 1996; 59:385-91. Segawa M, Ohmi K, Itoh S, Aoyama M, Hayakawa H. Childhood basal ganglia disease with remarkable response to L-Dopa. Shinryo (Tokyo) 1971; 24:667-72. Segawa M, Hosaka A, Miyagawa F, Nomura Y, Imai H. Hereditary progressive dystonia with marked diurnal variation. In: Eldridge R, Fahn S (Eds). Advances in Neurology. New York: Raven Press 1976:215-33. Sheffield VC, Weber JL, Buetow KH, Murray JC, Even DA, Wiles K, et al. A collection of tri- and tetranucleotide repeat markers used to generate high quality, high resolution human genome-wide linkage maps. Hum. Mol. Genet. 1995; 4:1837-44. Silver LM. Bouncing off microsatellites. Nat. Genet. 1992; 2:8-9. Soland VL, Bhatia KP, Sheean GL, Marsden CD. Hereditary geniospasm: two new families. Mov. Disord. 1996; 11:744-61. 204 Stallings RL, Ford AF, Nelson D, Tomey DC, Hildebrand CE, Moyzis K. Evolution and distribution of (GT)n repetitive sequences in mammalian genomes. Genomics 1991; 10:807-15. Steinberger D, Weber Y, Korinthenberg R, Deuschl G, Benecke R, Martinius J, et al. High penetrance and pronounced variation in expressivity of GCH1 mutations in five families with dopa-responsive dystonia. Ann. Neurol. 1998; 43:634-9. Steinlein OK, Mulley JC, Propping P, Wallace RH, Phillips HA, Sutherland GR, et al. A missense mutation in the neuronal nicotinic acetylcholine receptor alpha 4 subunit is associated with autosomal dominant nocturnal frontal lobe epilepsy. Nat. Genet. 1995; 11:201-3. Stockier S, Isbrandt D, Hanefeld F, Schmidt B, von Figura K. Guanidinoacetate methyltransferase deficiency: the first inborn error of creatine metabolism in man. Am. J. Hum. Genet. 1996; 58:914-22. Stocks P. Facial spasm inherited through four generations. Biometrika 1922; 14:311-5. Strachan T, Read AP. Human molecular genetics. Oxford: Bios scientific publishers 1996. Su YR, Klanke CA, Houseal TW, Linn SC, Burk SE, Varvil TS, et al. Molecular cloning and physical and genetic mapping of the human anion exchanger isoform 3 (SLC2C) gene to chromosome 2q36. Genomics 1994; 22:605-9. Szepetowski P, Rochette J, Berquin P, Piussan C, Lathrop GM, Monaco AP. Familial infantile convulsions and paroxysmal choreoathetosis: a new neurological syndrome linked to the pericentromeric region of human chromosome 16. Am. J. Hum. Genet. 1997;61:889-98. 205 Taylor TD, Litt M, Kramer P, Pandolfo M, Angelini L, Nardocci N, et al. Homozygosity mapping of Hallervorden-Spatz syndrome to chromosome 20pl2.3-pl3. N at Genet. 1996; 14:479-81. Tibbies JA, Barnes SE. Paroxysmal dystonic choreoathetosis of Mount and Reback. Pediatrics 1980; 65:149-51. Uitti RJ, Maraganore DM. Adult onset familial cervical dystonia: report of a family including monozygotic twins. Mov. Disord. 1993; 8:489-94. The Utah marker development group. A collection of ordered tetranucleotide-repeat markers from the human genome. Am. J. Hum. Genet. 1995; 57:619-28. Waddy HM, Fletcher NA, Harding AE, Marsden CD. A genetic study of idiopathic focal dystonias. Ann. Neurol. 1991; 29:320-4. Wadlington WB. Familial trembling of the chin. J. Pediatr. 1958; 53:316-21. Wali GM Paroxysmal hemidystonia induced by prolonged exercise and cold. J. Neurol. Neurosurg. Psychiatry 1992; 55:236-7. Walker ES. Familial paroxysmal dystonic choreoathetosis: a neurologic disorder simulating psychiatric illness. Johns. Hopkins. Med. J. 1981; 148:108-13. Wang DG, Fan JB, Siao CJ, Bemo A, Young P, Sapolsky R, et al. Large-scale identification, mapping and genotyping of single-nucleotide poymorphisms in the human genome. Science 1998; 280:1077-82. Warner TT, Fletcher NA, Davis MB, Ahmad F, Conway D, Feve A, et al. Linkage analysis in British and French families with idiopathic torsion dystonia. Brain 1993; 116:739-44. 206 Weatherall DJ. The new genetics and clinical practice. 2nd edition. Oxford: Oxford University Press 1985. Weber JL, May PE. Abundant class of human DNA polymorphisms which can be typed using the polymerase chain reaction. Am. J. Hum. Genet. 1989; 44:388-96. Weber MB. Familial paroxysmal dystonia. J. Nerv. Ment. Dis. 1967; 145:221-6. Weil D, D'Alessio M, Ramirez F, Steinmann B, Wirtz MK, Glanville RW, et al. Temperature-dependent expression of a collagen splicing defect in the fibroblasts of a patient with Ehers-Danlos syndrome type VII. J. Biol. Chem. 1989; 264:16804-9. Welt C, Abbs JH. Musculotopic organization of the facial motor nucleus in Macaca fascicularis : a morphometric and retrograde tracing study with cholera toxin B-HRP. J. Comp. Neurol. 1990; 291:621 -36. Williams AC. Dopamine, dystonia, and the deficient co-factor. Lancet 1995; 345:1130 Wood N. Genes and parkinsonism. J. Neurol. Neurosurg. Psychiatry 1997; 62:305-9. Xiong M, Guo SW. Fine-scale genetic mapping based on linkage disequilibrium: theory and applications. Am. J. Hum. Genet. 1997; 60:1513-31. Yoshimoto K, McBride W, Lumeng L, Li T. Alcohol stimulates the release of dopamine and serotonin in the nucleus accumbens. Alcohol 1991; 9:17-22. Zeman W, Dyken P. Dystonia musculorum deformans: clinical, genetic and patho- anatomical studies. Psychiatr. Neurol. Neurochir. 1967; 70:77-121. Ziegle JS, Su Y, Corcoran KP, Nie L, Mayrand PE, Hoff LB, et al. Application of automated DNA sizing technology for genotyping microsatellite loci. Genomics 1992; 14:1026-31. 207 Zilber N, Korczyn AD, Kahana E, Fried K, Alter M. Inheritance of idiopathic torsion dystonia among Jews. J. Med. Genet. 1984; 21:13-20. 208 PUBLICATIONS RESULTING FROM WORK PRESENTED IN THIS THESIS Jarman PR, Bandmann O, Marsden CD, Wood NW. GTP cyclohydrolase mutations in patients with dystonia responsive to anticholinergic drugs. Journal of Neurology, Neurosurgery and Psychiatry 1997; 63:304-8. Jarman PR, Davis MB, Hodgson SV, Marsden CD, Wood NW. Paroxysmal dystonic choreoathetosis: genetic linkage studies in a British family. Brain 1997; 120:2125-30. Jarman PR, Wood NW, Davis MT, Davis PV, Bhatia KP, Marsden CD, Davis MB. Hereditary geniospasm: linkage to chromosome 9ql3-q21 and evidence for genetic heterogeneity. American Journal of Human Genetics 1997; 61:928-33. Jarman PR, Warner TT. The dystonias. Journal of Medical Genetics 1998; 35:314-8. Jarman PR, Bhatia KP, Heales S, Tuijanski N, Marsden CD, Wood NW. Paroxysmal dystonic choreoathetosis: clinical features and investigation of pathophysiology in a large family. Submitted to Movement Disorders. Jarman PR, del Grosso N, Valente EM, Leube B, Cassetta E, Bentivoglio AR, et al. Primary torsion dystonia: the search for genes is not over. Submitted to the Journal of Neurology, Neurosurgery and Psychiatry. 209 SELECTED ABSTRACTS Jarman PR, Warner TT, Williams LD, Harding AE. Genetic linkage studies in autosomal dominant torsion dystonia. Journal of Neurology 1995; 242 (Suppl.2):81. Jarman PR, Wood NW, Harding AE. GTP cyclohydrolase I mutations in patients with anticholinergic responsive dystonia. Journal of Medical Genetics 1996; 33 (Suppl 1): 55. Jarman PR, Davis MB, Hodgson SV, Marsden CD, Wood NW. Genetic linkage and candidate gene studies in a British family with paroxysmal dystonic choreoathetosis. Neurology 1997; 48 (Suppl 2):396 Jarman PR, Davis MB, Hodgson SV, Marsden CD, Wood NW. A genetic study of PDC in a British family. Journal of Neurology, Neurosurgery and Psychiatry. 1997;63:263 Awarded Charles Symonds Prize Jarman PR, Wood NW, Davis MT, Davis PV, Bhatia KP, Marsden CD, Davis MB. Linkage of a gene for hereditaiy geniospasm to chromosome 9q. Movement Disorders 1998 (in press). 210