The molecular genetics of Charcot-Marie-Tooth disease

(1)

THE MOLECULAR GENETICS OF

CHARCOT-MARIE-TOOTH DISEASE

PAULA JANE HALLAM

Submitted for the degree of Doctor of Philosophy July 1993

(2)

ProQuest Number: U063389

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 complete 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 U063389

Published by ProQuest LLC(2016). Copyright of the Dissertation is held by the Author.

This work is protected against unauthorized copying under Title 17, United States Code. Microform Edition © ProQuest LLC.

ProQuest LLC

789 East Eisenhower Parkway P.O. Box 1346

(3)

(4)

ACKNOWLEDGEMENTS

This project was supervised by Dr S Malcolm and supported by the M uscular

D ystrophy G roup of G reat Britain and N orthern Ireland.

I w ould like to thank the following people for their invaluable assistance

during this project: Dr H M iddleton-Price for unravelling the mysteries of the

LIPED and LINKAGE packages and P R utland for com puting generally; Dr P

Scambler for the use of the Phospho-Im aging equipm ent and D Kelly for

guiding m e through its application; J Eckersley and Calvin for their patience

w ith the preparation of the figures; all the patients and clinicians w ithout

w hom there w ould have been no project; and all the members of Club 5022

w ho offered support and encouragem ent w hen it was so desperately needed.

Finally, I w ould like to thank Michael for his unw avering support and limitless

(5)

ABSTRACT

Charcot M arie Tooth disease is the commonest of the inherited peripheral

neuropathies, w ith an incidence of 1/2500. The m ost frequent form of

inheritance is one of autosom al dominance. The majority of families show ing

this m ode of inheritance have show n linkage to probes from the pericentric

region of chrom osome 17 although genetic heterogeneity does exist, w ith some

families show ing linkage to chrom osome 1 probes.

Eight large families, six of w hom had previously show n linkage to probes

from the pericentric region of chromosome 17 were exam ined w ith probes

from the short arm of chrom osome 17. Two-point and m ulti-point linkage

analyses w ere used further to localise the genetic location of the disease gene

to the 12cM region betw een probes pl0.41 and EW503 on 17pl 1.2-12. M ulti

point analysis gave a m axim um lod score of 9.97 for the interval betw een these

tw o probes. Cross-overs still rem ained, in different families, betw een the

disease locus and both flanking probes.

Following the identification of sub-microscopic duplication of part of 17pll.2

in patients w ith C M T la from tw o other groups, the families in our study

group w ere analyzed for evidence of this duplication. The duplication was

detected in all eight families, as well as in twenty-seven smaller families

previously too small for conventional linkage analysis and num erous sporadic

cases.

The region of duplication seems consistent in the eight families studied in

detail, including probes VAW409R3 (D17S122), VAW412R3 (D17S125) and

pEW401 (D17S61), a region thought to be approxim ately 1Mb. Cross-overs

still exist w ithin the duplicated region, w ith probes VAW409R3 and pEW401

(VAW412R3 being uninform ative), although there is complete concordance of

the duplication w ith the disease phenotype.

(6)

CONTENTS

Page

TITLE PAGE 1

ABSTRACT 2

ACKNOWLEDGEMENTS 4

LIST OF FIGURES 11

LIST OF TABLES 14

1 .0 INTRODUCTION 16

1.1 CHARCOT-MARIE-TOOTH DISEASE 17

1.2 ANATOMY OF NEURONES 21

1.3 MYELIN 24

1.4 GENE MAPPING APPROACHES 26

1.4.1 Cytogenetic abnorm alities 27

1.4.2 As applied to CMTl 27

1.4.3 M urine hom ology 27

1.4.4 M urine hom ology for CMTl 28

1.4.5 Linkage approach 33

1.4.6 Linkage approach for CMT 33

1.5 GENE MAPPING BY PHYSICAL METHODS 39

1.5.1 Banding of chromosomes 40

1.5.2 Somatic cell hybrids 40

1.5.3 D eletion m apping 41

1.5.4 Translocation m apping 41

1.5.5 In-situ-hybridisation 41

1.5.6 Pulsed-field gel electrophoresis 42

(7)

1.6 GENETIC MAPPING 43

1.6.1 Linkage 43

1.6.2 Genetic m ap distance 43

1.6.3 M apping functions 44

1.6.4 Genetic m arkers 46

1.6.4.1 Restriction Fragm ent Length Polym orphism s 46

1.6.4.2 Minisatellite sequence polym orphism s 46

1.6.4.3 Microsatellites 48

1.6.4.4 Single stranded conform ation polym orphism s 48

D enaturing gradient gel electrophoresis

1.7 LINKAGE ANALYSIS 48

1.7.1 Two-point linkage analysis 50

1.7.2 Levels of significance 59

1.7.3 M ulti-point linkage analysis 60

1.7.4 C om puter program s 61

1.8 CHROMOSOME 17 61

1.8.1 Regions of synteny w ith m ouse chromosome 11 61

1.9 AIMS OF THIS PROJECT 64

2.0 MATERIALS AND METHODS 65

2.1 MATERIALS 66

2.1.1 Family data 66

2.1.2 Probes 67

2.1.3 Chemicals and reagents 67

2.1.4 Enzymes 68

2.1.5 N ylon m em branes 68

2.1.6 Photography, autoradiography and Phospho 68

(8)

2.2 SOLUTIONS 74

2.3 METHODS 76

SOUTHERN BLOTTING OF HUMAN DNA 76

2.3.1 Preparation of hum an DNA from blood 76

2.3.2 Digestion of hum an genomic DNA 76

2.3.3 Electrophoresis of genomic DNA 77

2.3.4 DNA transfer to H ybond-N + 77

2.3.5 H ybridisation of 32-P-labelled DNA to filters 78

PROBE DNA 79

2.3.6 PREPARATION OF PLASMID DNA 79

2.3.7 Transform ation of bacteria w ith plasm id DNA 79

2.3.8 Preparation of glycerol stocks of transform ed

bacteria 80

2.3.9 Isolation and purification of plasm id DNA

2.3.10 Isolation of plasm id insert from vector DNA 81

2.3.11 OLIGO-LABELLING OF PLASMID OR INSERT

DNA 82

2.3.12 Removal of unincorporated nucleotides 82

2.3.13 PRE-ANNEALING OF PROBES 82

2.3.14 Preparation of hum an DNA for pre-annealing

from blood 82

2.3.15 Preparation of hum an DNA for pre-annealing from

placental tissue 82

2.3.16 Pre-annealing probes 84

DATA ANALYSIS 85

2.3.17 Tw o-point linkage analysis 85

DOSAGE ANALYSIS 86

2.3.19 W ith A4spJ-digested DNA 86

(9)

3.0 RESULTS 88

3.1 RFLP data 89

3.2 Tw o-point linkage analysis 108

3.3 M arker order in the centromeric region of

chrom osome 17 119

3.4 Cross-over analysis 120

3.5 M ulti-point linkage analysis 121

3.6 D uplication of 1 7 p ll.2 126

3.6.1 Presence of segm ental trisomy of 1 7 p ll.2 126

3.6.2 H aplotype frequency of alleles in duplicated region 130

3.6.3 Extent of the segm ental trisom y in these families 131

3.6.4 Presence of the duplication as a diagnostic tool

w ithin families 140

3.7 Cross-overs betw een probes within the

duplicated region and C M Tla 142

3.7.1 Failure to recognise the dosage effect and

apparent cross-overs 142

3.7.2 "True" cross-overs w ithin the duplicated region 144

3.8 CMT2 families 152

3.9 Small families and other sam ples 152

3.10 D uplication detection using probe VAW409R3a 154

(10)

3.11 D uplication detection using probes VAW409R3a,

VAW412R3HEb and E3.9 and EcoRI-digested DNA 155

3.11.1 Experimental design 155

3.11.2 Validation of the technique 155

3.11.3 Results of dosage analysis on clinical sam ples 169

Familial cases 195

Sporadic cases 196

Testing of asym ptom atic family m em bers 196

Recessive CMT 197

X-linked CMT 198

CMT2 199

Differential diagnosis 199

Pre-natal diagnosis 200

4.0 DISCUSSION 202

4.1 LINKAGE AND CROSS-OVER ANALYSIS 203

4.1.1 Tw o-point linkage analysis 203

4.1.2 M arker order in the centromeric region of

chrom osom e 17 205

4.1.3 Cross-over analysis 206

4.2 DUPLICATION OF 17pll.2 210

4.2.1 Presence of the duplication in the eight families 210

4.2.2 Extent of the duplication 212

4.3 MECHANISMS CAUSING THE DUPLICATION 214

4.3.1 Cross-overs w ithin the duplication 217

4.4 MECHANISM OF DUPLICATION INDUCING

PATHOLOGICAL EFFECTS 220

(11)

4.6 CMT2 FAMILIES 225

4.7 DUPLICATION DETECTION IN SPORADICS

AND SMALL FAMILIES 226

4.7.1 Applicability of dosage analysis in clinical cases 229

4.8 FUTURE FOR THE STUDY 231

(12)

FIGURES

1.1 PNB m yelination 23

1.2 M olecular organisation in PNB m yelin 25

1.3 Recombination fraction and genetic distance 47

1.4 Gene ordering w ith two-factor crosses 52

1.5 Assignm ent of gene order from two-factor crosses 52

1.6 Use of three-factor crosses to confirm gene order 53

1.7 Pedigree show ing segregation of autosomal

dom inant condition in a three generation family 56

1.8 Ideogram of hum an chromosome 17 62

1.9 Syntenic regions of hum an 17 and m ouse chromosome 11 63

3.1 Pedigree of Family SM 90

3.2 Pedigree of Family MO 92

3.3 Pedigree of Family PE 95

3.4 Pedigree of Family W H 98

3.5 Pedigree of Family GV 100

3.6 Pedigree of Family HED 102

3.7 Pedigree of Family FG 104

3.8 Pedigree of Family HEN 106

3.9 Schematic representation of recom binant chromosomes 122

3.10 M ulti-point linkage analysis including duplicated probes 124

3.11 M ulti-point linkage analysis excluding duplicated probes 125

3.12 Genomic DNA hybridised w ith probe VAW409R3a 128

3.13 A utoradiograph of Family PE DNA hybridised with

probe VAW409R3a 129

3.14 Genomic DNA hybridised w ith probe VAW412R3HEb 133

3.15 Genomic DN A hybridised w ith probe VAW401HE 135

3.16 Genomic DNA hybridised with probe 1041 136

3.17 Genomic DNA hybridised w ith probe VAW411R2 138

3.18 Schematic representation of duplication status

(13)

3.19 Genomic D N A hybridised w ith probe VAW409R3a 141

3.20 Family GV: Genomic DNA hybridised with

probe VAW409R3a 143

3.21 Part of Family MO: show ing cross-over betw een

probe VAW409R3a and the disease locus 145

3.22 Part of Family MO: RFLP results 146

3.23 Family MO, individual III-4 - explanation of

recom bination event 147

3.24 Family SM: RFLP results 148

3.25 Family SM, individual I-l - explanation of

3.26 Family PE: RFLP results 150

3.27 Family PE, individual HI-3 - explanation of

3.28 Genomic DN A from individuals w ith CMT2

hybridised w ith probe VAW409R3a 153

3.29 Genomic DN A hybridised sim ultaneously w ith probes

VAW409R3a, VAW412R3 and E3.9 156

3.30 Regression analysis of Phospho-Imaging control

d ata from Experiments 1-8 168

3.31 Regression analysis of Phospho-Imaging data from

Experim ent 1 173

Experim ent 2 176

3.33 Regression analysis of Phospho-Im aging data from

Experim ent 3 179

Experim ent 4 182

Experim ent 5 185

(14)

Experim ent 7 191

Experim ent 8 194

(15)

TABLES

1.1 Com position of hum an PNS myelin 24

1.2 Lod scores betw een CMTl and Duffy m arker 35

1.3 Lod scores betw een CMTl and chromosome 17 m arkers 37

2.1 Probes used in this study 69

3.1 RFLP data - Family SM 91

3.2 RFLP data - Family MO 93

3.3 RFLP data - Family PE 96

3.4 RFLP data - Family W H 99

3.5 RFLP data - Family GV 101

3.6 RFLP data - Family HED 103

3.7 RFLP data - Family FG 105

3.8 RFLP data - Family HEN 107

3.9 Two-point linkage analysis of CMTl and probe EW206 109

3.10 Two-point linkage analysis of CMTl and probe HHH202 110

3.11 Tw o-point linkage analysis of CMTl and probe EW301 111

3.12 Two-point linkage analysis of CMTl and probe cl516 112

3.13 Tw o-point linkage analysis of CMTl and probe 1041 113

3.14 Tw o-point linkage analysis of CMTl and probe VAW409R3 114

3.15 Tw o-point linkage analysis of CMTl and probe VAW412R3 115

3.16 Two-point linkage analysis of CMTl and probe EW401 116

3.19 H aplotype of MspI alleles w ithin duplication 131 3.20 H aplotypes of MspI alleles in the duplication

detected w ith probe VAW409R3a 131

3.21 Phospho-Im ager results from control sam ples from

Experiments 1-8 160

3.22 Phospho-Im ager results from Experiment 1 171

(16)

(17)

(18)

1.1 CHARCOT-MARIE-TOOTH DISEASE

Peroneal m uscular atrophy was first clearly described by Charcot and Marie,

and Tooth in 1886, although earlier similar cases had been documented. This

neuropathy was characterised by a denervation process which resulted in a

slow progressive w eakness and muscle atrophy. The lower limbs w ere

prim arily affected and this included involvem ent of the peroneal muscle, foot

deform ity and some sensory loss. A familial tendency was noted. Charcot

and M arie proposed a myelopathological basis for the disease, whilst Tooth

favoured a peripheral nerve disturbance.

M uch confusion arose following these original descriptions. Gom bault and

M allet described a patient in 1889 w ith hypertrophic neuropathy and club foot.

In 1893, Dejerine and Sottas show ed similar changes in a brother and sister,

w ith a severe progressive sensory and m otor neuropathy and thickened

peripheral nerves. The presence of nerve hypertrophy became synonym ous

w ith Dejerine - Sottas disease, and it was not until the 1950"s that it was

realized that hypertrophic neuropathy is a non - specific consequence of a

chronic segm ental dem yehnation and rem yelinadon which is seen in a w ide

variety of inherited and acquired disorders. Roussy and Levy added to the

confusion w ith their description in the 1920's of cases similar to those

described by Charcot, M arie and Tooth b u t w ith an additional static trem or of

the hands (H arding and Thomas, 1980 a.).

The classification of peroneal m uscular atrophy and related disorders became

m uch clearer w ith the introduction of nerve conduction studies in the m id -

1950's. Severe slowing of m otor nerve conduction was found in some patients

w ith peroneal m uscular atrophy w hilst other cases failed to show this. Dyck

and Lam bert (1968 a.) found that this phenom enon was consistent w ithin

families. They w ere also able to correlate the changes in conduction w ith

nerve biopsy findings. They w ere thus able to define tw o major groups based

on physiological and pathological criteria. The first group comprised cases

(19)

dem yelination sometimes associated w ith hypertrophic changes. This group

w as further subdivided into the hypertrophic Charcot-Marie-Tooth disease

T ype 1 (CMTl) and Dejerine-Sottas type neuropathy. The second major group

show ed neural degenerations w ith only slight, if any, reduction in nerve

conduction velocities and little evidence of segm ental demyelination. W ithin

this group fell the neuronal form of Charcot-Marie-Tooth Type 2 (CMT2),

hereditary spastic paraplegia and spinocerebellar degeneration w ith peroneal

m uscular atrophy (H arding and Thomas, 1980 a.) . Thomas et al. (1974) introduced the term hereditary m otor and sensory neuropathy (HMSN) for a

g roup of peroneal atrophies including the Charcot-Marie-Tooth and Dejerine-

Sottas disorders. This term w as later expanded to include CMTl as HMSN I,

CMT 2 as HM SN H, an d Dejerine-Sottas disease as HMSN m .

Dyck and Lam bert (1968 a.) concluded that Roussy-Levy syndrom e is an

expression of the gene for type 1 Charcot-Marie-Tooth disease since it is found

in patients w ith other family m em bers show ing m ore classical features of

CMT. There is no clear distinction in the clinical features of both conditions,

m ore a continuous spectrum of clinical findings. The neurophysiological and

histological features of the tw o disorders are identical.

H ence, CMT disease m ay be divided into tw o types on the criteria of m otor

nerve conduction velocities and pathological evidence of demyelination and

rem yelination giving rise to "onion bulb" form ations and axonal degeneration.

Following the study of 228 affected individuals by H arding and Thomas (1980

a.) the separation of CMT 1 and 2 w as evident in that individuals affected w ith

type 1 h a d m edian m otor nerve conduction velocities less than 38 metres per

second. The division of CMT disease into tw o types allowed for observations

to be m ade on the clinical features of each (H arding and Thomas, 1980 a.).

A bout tw o - thirds of individuals w ith CMT 1 have onset of sym ptom s in the

first decade. In about 60% of cases the initial low er limb weakness is followed

by involvem ent of the u p p er limbs. Total tendon areflexia was found in about

(20)

Distal sensory loss w as reported in over half the cases, pes cavus in nearly

three - quarters and scoliosis in 14% of individuals. The progression of the

disease is slow b u t w hile about 30% developing m arked weakness of the

ankles and feet few experience severe difficulty in getting around. Indeed,

one fifth of the affected relatives of the index patients in the study by H arding

and Thomas (1980 a.) w ere asymptomatic. There appears to be no correlation

of the severity of the disease and the degree of reduction in m otor nerve

conduction velocity.

M otor nerve conduction velocities are norm al or only slightly reduced in

individuals w ith CMT 2. The abnorm alities of sensory nerve conduction are

less obvious th an those seen in type 1 CMT. The neuropathy is one of axonal

origin and not dem yelination as show n in CMT 1. generally, the condition is

less severe and w ith a later age of onset than CMT 1. H arding and Thomas

(1980 a.) show ed the peak age of onset in type 2 patients to be in the second

decade w ith a significant num ber not developing sym ptoms until later,

som etim es n o t until the seventh decade. It is less comm on than in CMT 1 for

the w eakness to progress to involve the u pper limbs cind the upper limb

reflexes are usually preserved.

For both CM Tl and CMT2 the m ost common pattern of inheritance is one of

autosom al dom inance and all the families included in the linkage analysis

portion of this study show this pattern of inheritance. However, alternative

m odes of inheritance have also been reported.

In 1888 H erringham described a family w hich raised the possibility of an X-

linked form of the disease. Allan reported another family in 1939 which also

suggested the existence of an X-linked form (reviewed Vance, 1991). This was

d oubted by some investigators, and further exam ination of the H erringham

(21)

chromosome. The phenotypic expression of the X-linked from of CMT is

sim ilar to the autosom al dom inant form (Dyck et al. 1968a) Affected males in this family show ed slow NCVs and "onion bulb" form ation of sural nerve

biopsy. Some heterozygous females w ere asym ptom atic w ith norm al NCVs,

w hilst others h ad slow ed NCVs and severe m anifestations of the disease.

O ther families have been reported w ith X-linked CMT including a German

family w ith atrophy and weakness, tendon areflexia and reduced NCVs

compatible w ith a dem yelinating neuropathy. Males were m ore severely

affected than females (reviewed in Rozear et al. 1987). A nother family, described by Phillips et al. (1985), was thought to have X-linked dom inant CMT w ith males m ore severely affected than females and slow NCVs in m en

b u t borderline in wom en. Fischbeck et al. (1986) reported tw o families w ith an X-linked m ode of inheritance. In one of these families, males w ere severely

affected in childhood and obligate heterozygous females were norm al

clinically, including their NCVs. The neuropathy appeared to be axonal, not

demyelinating. The second family w as similar to that described by Rozear et al. (1987) w ith heterozygous females show ing variable degree of severity of sym ptoms. Most of the families reported, w ith the exception of that of

Fischbeck et al. (1986) w ere consistent w ith X-linked dom inant inheritance and w ithin m ost of these families the disease appears to be demyelinating. It has

been suggested that the X-linked dom inant and recessive forms are due to

variable expression of the same gene, rather th an separate genetic entities

(Vance, 1991).

H arding and Thomas (1980 b. and c.) have reported several families w ith

autosom al recessive forms of the disease. W ithin these families, there are

several features w hich support recessive inheritance; sym ptom-free parents of

affected individuals, the presence of m ultiple affected siblings, and increased

consanguinity. N erve conduction velocity studies suggest that autosomal

recessive form s exist of both CMT 1 and 2. These families cannot be

(22)

presentation or electrophysiological findings (Loprest et ah 1992).

1.2 ANATOMY OF NEURONES

All axons in the peripheral nervous system (PNS) and central nervous system

(CNS) are sheathed in accessory cells. Structurally, three types of nerve fibres

occur; one is enclosed in a sheath of fatty substance, myelin, and an outer

m em braneous sheath called the neurolem m a (sheath of Schwann) and is

know n as m yelinated or m edullated fibre. A second type has no myelin

sheath b u t does have a neurolem m a and is referred to as non-m yelinated or

non-m edullated fibre. Most non-m yelinated fibres belong to the autonomic

nervous system, which is concerned w ith visceral activity. A third type of

neuronal process lacks a neurolemm a. All fibres w ithin the central nervous

system and the optic and auditory nerves fall into this last category.

The m yelin sheath is form ed by the specialized glial cells - Schwann cells in

the peripheral nerves an d oligodendrocytes in the central nervous system. The

m yelin sheath is divided into segm ents corresponding to the territories of the

individual glial cells. In the m yelinated PNS of m am m als, the segments range

from 200 to over 2,000pm in length. The segm ents join at nodes (nodes of

Ranvier).

The m yelin sheath forms a highly resistive covering which isolates the

internodal axonal m em brane from extracellular electrical influences. The

nodes of Ranvier are the site of electrical excitation. Local currents generated

at one node of Ranvier pass through the cytoplasm of the axon to the next

node. Because of the insulation afforded by the myelin, m ost of this current

leaves at the next node, depolarising it to the point th at it also becomes excited

and begins to generate enough current to excite the next node, and so on.

Through the action of these local currents, the nerve im pulse moves dow n the

fibre, w ith the site of active excitation jum ping from one node to the next.

(23)

nerve impulses - an im pulse is conducted about ten times m ore rapidly than

in the absence of myelin. This m eans that for a given conduction velocity,

m yelination reduces the size of fibre required. Secondly, since excitation only

occurs at the nodes of Ranvier, less energy is spent in conduction of the action

potential, since it is only at the nodes that an "energy debt" is generated that

m ust be repaid during the resting period (Ritchie, 1984).

Ontogenetically, Schwann cells are thought to be of ectoderm al origin,

m igrating from the neural crest into the PNS during development. As axons

em erge from the CNS and enter the PNS, they are followed by m igrating

Schwann cells. Schwann cell proliferation continues until each axon lies w ithin

a furrow along the long axis of the Schwann cell. As the Schwann cell grows

it w raps itself aro u n d the axon. The processes of the Schwann cell that are

spiralled around the axon extrude their cytoplasm to form compact myelin

(Raine, 1984). This process is represented in Figure 1.1

Unlike m ost cells in the body, the term inally differentiated nerve cells caimot

reproduce by mitosis to replace any that are destroyed. The nerve fibres are

outgrow ths of the cell body of the neurone. If a fibre's connection w ith the

cell body is interrupted, the distal fragm ent ceases to function and

degenerates. The fibre m ay regenerate and restore function, provided it has

a neurolem m a. Since the nerve fibres w ithin the central nervous system lack

a neurolem m a, they are incapable of regeneration to restore function.

W hen a fibre w ith a neurolem m a is dam aged, the distal fragm ent disintegrates

and the debris rem oved by phagocytic cells. The Schwann cells of the

neurolem m a proliferate, form ing strands and a pathw ay along the course of

the degenerated portion of the fibre. The distal tip of the viable portion of the

fibre begins to extend bud-like processes and one of these extends into the

tube-like pathw ay form ed by the strands of Schwann cells. The process

continues to grow until it reaches the peripheral destination. N on-m yelinated

(24)

SCHWANN CELL

AXON

EXTRACELLULAR

SPACE

NUCLEUS

Figure 1.1: Schematic representation of PNS myelination showing the process

of specialization of the Schwann cell cytoplasm and compaction.

(25)

The ability of nervous tissue to re-create internodes of myelin following their

destruction and rem oval has been well docum ented (Raine, 1984). Regardless

of the dem yelinating event, the peripheral nervous system (PNS) has the

ability to rem yelinate all affected fibres w ithin a few weeks. Schwann cells

undergo rapid division, and the daughter cells m igrate to positions along the

naked nerve fibre. However, even w hen re-m yelination is complete, axons of

reduced diam eter and thinner-than-norm al PNS m yelin sheaths are still

evident over all the re-m yelinated segments.

1.3 MYELIN

Myelin, like other plasm a m em branes, is m ade up of lipids, carbohydrates in

the form of glycolipids and glycoproteins, and a smaller am ount of protein

Table 1.1: Composition of human PNS m yelin

W ater 35-45% total volume

Total protein 28.7% dry weight

Total lipid 71.3% dry weight

Of which cholesterol 23.0%

galactolipid 22.1 %

phospholipid 54.9%

(From N orton an d Gammer, 1984)

The lipid m atrix of myelin, like th at of other m em branes, is a bilayer of

phospholipids an d glycolipids. The lipids appear to be distributed

asym m etrically across the bilayer. There is about twice as m uch cholesterol

in the outer leaflet than in the inner leaflet of the bilayer. Most of the

cereboside (the m ajor galactolipid) is also found in the outer leaflet. The inner

(26)

MAG

Ext,

MBP

Cyt _MBP

Figure 1.2: Diagrammatic representation of the molecular organisation in PNS

myelin. (Pq) major glycoprotein; (MBP) myelin basic protein; (MAG) myelin

associated protein.

(27)

The m ajor protein com ponents of the compact m yelin sheath include myelin

basic protein (MBP) and myelin-associated glycoprotein (MAG), which are

present in both the central and peripheral nervous systems. Two other

proteins are know n w hich are specific to the tw o parts of the nervous system:

proteolipid protein is found only in the central nervous system; and protein

zero (Pq) which is only found in the peripheral nervous system (Braun, 1984).

The arrangem ent of these proteins across the lipid bilayer is show n in Figure

1.2. The m ajor structural protein, Pq, is a glycoprotein w ith a m olecular weight

of about 28,000-30,000 and accounts for 50-60% of the protein in the PNS.

Im munocytochemical studies have show n that Pq is specific to peripheral

m yelin and is distributed throughout the m yelin lipid bilayer. The myelin

basic protein (MBP) is identical in both the PNS and CNS w ith a molecular

w eight of 18,500. It is thought to exist in heterodim er units and accounts for

40-50% of the total protein. It is an extrinsic protein w ith one or more

dom ains in contact, to a lim ited extent, with the hydrophobic interior of the

bilayer. The P2 basic protein is a m inor component (usually less than 2% of the total protein) and is also of this extrinsic type. The myelin associated

glycoprotein (MAG) constitutes less than 1% of the total protein. It is thought

to be a transbilayer protein w ith m ost of its bulk exposed to the exterior leaflet

of the bilayer (Braun, 1984; Lees and Brostoff, 1984)

1.4 GENE M A PPING APPROACHES

The forty-four autosom es and tw o sex chromosomes which m ake up the

hum an genom e carry betw een them almost all the genetic inform ation needed

for complete developm ent and functioning of all the cells within the hum an

body. The hum an genome contains 2.8 X 10^ base pairs of DNA, although less

than 3% of this codes for "unique" gene sequences. The rest consists of non

coding regions.

Early studies of hum an genetic diseases involved systematic searches of

(28)

biochemical finding w as observed. W hilst this approach provided the solution

to m any diseases, including the thalassaemias, other inherited disorders

involved sym ptom s too complex to allow the direct deduction of the

underlying biochemical defect.

The early 1980's saw the introduction of an alternative approach that allowed

the cloning of the gene by virtue of its chromosomal location. Since this

m ethod bypassed the initial need for know ledge of the basic defect of the

disease, b u t rather allowed this question to be answ ered after the gene had

been cloned, it became know n as "reverse genetics" or "positional cloning".

The chrom osomal localization of a disease locus is the first step in the isolation

and characterization of genes and the defects involved in genetic disorders.

1.4.1 Cytogenetic abnormalities

The identification of a visible cytogenetic abnorm ality present in affected

individuals can indicate the chromosomal location of candidate genes. Genes

know n to m ap near the region of abnorm ality m ust be considered as

candidates. This is especially so if the functional protein could be involved in

the disease process.

The disease locus for Duchenne m uscular dystrophy was initially localized to

the short arm of the X chromosome (Xp21) by the detection of structural

abnorm alities involving this region.

1.4.2 As applied to CM Tl

N o cytological abnorm ality or chromosomal rearrangem ent had been observed

in any patients w ith CM Tl to suggest a chromosomal location for the disease

locus.

1.4.3 Murine hom ology

The use of m an /m o u se syntenic regions represents a powerful tool in the

(29)

for Neurofibrom atosis type (1 N F l), or von Recklinghausen disease serves as

a good example of this com parative m apping strategy. The disease locus was

initially m apped to the long arm of chrom osome 17 by genetic linkage analysis

in affected families (Barker et al. 1987a; Seizinger et al. 1987). Cytogenetic abnorm alities were also im portant in isolating the N F l gene. Two

translocation breakpoints w ere m apped to 17qll.2. Com parative m apping

identified the hum an hom ologue for the m ouse ectropic viral integration site-2

gene (Evi-2). This m urine gene was localized to the syntenic region on m ouse chrom osome 11 and had been im plicated in leukaemogenesis (Buchberg et al.

1990). The hum an hom ologue was found to m ap to the interval betw een the

tw o translocation breakpoints described earlier (O'Connell, 1990). However,

m olecular analysis failed to reveal any m utation in N Fl patients. Examination

of further transcribed sequences in the region isolated a gene which was

finally confirmed as the N F l gene by the detection of several apparent disease-

causing m utations (Wallace et al.1990; C aw thon et al.1990)

1.4.4 Murine hom ology for CM Tl

The Trembler m ouse m utation (Tr) has been suggested as an animal m odel for

CM Tl disease (Vance, 1991). This autosom al dom inant trait was first described

by Falconer in 1951. The m utation manifests as a Schwarm cell defect (Aguayo

et al. 1977) and affected anim als suffer from a hypom yelinating neuropathy and continuing Schwann cell proliferation w ith onion bulb form ation in older

animals. The anim als m ove aw kw ardly and suffer from trem ors and transient

seizures. The Tr m utation m aps to m ouse chromosome 11, near a region that

is syntenic to hum an chrom osom e 17 (Buchberg et al. 1989).

Analysis of the protein com ponents of m yelin isolated from trem bler m ouse

sciatic nerve show s Pq and MBP are virtually absent from trem bler m yelin

(30)

Suter et al. (1992a) show ed that the Tr m ouse carries a point m utation in the peripheral m yelin protein-22 (PMP-22) gene on m ouse chromosome 11. PMP-

22 was initially identified as a grow th arrest specific (GAS-3) gene expressed

in serum starved m ouse 3T3 fibroblasts (Schneider et al. 1988). A transcript (CD25), hom ologous to the gas-3 gene and prim arily expressed in Schwann

cells, w as also identified by differential screening of rat cDNA libraries m ade

from crushed rat sciatic nerve versus contralateral, non-injured sciatic nerves

(Spreyer et al. 1991).

PMP-22 and gas-3 share 98% amino acid hom ology over the complete protein

sequence (Snipes et al. 1992). The rat hom ologue was show n to encode a m yelin protein (PM P-22/SRI3) expressed in rat peripheral nerve (Welcher et al. 1991). This 160 am ino acid protein has been show n to be expressed at a high level in the peripheral nervous system, accounting for u p to 0.2% of the

total mRNA. The transcript was m ost abundant in intact sciatic nerve and was

very low in other tissues including the central nervous system, suggesting a

high level of tissue specificity in the rat (Spreyer et al. 1991). Im muno- histochemical analysis (Welcher et al. 1991) dem onstrated that PMP-22 was associated w ith the m yelin sheath and correlates tem porally w ith the formation

of m yelin w hen com pared to the expression of MBP (Snipes et al. 1992).

Tem poral changes in the relative abundance of the mRNA in sciatic nerve

w ere analyzed using N orthern blots following tw o types of injury: firstly,

crush injury, which leads to W allerian degeneration in the distal nerve

segm ent prior to regeneration of axons from the proximal stum p into the distal

portion; an d secondly, transection of nerve and perm anent separation of both

portions to prevent axonal grow th into the degenerating distal portion.

In the proxim al portions of both crushed and transected sciatic nerves only

m inor changes in the abundance of mRNA w ere observed for at least four

(31)

rapidly decreased w ithin tw o days in the case of a crush injury and one week

following transection. In the regenerating distal portion of crushed nerve, the

transcript levels w ere show n to increase during the second week after the

injury, reaching control values of norm al nerve w ithin six weeks. Conversely,

in the distal portion of transected nerve, w here no axonal regeneration

occurred, the mRNA levels did not return to the norm al levels.

The p attern of expression of PMP-22 in the distal nerve stum p following sciatic

nerve crush injury w as comparable to the expression of other PNS myelin

proteins Pq and MBP (Snipes et a l 1992).

Spreyer et a l (1991) show ed that the re-expression of the transcript in crushed peripheral nerve w as closely related to axonal regeneration. Anastomosis of

a perm anently denervated distal stum p to its non-degenerated proximal nerve

segm ent lead to re-expression of mRNA in the distal stum p portions which

had received regenerating axons. These results suggested that Schwann cells

associated w ith axons express high levels of the transcript, and that

interruption of this relationship leads to a dow n-regulation in expression in

proliferating Schwann cells following nerve injury in vivo. Restoration of

Schwann cell-axon contact in regenerating nerve was thought to re-establish

the elevated levels of transcription to those of non-injured nerve.

H aving established PMP-22 as a putative m yelin protein, its expression was

com pared w ith other m yelin proteins (Snipes et a l 1992). The pattern of PMP- 22 expression in the PNS during developm ent was found to be essentially

identical to other PNS proteins such as Pq and MBP. In the im m ediate post

natal period PMP-22 mRNA is expressed at low levels (10% of the adult

m axim um ), b u t rapidly increases over the first three post-natal weeks w hen

it reaches near maximal levels.

The localisation of PMP-22 mRNA w ithin the m yelin sheath to the perinuclear

(32)

m em brane protein (Snipes et al. 1992). This pattern of perinuclear localisation is also seen in the PNS protein Pq and the CNS PLP. This is com pared to the

diffuse cytoplasmic localisation seen w ith MBP mRNA. These patterns of

mRNA expression are considered to dem onstrate the fact that both Pq and PLP

are integral m em brane proteins which m ust undergo intracellular processing

through the rough endoplasm ic reticulum and Golgi apparatus as opposed to

the soluble MBPs which are synthesised on free ribosomes. Lam perth et al.

(1990) have show n directly Pq mRNA processing using mRNA hybridisation

at the ultrastructural level to dem onstrate th at Pq mRNA is localised to the

rough endoplasm ic reticulum. Snipes et al. (1992) also show ed imuno- histochemically that PMP-22 w as not highly expressed in the Schwann cell

cytoplasm and was excluded from the cytoplasmic invaginations into the

m yelin sheath.

A dditional evidence that PMP-22 is an integral m em brane protein was

provided by analysis of its amino acid sequence which reveals four

hydrophobic dom ains thought to function as transm em brane spanning regions

and also a consensus site for N -linked glycosylation. PLP (which is restricted

to the central nervous system) was also predicted to be an integral m em brane

protein w ith four m em brane-associated zones (Pham-Dinh, 1991). M utations

in PLP h ad been identified as the prim ary defect in hereditary conditions

affecting the m yelination of the central nervous system in mouse, dog and

m an (review ed in Pham-Dinh, 1991).

The Trem bler m ouse show ed a point m utation, substituting an aspartic acid

residue for a glycine at position 150, in a putative m embrane-associated

dom ain of the PMP-22 protein (Suter et al. 1992a). This exchange introduces a charged am ino residue in the fourth putative transm em brane-spanning

dom ain and is likely to affect the structure and functioning of the protein.

(33)

Tr m ouse phenotype: the PMP-22 protein is expressed predom inantly by the

Schw ann cells of the peripheral nervous system and not w ithin the central

nervous system, the Tr m utation is m anifested as a Schwann cell defect of the

peripheral nervous system w ith no know n consequence in the central nervous

system. Secondly, the PMP-22 protein is localized in the compact myelin

sheath, consistent w ith the hypom yelination of peripheral nervous system

axons seen in Tr m utants. Thirdly, since PMP-22 mRNA expression has been

im plicated in cellular grow th arrest, the abnorm ality of PMP-22 could account

for the persistent Schwann cell proliferation seen in Tr mice.

Trembler-J (Tr^) is a spontaneous m utation th at arose independently from

Trembler. Tr and Tr^ w ere thought to be allelic since the two disorders were

phenotypically so similar and they both show ed linkage to the same m arker

(vestigial tail). Histologically both Tr and TP are characterized by a decrease

in axon m yelination and an increase in the num ber of Schwann cells.

H ow ever, there are m arked differences betw een the two. W hereas expression

of the Tr allele is truly dom inant, that of TP is semi-dominant, perhaps

show ing a dosage effect. Mice hom ozygous for the Tr^ m utation show the

m ost severe peripheral m yelin deficiency and die at 17 or 18 days. Mice

hom ozygous for the Tr m utation are longer-lived (Henry and Sidman, 1988).

Mice heterozygous for these m utations also show m arked differences. T r/+

mice have a m ore severe progression of sym ptom s w ith age than Tr^/+ mice.

This is especially apparent in the "onion bulb" structures, show ing repeated

cycles of dem yelination and rem yelination which are m uch m ore prom inent

in the T r / + animals.

The putative prim ary defect in Titm ice has been identified as a point m utation

at nucleotide position 190 - from thym ine to cytosine - which translates into

an exchange of proline for leucine at residue 16 (Suter et ah 1992b).

The tw o independent non-conservative m utations in the PMP-22 gene,

(34)

hereditary peripheral neuropathies in mice and possibly in hum ans.

1.4.5 Linkage approach

For the m ajority of genetic diseases in m an, gross chromosomal

rearrangem ents are not found as m arkers for the position of the disease gene.

In these cases gene localization m ay be achieved through linkage analysis w ith

know n genetic m arkers. The description of the localization of the gene for

cystic fibrosis (CF) illustrates that it is possible to identify a disease locus

solely on the basis of linkage analysis. N o gross chromosomal rearrangem ents

had been observed in affected individuals and no biochemical data was

available to assist in the localization of the gene. The high frequency of CF

m ade it possible to collect sufficient affected families for linkage analysis. The

first linkage to CF was detected w ith an enzym e m arker PON (Eiberg et al.

1985), b u t the chrom osomal location of CF resulted from linkage to an

arbitrary RFLP m arker D7S15 (Tsui et al. 1985). This m arker was m apped to chrom osom e 7 using a somatic cell hybrid (Knowlton et al. 1985). Two further closely linked probes w ere then described, MET (White et al. 1985) and D7S8 (W ainwright et al. 1985). These m arkers w ere show n to flank the CF gene. A variety of m ethods were used to generate probes within the region defined by

MET and D7S8. The CF gene w as eventually cloned through a series of

chrom osom e w alking and jum ping experiments, detection of cross-species

hybridization an d screening cDNA libraries m ade from tissues affected in CF

patients.

1.4.6 Linkage approach for CMT

A utosom al dom inant CMT disease is the m ost frequent form of the disease,

w ith a prevalence estim ated at 36/100,000 (Skre, 1974). The genetic basis for

autosom al dom inant CMT 1 has been extensively investigated in recent years.

In 1978 H eim ler et al. reported a single family w ith a dom inant hereditary neuropathy that appeared to be segregating together w ith the dom inant nevoid

basal cell carcinom a syndrom e. Previous evidence suggested linkage of the

(35)

therefore, it w as suggested that the locus for CMT disease could also be on

chrom osom e 1. This lead to the publication of a num ber of papers showing

linkage of the CMT disease locus to chrom osom e 1. Bird et ah (1982) presented tw o families comprising 3 and 4 generations, with a total of 23

affected individuals inform ative for the Duffy blood group locus know n to be

on the long arm of chromosome 1 (lq21 - q25). The m axim um lod score

obtained w as 2.297 at a recombination fraction of 0.10. In the same year

(Guiloff et al. (1982) obtained a slightly positive m axim um lod score of 0.890 at a recom bination fraction of 0.01 w ith the Duffy blood group locus in a study

of four families w ith 22 affected individuals. A stu d y of a single large kindred

of 50 individuals, 18 of w hom w ere affected (Stebbins and Conneally, 1982)

yielded a m axim um lod score of 3.11 at a recom bination fraction of 0.05 with

the Duffy locus. Dyck et al. (1983) described a family segregating for CMT I w ith a m axim um lod score of 1.19 at recom bination fraction 0.01 which added

further to the hypothesis th at type I CMT w as indeed linked to the Duffy locus

on the long arm of chrom osome 1.

Subsequently, several groups failed to confirm this linkage to chromosome 1.

These included Bird et al. (1983) w ith a lod score of -5.383 at a recombination fraction of 0.01 in a large three generation family, Dyck et al. (1983) w ith a lod score of -10.94 at a recom bination fraction of 0.01 and M arazita et al. (1985) w ith a lod score of -1.57 at a recom bination fraction of 0.01. Further

significantly negative lod scores were generated; lonasescu et al. (1988a) -10.83 at a recom bination fraction of 0.01, Griffiths et al. (1988) -15.33 at a recom bination fraction of 0.05 and M iddleton-Price et al. (1989) w ith a lod score of -20.74 at a recom bination fraction of 0.01. The lod scores between

CMT 1 and the Duffy locus are sum m arized in Table 1.2

Based on the genetic linkage results, at least tw o subtypes of autosomal

dom inant CMTl appeared to exist. Bird et al. (1983) proposed that CMT type la should refer to the disorder not linked to the Duffy blood group locus on

(36)

*

Table 1.2: Lod scores between CMT I and Duffy Marker (lq21 - q25)

Reference 0 (recom bination fraction)

0.01 0.05 0.10 0.20 0.30

Bird et a/.1982 0.36 1.99 2.30 1.96 1.30

Guiloff et fl/.1982 0.89 0.82 0.73 0.54 0.36

Stebbins & Conneally,1982 2.69 3.11 3.03 -

-Dyck et a/.1983 1.19 1.12 1.02 0.82

Bird et a/. 1983 -5.38 -2.72 -1.68 -0.78

Dyck et af.1983 -10.94 -4.98 -2.65 -0.77

-M arazita et a/. 1985 -1.57 -0.86 -0.55 -0.25

-lonasescu et a/.1988a -5.86 -2.32 -1.31 -0.33 -0.09

Griffiths et fl/.1988 - -15.33 -8.93 -3.71

(37)

In 1989 Vance et al. published linkage data on 194 individuals in six families segregating for autosom al dom inant CMT. This show ed significantly positive

lod scores w ith tw o probes m apping to the pericentric region of chromosome

17 and was suggestive of genetic heterogeneity of CMT type 1. Probe p i 0-41

(D15S71) gave a m axim um lod score of 7.36 at a recombination fraction of 0.06,

w hilst EW301 (D17S58) gave a m axim um lod score of 10.49 at a recombination

fraction of 0.05. This was shortly followed by other groups publishing data

from studies of their families w ith these chromosome 17 probes. In a large

pedigree Raeymaekers et al. (1989) w ere able to exclude chromosome 1 as the location for CM Tl and show linkage to chromosome 17. M iddleton-Price et al. (1990) studied eight families w ith CMT type I and confirmed linkage for D17S58 (EW301) w ith a m axim um lod score of 5.89 at a recombination fraction

of 0.08 and for D17S71 ( p i0-41) w ith a m axim um lod score of 3.22 at a

recom bination fraction of 0.08. A study of five Canadian families (MeAlpine

et al. 1990) gave a m axim um lod score of 10.83 at a recombination fraction of 0.00 w ith the probe EW301. Linkage analysis in a large French-Acadian

kindred segregating for CM Tl a (Patel et al. 1990) gave m axim um lod scores w ith EW301 and p i 0-41 of 2.37 (at a recombination fraction of 0.07) and 0.89

(at a recom bination fraction of 0.00). Tim m erm an et al. (1990) further analyzed the five-generation Belgian family affected w ith CMTl a initially reported by

Raeymaekers et al. (1989) for linkage w ith the chromosome 17 markers. Probe p i 0-41 gave a m axim um lod score of 6.13 at a recombination fraction of 0.00

and EW301 a m axim um of 11.29 at a recombination fraction of 0.08. The lod

(38)

CO

Table 1.3: Lod scores Betw een CM T I and Chrom osom e 17 M arkers.

6 recom bination fraction)

0.001 0.05 0.10 0.15 0.20 0.30 0.40

Vance e t ai.l989

EW301 6.21 10.49 10.18 9.30 8.18 5.42 2.29

plO-41 5.13 7.26 6.77 5.92 4.93 2.77 0.91

M iddleton - Price e t «1.1990

EW301 -0.14 5.68 5.85 5.49 4.87 3.29 1.52

plO-11 0.31 3.20 3.18 2.95 2.62 1.78 0.84

M cAlpine e t «1.1990

EW301 10.81 9.89 8.91 - 6.78 4.42 1.93

Patel e t «1.1990

EW301 1.89 2.34 2.25 - 1.78 1.14 0.45

plO-41 0.89 0.79 0.67 - 0.43 0.20 0.05

T im m erm an et «1.1990

EW301 2.42 11.04 11.21 - 9.54 6.76 3.33

(39)

The genetic relationship and location of D17S58 (EW301) and D17S71 (plO-41)

are well established (Goldgar et ah 1989). Pooled data from both sexes suggests th at EW301 lies approxim ately 5.5cM from the centromere on the

short arm of chrom osom e 17. Probe p i 0-41 is thought to lie 3.3cM distal to

EW301. W hilst initial m ulti-point analysis localized CMT type I to

chrom osome 17 it failed to resolve the relative order of the three loci. The fact

that some meioses show ed crossovers w ith both probes (Middleton-Price et ah

1990) m ade it less likely that the disease gene locus fell betw een the probe loci

than outside them. H ow ever it was unable to confirm w hether the disease

locus w as telomeric or centromeric.

W ith re-exam ination of m any CMTl families supporting a chromosome 17

locus and the low num bers of new chromosome 1-linked families, it was

suggested that the chrom osome 1 form of autosomal dom inant CMTl m ay be

not be a real entity, b u t m ay be a statistical artifact of the linkage data

(M iddleton-Price et ah 1989; Raeymaekers et ah 1989). However, it w ould appear th at genetic heterogeneity does exist for the autosomal dom inant form

of CM Tl. Chance et ah (1990) re-exam ined tw o families previously included in the stu d y of Bird et ah (1982). One pedigree which h ad show n a weakly positive lod score of 0.64 at 0 = 0.15 w ith the Duffy locus w as actually linked

to m arkers on chrom osom e 17 w ith a m axim um lod score w ith probe EW301

of 3.22 at 8 = 0.10. The other family segregating for CMTl had given a

m axim um lod score w ith the Duffy locus of 2.00 at 8 = 0.00 failed to show

linkage to the chrom osom e 17 probes w ith a m axim um negative lod score of

-3.52 w ith probe pA10.41 at 8 = 0.001.

A dditional evidence supporting genetic heterogeneity in CMTl has been

p rovided by Defesche et ah (1990) w ho described five CM Tl pedigrees, one of w hich show ed no linkage to chrom osome 17 m arkers and gave evidence for

(40)

lonasescu et al. (1992) described a fam üy segregating for autosomal dom inant CM Tl. Tw o-point and m ulti-point linkage analyses were strongly suggestive

of a CMT locus on chromosome Iq w ith a m axim um m ulti-point lod score of

2.70 at M U C l (0 = 0). M ulti-point analysis excluded the CMTl locus from

chrom osom e 17 m arkers in this family. Affected m em bers of this family also

failed to show evidence of the segmental trisom y detected w ith probe

VAW409R3a, th at has been described in families w ith CMTl a.

M uch less is know n about the genetic locus for CMT2. As CMTl and CMT

2 are clinically indistinguishable in an individual patient (without examination

of the NCVs), allelic heterogeneity has been suggested as a possible aetiology

for the differences in the tw o types. Previous linkage studies for CMT2

included a rep o rt by lonasescu et al. (1988b) of linkage data for CMT2 with the chrom osom e 1 serum amyloid m arker APCS in three families. A maxim um

lod score of 1.24 was achieved at 0 = 0.00. A later report by Loprest et al.

(1992) on three large families segregating for autosomal dom inant CMT2

excluded bo th the CM Tlb locus on chromosome 1 and the CMTl a locus on

chrom osom e 17.

1.5 GENE MAPPING BY PHYSICAL METHODS

Physical m aps m ay be cytogenetically or m olecularly based. Cytogenetically

based physical m aps order loci w ith respect to the visible banding pattern or

their relative position along the chromosome. The data for these m aps comes

prim arily from somatic cell hybrids and in-situ-hybridization. Molecularly

based physical m aps characterize large spans of DNA directly by establishing

m olecular landm arks including restriction endonuclease sites and "sequence

tagged sites". These m aps are usually produced from data generated from

(41)

1.5.1 Banding of chromosomes

Cytogenetically based physical m aps order loci w ith respect to the visible

banding p attern or their relative position along the chromosomes. Initially,

h um an chrom osom es were classified into one of seven groups on the basis of

their size. It was not usually possible to further differentiate betw een

chrom osomes unless polym orphic characteristics such as the uncoiler region

of chrom osome 1 allowed for recognition of individual chromosomes. In 1971

the Fourth International Congress of H um an Genetics introduced a num bering

system based on chromosome banding patterns produced by differential

staining techniques which m eant that each chromosome could be distinguished

and allowed for the description of a particular region on either the long or the

short arm of a particular chromosome.

1.5.2 Somatic cell hybrids

Identification of individual chromosomes also lead to the construction of

hum an-rodent cell hybrid panels - a range of rodent cells with a varied content

of hum an chromosomes. This has been used to m ap gene loci to

chromosomes. Somatic cell hybrids form ed a vital resource in the m apping of

the short arm of chromosome 11 in relation to W ilms' tum our. A hum an-

ham ster somatic cell hybrid that had chromosome 11 as its only hum an

com ponent (Puck et ah 1971) was used to generate a series of cell lines containing overlapping portions of chromosome 11. Deletion derivatives of

these lines form ed a deletion m ap of l i p based on the pattern of m arker

segregation in this panel. This has contributed to the developm ent of a

detailed m ap of this region of chromosome 11 by quickly localizing any

m arker or gene to a defined location. It has also proved invaluable in defining

the chrom osom al abnormalities associated w ith familial aniridia and WAGR

syndrom e (Pelletier et al. 1991).

(42)

m arkers from the region of chromosome 4 near to the H untington disease

locus (Cox et al. 1989).

1.5.3 Deletion mapping

The precise characterization of deletions has allowed the direct m apping of

som e disease loci. For example, the gene which codes for acid phosphatase 1

(APCl) in erythrocytes was localized to chromosome 2 by cytological

exam ination of tw o children (from tw o separate families) each w ith m ultiple

congenital abnormalities. Both children were show n to carry deletions of the

term inal portion of the short arm of chromosome 2. In one child, the

breakpoint was in the distal section of band 2p23. C ultured cell lines from this

child show ed norm al ACPI activity. The deletion in the second child extended

from the proxim al portion of band 2p23 to the telomere, cell lines containing

this chrom osome show ed no ACPI activity. From this evidence it was

inferred th at the ACPI gene is located w ithin the band 2p23 (Emanuel et al.

1979).

1.5.4 Translocation mapping

The locus for D uchenne m uscular dystrophy was also initially localized by the

detection of structural chromosome abnormalities. These structural

abnorm alities included X-autosome translocations in affected females

(Greenstein et al. 1977; Zatz et al. 1981) and a small deletion of Xp21.1 in a patient w ith m ultiple abnorm alities including DMD and McLeod syndrom e

(Francke et al. 1985).

1.5.5 In-situ-hybridization

Chrom osom e banding could be combined w ith techniques such as in-situ-

hybridization to enable direct m apping of gene probes to m etaphase spreads

(43)

(C O L lA l) to chrom osome 17q21-q22 (D'Eustachio and Ruddle, 1983). As well

as hybridizing isotopically labelled probes to chromosome spreads, fluorescent

molecules can be deposited at the sites of specific DNA sequences via

fluorescent in-situ-hybridization (FISH). Cosmids can be m apped w ith a

resolution of approxim ately 3Mb on m etaphase chromosomes, whilst using

FISH on interphase nuclei it has been possible to order seven probes m apping

w ithin 250kb in the DHFR region of Chinese ham ster cells (Trask et al. 1989).

1.5.6 PFGE

Pulsed - field gel electrophoresis (PFGE), which can separate DNA fragm ents

of several million base pairs, provides a m eans of constructing long range

physical m aps w hen used in conjunction w ith restriction endonucleases that

cut infrequently. This enables the identification of deletions and major

chrom osom al rearrangem ents (Schwartz et al. 1984). PFGE was used to generate a complete long-range restriction m ap of the region thought to be

involved in W ilm ' tum our (Rose et al. 1990). A specific deletion of l l p l 3 detected in W ilm 's tum our patients allowed for the characterisation of a

transcript encoding a zinc finger protein thought to represent the W ilm 's

tu m o u r gene (Call et al. 1990).

1.5.7 Chromosome walking and jumping

A chrom osom e w alk starts at a cosmid or YAC containing a stretch of DNA

th at has been m apped and orientated and is w ithin several megabases of the

target gene of interest. The cosmid or YAC is used to isolate other stretches

of DNA from a library w ith which it overlaps. This process is repeated until

the target gene is reached.

The distance covered by one step can be increased by extending the range of

m ethods used to isolate fragments. The circularization of large stretches of

DN A in this technique tu rn a walk into a chromosome jump. Chrom osom e

(44)

term inating a chrom osom e walk, it allows these sequences to be jum ped over.

Once a jum p has been m ade and positioned on a pulsed-field m ap,

conventional cosm id or phage walks can be initiated to yield probes suitable

for genetic analysis and physical m apping.

1.6 GENETIC MAPPING

Genetic m aps have been constructed from m any different types of data from

the first genetic linkage m ap produced by Sturtevant in 1913 to the complex

m olecular m aps of today. By 1991,2325 genes had been m apped in the hum an

genome, equivalent to approxim ately 5% of the estim ated 50,000 genes thought

to comprise the hum an genome. M ore than 10000 loci were defined by DNA

m arkers,of w hich about 3,000 are polym orphic (Williamson et a l 1991).

1.6.1 Linkage

Genetic linkage m aps are based on the co-inheritance of allele combinations

across m ultiple polym orphic loci. The parental combinations are usually

transm itted if the loci are molecularly close, b u t recombination at meiosis m ay

generate non - parental combinations m ore frequently if the loci are further

apart. The prim ary source of linkage data is the observation of allele

combinations in gametes. The allelic constitution has been determ ined

indirectly by fam ily studies b u t direct molecular analysis of gametes and single

chrom osomes has recently become possible. Linkage m aps are described in

term s of recom bination fraction and Morgans.

1.6.2 Genetic map distances

The term "recombination fraction" was introduced to describe the proportion

of non - parental recom binants in the total num ber of gametes and, therefore,

the genetic distance betw een them. The M organ (M) w as introduced as the

unit distance over which one crossover should occur in every gamete. Hence,

(45)

recom bination accounts for one in every h u n d red gametes. Unlike physical

distances betw een loci m easured in kilobases of DNA, distances m easured as

recom bination fraction are not additive. Therefore,whilst cM units are based

on the recom bination fraction, they are only directly proportional to it over

small distances.

The am ount of recom bination that occurs varies betw een the different

chrom osomes an d also betw een the sexes. The sites at which homologous

chrom osomes appear to exchange genetic m aterial during meiosis are term ed

chiasmata. A t least one chiasma occurs per chromosome arm per meiosis.

A bout 53 chiasm ata occur in the autosom es of the male giving an estim ated

m ale autosom al m ap length of 26.5M.

Recombination frequencies per megabase of DNA vary considerably by

chrom osom e region and by sex. Overall, there are higher frequencies of

recom bination in female meioses than in m ale meioses although the ratios of

sex specific m ap lengths differ am ong the chromosomes. The pattern of

regional variation shows that centromeric regions have proportionally higher

frequencies of recom bination in female meioses, and that telomeric regions

have m ore recom bination in m ale meioses.The higher recombination in females

results in an autosom al m ap 1.5 times that of the male. The predicted total

length for the sex - averaged linkage m ap is 3300cM (Morton, 1982) w ith the

average length of a chrom osome being 1.5M, that is, experiencing an average

of 1.5 chiasm ata per chromosome (Ott, 1991).

1.6.3 M apping functions

W hilst the frequency of recom bination is directly proportional to the physical

m ap interval over small intervals, the relationship betw een these two

param eters is complex. For larger intervals, this direct relationship is not