MOLECULAR GENETICS OF THE
HUMAN X CHROMOSOME
BY
DAVID ANDREW HARTLEY
a thesis submitted for the degree of
Doctor of Philosophy
in the
University of London
January, 1984 Department of Biochemistry St. Mary's Hospital Medical School London W2 IPG
1 Abstract
The comparison of frequency of recombination with
chromosomal position has been possible only in Drosophila melanogaster but it is known that chiasma distribution in
the human male (and many other species) is not random. A number of cloned DNA sequences have been isolated from a
library enriched for X chromosome DNA by flow cytometry.
The chromosomal loci complementary to these cloned
sequences have been investigated by hybridization to
somatic cell hybrid DNAs and by direct hybridization 1 in
situ' to human metaphase spreads. In this way a
cytological map of the X chromosome defined by DNA
sequence markers has been constructed. Genetic distances
complementary to the physical separations determined have
been estimated by determining the segregation of DNA
sequence variants at each locus in three generation
p e d i g r e e s .
The sequence variation is manifested as altered
restriction fragment migration in agarose gels. It is
intimated from these analyses that the pattern of
recombination frequency along the human X chromosome mirrors the non-random chiasma distribution seen along male autosomes. This supports the chiasmatype theory and has implications on the availability of genetic components
of the X chromosome to recombination.
2 To ray father in the hope of convincing him of life outside
o f E.coli
3 ...scientists getting their kicks when deadly disease can do as it pleases
results ain't hard to predict
Gil Scott-Heron
4 Ack n ov; ledgements
To my supervisor Bob Williamson for support. To my other supervisor Kay Davies for tolerance and PR. To
Graham Casey, Marion Stone, Teresa Davies and Paul Barton for teaching me the true meaning of cytogenetics. To Peter
Goodfellow for hybrid cells. To Paul North for microchip technology. To Helen Kingston and Dennis Drayna for
families and Lod scores.
To all the Biochemistry Department and in particular the 'X group' for tolerating my singing and my socks. To
Julian and Hark for endless attempts at the meaning of life and for going for it. To Jo, Sue, Teresa, /Auntie
Janet, Sissel, Pauline and Susie for demonstrating the superiority of two X chromosomes. To Jack, Ali, Rob,
Graham, Tim, Conrad, Ola and Simon for coming a close s ec o n d .
To my family and friends.
To my skin and blister for never letting up in giving me a hard time (no more than I deserved).
To Jane.
5 Abbreviations dATP adenosine deoxyribonucleotide triphosphate AP RT adenosine phosphoribosyl transferase b p ba se pair 3 kb p 10 base pairs BSA bovine serum albumin o C degrees Centigrade u C i ,m C i ,Ci micro, milli, Curie dCTP cvtidine deoxyribonucleotide triphosphate DNA deoxyribonucleic acid DNAase deoxyribonuclease dpm disintegrations per minute E. coli Escherichia coli EDTA ethylene diamine tetraacetic acid G6 PD H glucose-6-phosphate dehydrogenase pg / ng , ug , mg , g pico, nano, micro, milli, gram dGTP guanosine deoxyribonucleotide triphosphate hrs hours HAT hypoxanthine adenine thymidine HGPRT hypoxanthine-guanosine phosphoribosyl transferase u l ,m l ,1 micro, milli, litre Lod logarithm of the odds nm,um,mm,cm nano, micro, milli, centimetre 2 cm squared centimetre mins minutes u M ,m M ,M micro, milli, molar cM centiMorgan PPO 2,5-diphenyloxazole PEG polyethylene glycol P probability RFLP restriction fragment length polymorphism RFLV restriction fragment length variant rpm revolutions per minute RNA ribonucleic acid RNAase ribonuclease RPMI-1640 Roswell Park Memorial Institute-1640 m e d i u m SSC saline sodium citrate secs seconds SDS sodium dodecyl sulphate SD standard deviation Tris 2-amino-2(hydroxymethyl)-propane-1,3-diol dTTP thymidine deoxyribonucleotide triphosphate UV ultraviolet V volts v / v volume per volume of water w / V weight per volume of water
6 F i g u r e s ge
Secondary screenings of X-enriched library recombinants with human, Horl9X and mouse DMAs. 60
Screenings of X-enriched library with Horl9X and
MCP-6 DMAs. 62
EcoRI digests of recombinants DX1-12. 64
Recombinants DX1-12 hybridized to human and lambda DMAs. 66
EcoRI digests of human, Horl9X, IV7I-5, MCP-6 and mouse DMAs. 71
Somatic cell hybrid DMAs hybridized to DX24 and
RD6 . 73
Somatic cell hybrid DMAs hybridized to S21 and
DX4 . 75
Somatic cell hybrid DMAs hybridized to DX7 and
DX6 . 76
4C11 hybridized to a metaphase spread. 82
Distribution of grains over chromsome 6s after
'in situ' hybridization to 4C11. 85
Somatic cell hybrid DMAs hybridized to 4C11. 86
Repetitive seguences hybridized 'in situ' to metaphase spreads. 83
RC8 hybridized to a metaphase spread. 90
Distribution of grains over all chromosomes after
'in situ' hybridization to RC8. 91
7 15. Distribution of grains over the X chromosomes
after 'in situ' hybridization to RC8. 93
16. LI.28 hybridized to a metaphase spread. 95
17. Distribution of grains over all chromosomes after
'in situ' hybridization to LI.28. 96
18. 52A hybridized to a metaphase spread. 99
19. Distribution of grains over all chromosomes after
hybridization to 52A. 100
20. S21 hybridized to a metaphase spread. 102
21. Distribution of grains over all chromosomes after
hybridization to S21. 103
22. DP31 hybridized to a metaphase spread. 106
23. Distribution of grains over all chromosomes after
hybridization to DP31. 107
24. DX13 hybridized to a metaphase spread. 109
25. Distribution of grains over all chromosomes after
hybridization to DX13. 110
26. Distribution of grains over the X chromosome
after hybridization to RC8, LI.28, DP31, S21, 52A
and DX13. 113
27. Cvtological map of the X chromosome. 114
28. The TaqI polymorphism detected by RC8. 117
29. The TaqI polymorphism detected by S21. 119
30. The TaqI variants detected by LI.28 and DP31 in
the pedigree D3. 123
31. Pedigree D3. 124
32. Pedigree D2. 126
8 33. The TaqI and Bglll variants detected by DP31 and
DX13 in the pedigree BD26. 128
34. Pedigree BD26. 129
35. The physical and genetic map of the human X
chromosome. 133
9 Tables
Page
1. Recombinants picked from the X-enriched library -
insert sizes and hybridization to total human
DliA. 69
2. Hybridization of recombinants to somatic cell
hybrids - chromosomal localization. 78
3. Statistical analysis of 4C11 'in situ'
hybridization. 84
4. Statistical analysis of RC8 'in situ'
hybridization. 92
5. Statistical analysis of LI.28 'in situ'
hybridization. 97
6. Statistical analysis of 52A 'in situ'
hybridization. 101
7. Statistical analysis of S21 'in situ'
hybridization. 104
8. Statistical analysis of DP31 'in situ'
hybridization. 108
9. Statistical analysis of DX13 'in situ'
hybridization. Ill
10. Restriction fragment variants in eight mothers
detected by RC8, LI .28,DP31, S21, 52A and DX13. 121
11. Lod scores at different values of recombination
fraction between the loci detected by LI.28 and
DP31 and between RC8and Ll.28. 131
1 0 Contents Page
Abstract 2
Dedication 3
Quotation 4
Acknowledgements 5
Abbreviations 6
List of Figures 7
List of Tables 10
1 . Introduction 14
1 .1 . The eukaryotic chromosome 15
1 .2. Meiosis and the reassortment of genes 20
1.3. Recombination between linked genes and the
construction of linkage maps 22
1.4. Crossing over and chiasmata 25
1 .5 . Localized recombination 27
1 .6 . Cytological and genetic markers inman 30
1 .7 . Flow cytometry and molecular cloning of the
human X chromosome 33
1.3 . Summary and aims 35
2. Materials and Methods 37
2.1. Materials 38
2.2. Bacterial and phage strains 41
11 2.3. General methods 41
2.3.1. Restriction enzyme digestion 41
2.3.2. Ethanol precipitation of nucleic acids 42
2.3.3. Agarose gel electrophoresis 43
2.3.4. Autoradiography 43
2.3.5. Gel filtration 43
2.3.6. Determination of radioactivity 44
2.4. Specific methods 44
2.4.1. Preparing high molecular weight DHA from
tissue culture cells 44
2.4.2. Small scale preparation of bacteriophage DMA 45
2.4.3. Large scale preparation of bacteriophage DNA 46
2.4.4. Labelling DMA by nick translation 47
2.4.5. Screening lambda recombinants 43
2.4.6. Southern transfers 50
2.4.7. Competing repetitive sequences in Southern
blot hybridizations 51
2.4.8. Preparation of metaphase spreads 51
2.4.9. G-banding of metaphase chromosomes 52
2.4.10. 'In situ' hybridization 53
2.4.11. Scoring hybridized spreads 55
2.4.12. Pedigree analysis 56
12 3. Results 57
3.1. Screening recombinant bacteriophage 'in situ' 53 3.2. Isolation and characterization of individual
recombinant bacteriophage S3 3.3. Chromosomal localization I. Somatic cell
hybrids S3 3.4. Chromosomal localization II. 'in situ'
hybridization 79 A. Introduction 79
B. Cytological mapping of the X chromosome 29 C. Summary of 'in situ' hybridization data 112 3.5. Restriction fragment length polymorphisms 115
A. RFLP and heterozygote detection 115 B. Inheritance studies 122 3.6. Genetic distances - a genetic map of the X 130
4. Discussion 135
4.1. DBA sequence organization of the human X chromosome 136
4.2. The construction of chromosome maps in
r1 amma 1s 133 4.3. Localized recombination 143
5 . Summary and perspectives 154
6 • References 15 3 hrrata 16 6
13 C H AP TE R 1
Introduction
14 1.1. The Eukaryotic chromosome
The chromosome, first observed by Waldeyer in 1888, can be defined as a linkage structure consisting of a specific linear sequence of genetic information. The eukaryotic chromosome has two main biological functions.
The first is the transmission of genetic information from cell to cell and generation to generation and the second the ordered release of this information to control cellular function and development. The chromosomes of higher eukaryotes are characterized by their complex structure and by their behaviour during mitosis and meiosis when they undergo extraordinary changes in morphology.
The dissection of the chromosome by microscopic and biochemical techniques has revealed much of its molecular architecture. By thin section techniques and surface spreading of lysed cells the electron microscope reveals
'thick' 20-30nm fibres which unfold at low ionic strength to the characteristic 'beads on a string' structure
(DuPraw, 1970). This beaded structure corresponds to the now well established nucleosome which consists of two molecules each of the histones H2A, H2B, H3 and H4 with
140-200 base pairs of DMA (Oudet et al., 1975). The 20-
30nm fibre is interpreted variously as a solenoid of nucleosomes (Finch and Klug, 1976) or a repeating unit of
6-10 nucleosomes as condensed 'superbeads' (Hozier et al.,
1977).
15 A series of experiments have addressed themselves to the question as to whether the DNA is continuous and unineme in each chromosome. Observations on the deoxyribonuclease digestion of lampbrush chromosomes
(Gall, 1963), on the isotopic labelling of replicating chromosomes demonstrating semi-conservative inheritance of labelled DNA (Taylor, 1963), on the bromodeoxyuridine label distribution following sister chromatid exchange
(Latt, 1981) and on the viscoelastic properties of large
Drosophila melanogaster DNA molecules (Kavenoff and Zimm,
1973) are most simply explained by the conclusion that there is one uninterrupted DNA duplex per chromosome.
Interphase chromatin is visualized in the electron microscope as twisted 20-30nm fibres. At mitosis these fibres condense to yield the characteristic distinguishable structures of metaphase chromosomes.
Models to explain the compaction of the 20-30nm fibre have included further helical packing in a 'supersolenoid' (Bak et al., 1977) and a 'folded fibre' arrangement of repeated transverse and longitudinal folding of the 20-30nm fibre
(DuPraw, 1968; Comings, 1972). The organization of metaphase chromatin around a central backbone or core was first proposed by Taylor in 1958. Elegant work from the laboratory of Laemmli has confirmed this model. By competing out the histones using polyanions Laemmli and his coworkers have demonstrated that a protein scaffold is retained with loops of DNA emanating outwards, each loop
16 varying in length from 15-30um or approximately 45-90kbp
(Paulson and Laemmli, 1977). They have proposed the model that the histones coil up the DMA into the 20-30nm solenoid fibres which are attached to a non-histone protein scaffold (Laemmli et al., 1977). Using different procedures to isolate metaphase chromosomes the packing of the DNA loops is revealed. Electron microscopy of chromosomes isolated in the presence of divalent cations reveals the lOnm DMA duplex and in the presence of ImM
MgCl^ reveals the 20-30nm nucleosome solenoid. Chromosomes extracted in hexylene glycol reveal 50-60nm knobs and retain the dimensions of the intact metaphase chromosome.
These knobs are interpreted as supercoiled solenoids
(Marsden and Laemmli, 1979).
The determination of the metaphase chromosome structure, dividing the DNA into loops or domains directly reflects the structure of the lampbrush chromosome of meiotic cells (Miller and Hamkalo, 1972).
The proposal that interphase chromatin is also maintained in domains is strengthened by nuclease digestion studies of interphase nuclei (Igo-Kemenes and Zachau, 1977).
Centrifugation studies of lysed Drosophila melanogaster cells revealed supercoiled circular DNA structures of approximately 85kbp (Benyajati and Uorcel, 1976) and electron microscopic observations of mouse interphase nuclei suggest a non-histone protein skeleton with approximately 54kbp loops of DNA emanating outwards (Igo-
Kemenes et al., 1982).
17 Purification and reassociation analysis of human genomic DMA has led to our understanding of the DMA sequence arrangement in the chromosome. Human DMA sequences may be broadly classified into three classes
(Jelinek and Schmid, 1982). The first is the non- repetitive fraction present in less than thirty copies of each sequence and consisting of the majority of the coding and intervening sequences. The second type are the moderately repetitive sequences present in 50-1000 copies consisting of long and short sequences, some of which may be coding, interspersed with unique DMA. The third class are the highly repetitive sequences present in greater 4 than 10 copies and consisting of fairly short sequences in clusters. Dispersed repetitive sequences may be further divided into families of repeats which are often defined by common restriction sites. These type of sequences make up 15-30% of the genome by mass (Jelinek and Schmid, 1982) and their paradigm is the so-called Alu family which alone accounts for 3-6% of the genome (Houck et al.,
1979). Alu sequences have been found both between and within genes (Jelinek and Schmid, 1982). Other examples of dispersed repeats include the Hindlll repeat (Manuelidis and Biro, 1982), the 6.4kbp repeat found in the beta- globin cluster (Jelinek and Schmid, 1982), the Kpnl repeat
(Shafit-Zagardo et al., 1982) and the genes encoding the small nuclear RNAs (Jelinek and Schmid, 1982). Repetitive
DMA sequences are also found clustered, particularly the
18 highly repetitive sequences which make up the bulk of the heterochromatin found predominantly in pericentric and telomeric locations (Brutlag, 1980). Moderately repetitive sequences may also be clustered and examples include the so-called ocRI repeat (Darling et al., 1982) and the 2kbp
BamHI repeat (Yang et al., 1982; Willard et al., 1982).
In terms of function, repetitive DNA sequences remain an enigma. It has been proposed that they are pu rely parasitic and any function is incidental to their maintenance in the genome (Doolittle and Sapienza, 1980;
Orgel and Crick, 1980). An alternative hypothesis is that they regulate transcription or the processing of primary transcription products to mature messenger RNA molecules
(Davidson and Britten, 1973). An attractive possibility is that they organize the DNA into the domains seen by electron microscopy and biochemical techniques. This hypothesis is supported by the observation that repetitive
DNA is enriched in the insoluble fraction after nuclease digestion of interphase nuclei (Jeppersen and Bankier,
1979) .
Since the domain structure of chromosomes is apparently conserved throughout the cell cycle and the DNA sequence also exhibits an interspersed pattern, this has provoked the hypothesis that these domains are functional units. They may be units of replication (Hand, 1978) and/or transcription (Stadler et al., 1980). Indeed, with recent interest in the role that different configurations the DNA helix might adopt in distinguishing active and
19 *
inactive chromatin, a model for DMA domain activation has
been proposed (Rich, 1982). In this model the DMA loops
out in segments from a nuclear matrix/scaffold protein
which isolates the DMA for supercoiling between the
different domains. Individual domains are seen as inactive
when a key control region takes the so-called Z-DNA
configuration but become activated when they return to the
normal Watson-Crick B-DNA configuration. This activation
would be associated v/ith a change in superhelical density
leading to a non-random uncoiling and increased
availability of genes located in the domain to polymerases
and/or protein effectors.
1.2. Meiosis and the reassortment of genes
The evolutionary success of sexually reproducing
organisms is evident even if the selective forces
responsible for the origin and maintenance of eukaryotic
sex are not (Maynard-Smith, 1978). Oscar Hertwig first
discovered in 1875 that fertilization, the basis of sexual
reproduction, concerns the nucleus. Waldeyer's subsequent
cytological discovery of the chromosomes revealed that
they too were involved in fertilization. Our understanding
of the second essential characteristic of eukaryotic sex,
meiosis, springs from two suggestions made by Weismann
(Weismann, 1887). The first was that meiosis would be
2 0 found to compensate for fertilization in the life cycles of both sexes and all organisms. The other, added with the purpose to make a clean separation between Darwin and
Lamarck, was that the development of sexual reproduction in evolution depends on the value of meiosis in exposing the results of genetic recombination to natural selection.
The implications of this statement have led to the development of much of what are today the sciences of modern genetics and evolutionary biology.
The behaviour of the chromosomes during meiosis has been characterized by microscopic analysis. The four copies of each chromosome are distributed to different progeny cells by successive divisions. Preceding the first division is a lengthy prophase which, in some organisms, can last several months. In leptotene at the beginning of prophase, the chromosomes are seen as very fine extended threads which approach each other during zygotene and begin to pair side by side. At the end of leptotene, lateral components have been formed in the interval between two chromatids of unpaired chromosomes
(Westergaard and von Wettstein, 1972) and subsequently as partner chromosomes synapse in zygotene, starting at the telomeres (Callan and Perry, 1977), the lateral components associate with a central core to form the synaptonemal complex (Moses, 1958). Synapsis is complete by the stage of pachytene. The double nature of the synaptonemal complex is clearly revealed at diplotene when the chromosomes separate along much of their length except at
2 1 certain restricted points termed chiasmata. These
chiasmata are thought to mark the site at which genetic
exchange has taken place (Janssens, 1909). From
experimental alterations of cross over frequencies and
biochemical studies it has traditionally been held that
synapsis must be complete before exchange can occur. This
viewpoint has not remained unchallenged however (Maguire,
1977). The final stage of meiotic prophase is diakinesis
when the bivalents coil up to form shorter and thicker
chromosomes and the chiasmata have a tendency to move
towards the ends of the the paired homologues. This
movement is more pronounced in shorter chromosomes and may
be due to torsional strain induced by chromosome
condensation. Chiasma terrninalization may be complete,
partial or entirely lacking (Darlington, 1931).
The spindle is organized in prometaphase I and
pairing configurations line up on the equatorial plate in
metaphase I, to be distributed to opposite poles in
anaphase I. The second meiotic division which is
mechanically similar to mitosis then proceeds, a membrane
surrounding the four haploid nuclei at telophase II.
1.3. Recombination between linked genes and the
construction of linkage maps
The segregation and transmission of chromosomes at
22
» meiosis provided Sutton and Boveri with a physical manifestation of Mendelian inheritance (Sutton, 1903).
VJith the discovery of an increasing number of genes in
Drosophila melanogaster came the realization that some of these genes must reside on the same chromosome. When these genes were studied in nonallelic pairs or groups, they yielded ratios that frequently departed from random assortment. This led to the definition by Morgan of linkage or synteny in 1911 (Morgan, 1911) and of crossing over in 1912 (Morgan and Cattell, 1912). The proposition was that genes are arranged in a linear order, each gene having a fixed position on the chromosome. Sturtevant subsequently proposed that the extent of recombination between two genes was a measure of their distance apart on the chromosome and that recombination 'distances' were additive (Sturtevant, 1913). The unit distance corresponds to 1% recombination frequency and is often referred to as the centiMorgan (cM).
The pioneering work of Morgan, Muller, Bridges and
Sturtevant has led to the accumulation of genetic data on the frequencies of crossing over, thus enabling the construction of genetic or chromosome maps in which the serial order of the genes and their genetic spacing have been accurately determined.
The principles governing gene order, linkage and crossing over are undoubtedly applicable to analysing inheritance of Mendelian traits in humans. However, human families are small, the lifetime of the experimenter is
23 that of the subject and genetic manipulation is prohibited
on ethical grounds. Consequently a likelihood test for
linkage has been devised by Morton and is referred to as
the sequential or lods method (Morton, 1956). This yields
likelihood ratios as to whether the data obtained from any
particular pedigree are the consequence of recombination
occurring at a certain frequency (theta) or the
consequence of independent assortment. By giving theta in
the numerator of the equation a series of values from 0 to
50 and by keeping theta in the denominator at 50 ( a value
indicating no linkage), each pedigree can be analysed and
a series of likelihood ratios obtained. By making use of
the logarithm of the odds (Lod), the likelihood ratios
from different pedigrees can be summed and the
recombination frequency giving the highest lod value
determined. This value for
Lod=log^Q x likelihood of observed pedigree assuming 0=0 likelihood of observed pedigree assuming 0=50
theta is known as the maximum likelihood estimate and a
lod value of +3.0 is very strong evidence of linkage whilst a value of -2.0 is equally strong evidence of
independent assortment. Any value between would provide
only an equivocal estimate.
24 1.4. Crossing over and chiasmata
The proposal that the chromosomal basis of crossing
over lay in the chiasmata observed at diplotene originated
with Janssens in 1909. Stern, using Drosophila and
Creighton and McClintock in maize provided experimental
evidence that crossing over involved a physical exchange of chromatin by using homologues that could be
distinguished cytologically by the presence of
translocations (Stern, 1931; Creighton and McClintock,
1931). The first direct physical evidence in favour of Janssens' chiasmatype theory was provided by 3 autoradiographic analysis of H- thymidine labelled
Romalea microptera rneiotic chromosomes (Taylor, 1965) and
subsequently using BUdR labelled chromosomes in Locusta
migratoria (Tease and Jones, 1973). Chiasma frequencies have been used to estimate genetic distances in those organisms whose genetic map is inadequate. To what extent this is justified can be
determined from attempts to quantitatively correlate
chiasma frequencies and cross over frequencies. In the
lily, the exhaustive study of Brown and Zohary has
demonstrated a 1:1 correspondence of chiasmata to
chromosome exchange (Brown and Zohary, 1955). In maize, a comparison of the total map lengths of the known linkage groups to average chiasma frequencies approaches 1:1
correspondence as the chromosomes become more accurately
mapped (Neuffer and Coe, 1976). Observations on chiasma
25 i frequencies in human male spermatocytes (Hulten, 1974) have led to the estimation of 30 Morgans for the total map length of the human genome (McKusick, 1980).
Comparisons of genetic maps based on recombination and chiasma frequencies betv/een species reveals a decrease in recombination frequency per unit DNA with increasing genome size. It is tempting to speculate that this reflects a constant level of recombination per gene as the amount of DNA per gene appears to increase with increasing genome size.
Certainly it is apparent that the number of chiasma is under genetic control and may be selected for (Shaw,
1974). Studies on recombination in meiotic mutants of lower eukaryotes, Drosophila and plants (Baker et al.,
1976) support this observation. The intrinsic factors affecting the frequency of crossing over appear to be sex, age and genotype. Thus the frequency of crossing over may be identical between the two sexes (or in the anthers and ovaries of dioecious plants), as in the garden pea; higher in females than in males, as in the mouse, man and rat; or higher in males than in females as in the pigeon. Extreme sexual divergence in regard to frequency of exchange is also known. In male Drosophila melanogaster there is no crossing over and no synaptonemal complex formation, whilst the female silkmoth, Bombyx mori experiences no crossing over at meiosis (White, 1973). Where linkage differentials exist between the sexes and where adequate
26 data are available, it is the heterogametic sex in which crossing over is either lowered in frequency or abolished.
1.5. Localized recombination
The first demonstration that recombination does not occur randomly came from early studies in Drosophila which indicated that one cross over has a tendency to interfere with the occurrence of another in its vicinity (Muller,
1916). This corresponds to early analysis of the distribution of chiasmata along the chromosome demonstrating a significant departure from the Poisson distribution expected for random location (Haldane, 1931).
Mather first proposed that the location of the first chiasma to form on any one chromosome is relatively fixed whilst the second (i f any) is determined by an interference exhibited by the first (Mather, 1938) and work on locust chromosomes is consistent with this hypothesis (Fox, 1973).
There are many examples of localized chiasmata in nature. Virtually all chiasmata in the male meiosis of the grasshopper Stethophyma grossum occupy a procentric location whilst in the female greater than 94% of the chiasmata are located interstitially or distally. In the palmate newt Triturus helveticus male chiasmata have a proterrninal localization but in the female they are all intercalary. The opposite occurs in the Great Crested newt
27 Triturus cristatus where the chiasmata in the female are procentric in contrast to the liberal distribution found in the male (Callan and Perry, 1977).
The chiasma distribution for the human male has been described (Hulten, 1974) and it appears that chiasmata occur more frequently at a proterminal location (Laurie et al., 1982) and that interference extends for 3.5% of the total autosomal length although not at all across the centromere. Variation in crossing over frequency at different cytological locations has been well characterized. The classic example is the reduction of crossing over seen in the vicinity of the centromeres of Drosophila melanogaster. Translocations or inversions bringing euchromatic material adjacent to the heterochromatic centromeric regions bring about a reduction in crossing over frequency in that euchromatin. In addition, proximal markers can be induced to recombine more freely when translocated away from their procentric location (Schalet and Lefevre, 1976). Other local differences in crossover frequencies have been found. For example, in the distal portion of the left arm of the X chromosome in Drosophila the genes y and w are separated by 75-80 bands and recombine at a frequency of 1.5%. The same recombinational frequency is found between the genes w and _fa yet these are separated by only 4-5 bands. However, it is not clear whether the haploid DNA content per band is relatively
28 constant in Drosophila (Lefevre, 1971).
The study of meiotic mutants provides additional
evidence for the differing affinities of different regions
of the chromosome for recombination (Baker et al.# 1976).
The mutation rec-1 in Neurospora crassa causes a 10-25
fold increase in heteroalleiic recombination of the his-1
and nif-2 loci without any affect on 19 other markers. In
Drosophila/ a group of mutants (mei-41, mei-218, mei-251, mei-S282, mei-B, abo, mei-68^1^) decrease exchange most
severely in regions where crossing over per unit length is
relatively high in the wild type (distal intercalary
positions) and less severely where it is relatively low
(procentric and telomeric positions). A mutant has also been described that governs only the distribution and not
the overall frequency of recombination, mei-352, which
seems to allow a more random distribution of exchanges.
It is clear that a precise determination of the
position and frequency of crossing over in relation to the chromosome is facilitated by the genetic and chromosome maps available in Drosophila. In no other eukaryote has
such a comparison been possible. The extensive genetic map of the mouse has permitted the conclusion to be drawn that
the known genes do not follow the Poisson distribution
expected if they are evenly distributed in the genome and chiasmata form randomly (Lyon, 1976). However, it remains
unclear whether the discrepancy lies with the distribution of genes or chiasmata or both.
29 1.6. Cytological and genetic markers in man
In recent years two significant advances have changed the face of mammalian gene mapping. The first is the development of somatic cell genetics permitting the
isolation of chromosomes and the second the introduction of recombinant DMA techniques allowing the purification of
specific RNA and DMA sequences which can detect homologous genomic loci.
Production of interspecific cell hybrids can be achieved at will using inactivated Sendai virus or polyethylene glycol to mediate cell fusion (Ringertz and
Savage, 1976). Subsequent growth of human x rodent hybrids results in a continual disappearance of human chromosomes
(VJeiss and Green, 1967). However, specific human chromosomes will be retained if that chromosome carries a selectable marker in an appropriate rodent background.
Thus human x mouse hybrids can be compelled to retain the gene for hypoxanthine-guanosine phosphoribosyl transferase
(lIGPRT) by growth on HAT (hypoxanthine, aminopterin, thymidine) medium. This enzyme activity segregates with human glucose-6-phosphate dehydrogenase (G6PDH) activity and the presence of the human X chromosome (Nabholz et al., 1969) and thus facilitates the assignment of any human genes expressed in such hybrids to the X chromosome.
The use of human parent cells containing derivative X chromosomes such as translocations and deletions generates
HAT resistant hybrids containing differing amounts of X
30 chromosomal material (Ricciuti and Ruddle, 1973). This permits the intra-chromosomal localization of X-linked genes expressed in the hybrid.
The isolation of restriction enzymes which cleave at specific DNA sequences and ligases and polymerases which can link fragments together has facilitated the construction of hybrid or chimeric DNA molecules (Jackson et al., 1972). In this way, hybrid plasmid or bacteriophage DNA molecules can be obtained in large amounts by growth in bacteria and specific DNA sequences may be purified. For example, a 'library' of cloned sequences of the human genome can be constructed by linking human DNA, partially digested with a restriction enzyme, to bacteriophage DNA and growing the resultant recombinant phage in bacteria (Maniatis et al., 1978).
Many techniques now exist for purifying individual DNA or
RNA segments or molecules which code for a known product
(Maniatis et al., 1982). Cytological localization of genes is greatly facilitated by the availability of cloned DNA due to its ability to reassociate with its complementary copy in the human chromosome in a sequence specific fashion. A method for detecting specific sequences in the human genome has been devised by Southern (1975). Using this method with genomic DNA from a human x rodent somatic cell hybrid, the cytological location of the DNA complementary to the cloned probe fragment can be determined.
31 An alternative and powerful approach to the use of recombinant DMA probes in conjunction with somatic cell hybrids is the technique of 'in situ' hybridization (Gall and Pardue, 1969). /applicable only to the cytological localization of reiterated sequences initially, the technique has been refined for the location of DNA sequences present in a single copy in the human genome
(Harper and Saunders, 1981). Radioactively labelled DNA probes are used to reassociate with the chromosomal DNA of human metaphase spreads fixed to glass slides and hybrids detected by autoradiography. In this fashion, the chromosomal locus for a number of genes has been determined (Harper et al., 1982; Malcolm et al. , 1982;
Trent et al., 1982). This technique offers the advantage of a single high resolution localization, often to within a single chromosome band.
It is manifest that cloned DNA fragments can detect cytological loci present in a single copy in the genome.
The same fragments may also detect genetic loci as they can demonstrate DNA sequence variants on different chromosomes which will be inherited in a Mendelian fashion. Such variants are detected most readily as restriction digest fragments whose mobility in agarose gels is altered. This restriction fragment length polymorphism (RFLP) is generated by mutation leading to the loss or gain of restriction sites or insertion or deletion of DNA between two sites. It has been proposed that such RFLP loci can be used to generate linkage maps
32 of whole chromosomes and eventually the entire human genome (Botstein et al., 1930; Davies, 1981). Such chromosome linkage maps can be compared to cytological maps defined by the same DMA loci. In this way, the position and frequency of crossing over can be precisely determined in a fashion analogous to the Drosophila genetics of the first half of this century.
1.7. Flow cytometry and molecular cloning of the
human X chromosome
The human chromosome most amenable to linkage analysis is the X due to the hemizygous nature of the male and subsequent absence of detectable crossing over. The odds that any DNA fragment (from a 'library' of similar sized fragments) is derived from the X are approximately
18:1. This makes the construction of an X chromosome map from a total genomic DNA library impractical. The two most powerful approaches of making DNA libraries enriched for a particular chromosome are the use of interspecific somatic cell hybrids and the use of flow cytometry and flow sorting.
Isolating DNA sequences specific to a human chromosome from a somatic cell hybrid has been used for chromosome 11 (Gusella et al., 1980). The technique relies on the dispersed nature of most repetitive DNA sequences and their species specificity. Human DNA sequences in the
33 human x Chinese hamster hybrid were identified by reacting individual recombinants from a genomic library with radiolabelled total human DNA. Only the repetitive DNA sequences are in sufficient concentration in the probe to allow autoradiographic detection. Human single-copy DNA sequences found on the same recombinant molecule were isolated and used as a probe for intra-chromosomal localization.
Flow cytometry is a recent approach to cytogenetics that enables separation of metaphase chromosomes on the basis of their reaction with an appropriate flourochrome.
Chromosomes are stained in aqueous suspension and constrained to flow at high speed through a narrow laser beam that excites the stain. The photometrically measured emitted flourescence permits the human chromosomes to be sorted into twelve main groups (Carrano et al., 1979). The molecular cloning of DNA extracted from sorted chromosomes facilitates the construction of recombinant 'libraries' greatly enriched for particular chromosomal DNA sequences.
A chromosome preparation enriched for the X by flow sorting had its DNA extracted and digested to completion with the restriction endonuclease EcoRI. Fragments were ligated to bacteriophage lambda DNA and individual recombinant molecules packaged into infectious viral particles 'in vitro' (Davies et al., 1981). The 50,000 recombinants thus obtained provide a starting point for the construction of a cytogenetic map of the human X chromosome.
34 1.8. Summary and aims
The human chromosome contains a single, continuous
longitudinally arranged DDA molecule. This molecule
appears to be organized in domains with 'loops' of DtIA constrained to a non-histone protein matrix seen as a
scaffold structure at metaphase. Repetitive DNA sequences might play a role in the organization of such a structure, may be involved in gene regulation or indeed have no function other than self maintenance. The domain structure of the chromosome may be crucial to its function in the way genes are replicated, transcribed and recombine at m e i o s i s .
Little is known of the relation of the chromosome to meiotic recombination as genetic maps are extensive in only a few eukaryotic organisms, principally Drosophila melanogaster. It seems clear that the cytologically observed chiasmata are the sites of genetic exchange and evidence for non-random positioning of chiasmata is abundant; their frequency and positioning appears to depend on sex, age and genotype. Evidence for localized recombination is present only for Drosophila whose salivary chromosome bands have been correlated to transcription units. Recombination in the human can be investigated using cloned DNA sequences. These can detect chromosomal loci and reveal inheritable sequence variants.
A genomic library of 50,000 recombinants greatly enriched for the human X chromosome DMA provides the
35 starting material for this study. The cytological location
of individual DNA fragments is determined by the use of
interspecific somatic cell hybrids and 'in situ' hybridization. Genetic distances are estimated by
following the inheritance of variant restriction fragments
in pedigrees and summing information by the sequential or
lods method. In this way the frequency of crossing over can be determined and compared in different parts of the chromosome.
3G CH AP TE R 2
Materials and Methods
37 2.1. Materials
Ch em ic al s
With the following exceptions, all chemicals were
Analar grade supplied by British Drug Houses (BDH), Poole,
England. Sodium chloride, Trizma base, dextran sulphate, ethidium bromide, orange G and bromophenol blue were from the Sigma Chemical Co., St. Louis, Missouri, U.S.A.
Spermidine trihydrochloride was from Fluka AG, Buchs,
Switzerland. Ficoll and polyvinyl pyrrolidone were from
Pharmacia, Uppsala, Sweden.
Isotopes
All radioisotopes were supplied by Amersham
International, Amersham, England.
Chromatography, electrophoresis, transfer, and
hybridization
Sephadex was from Pharmacia. Polymer wool was from
Interpet, Dorking, Surrey. Agarose was supplied by Miles
Laboratories Inc., Kanakee, Illinois, U.S.A. or by
Bethesda Research laboratories (BRL), Rockville, Maryland,
U.S.A. Nitrocellulose filters were from Schleicher and
Schull, Dassel, West Germany. Bovine serum albumin (BSA) was from Sigma.
Photography and autoradiography
PanF film, Phenisol and ID-II developers, llypam fixer
38 and K2 and L4 nuclear track emulsions were from Ilford,
Basildon, England. Polaroid film came from Polaroid (UK)
Ltd., St. Albans, England. RX medical X-ray film and Each
II intensifying screens were from the Fuji Photo Co. Ltd.,
Tokyo, Japan. The UV light source was a short wave transilluminator from UV Products Inc., San Gabriel,
California, U.S.A. NTB-2 nuclear track emulsion and X-OMAT
XAR-5 film was from Eastman Kodak, Rochester, New York,
U.S.A.
Enzymes
EcoRI, Hindlll, TaqI, Bglll, SstI and BamHI were from
BRL. Additional EcoRI was a kind gift from Dr. L. VJoodhead in this laboratory. DNAase I, E.coli DNA polymerase I and proteinase K were from Boehringer Corp., Mannheim, West
Germany. RNAase A was from Sigma. Nick translation kits were supplied by Arnersham.
Nucleic acids
Herring sperm DNA and polyadenvlic acid were from
Sigma Chemical Co. Deoxyribonucleotides were from
Boehringer. Bacteriophage lambda DNA was from BRL.
Bacteriological reagents
Agar v/as from Difco Laboratories, Detroit, Michigan,
U.S.A. Yeast extract and tryptone were supplied by Oxoid
Ltd., Basingstoke, England. Sterile 90mm petri-dishes were
39 from Sterilin Ltd., Teddington, England. Bio-assay dishes were from Nunc, Roskilde, Denmark.
Tissue culture reagents
RPMI-1640 medium, foetal and new born calf serum and non-essential amino acids were from Gibco Europe Ltd.,
Uxbridge, England. Glutamine, kanamycin, colchicine and heparin were from Sigma. Phytohaemagglutinin was from
Wellcome Diagnostics, Dartford, England. Tissue culture grade sterile plastic flasks were from Sterilin Products.
Cytogenetic reagents
Lipsol detergent was from Lip (Equipment and
Services) Ltd., Shipley, England. Phosphate buffer tablets, Giemsa and Leishmann's stain were from BDH.
Microscope slides and coverslips were from Chance Propper
Ltd., Warley, England and Holdtite rubber adhesive was purchased from F.W. Woolworths Ltd., England.
Microscopy
All brightfield optics used a Reichert Polyvar equiped with a lOx eyepiece, lOOx objective and automatic camera all from C. Reichart, Vienna, Austria.
40 2.2. Bacterial and phage strains
LE392 and LE831 are derivatives of ED803 (hood,
1966). Their genotypes are:- — + + LE392 : rk mk recA supE supF gal met — + — LE831 : rk mk recA supE supF gal met
Plasmid recombinants were propagated on strains
HB101, (Boyer and Roulland-Dussoix, 1969) or MC1061
(Casadaban and Cohen, 1980).
Their genotypes are:-
HB101 : s t rr rB mB recA pro gal r _ + + M C I 061 : str rB mB recA ara leu
All bacteriophage were human recombinants in the vector XgtV/ES. "Xb (Tiemeier et al., 1976) v/hose genotype
is
"X nin5 att-red Wam403 EamllOO SamlOO
Plasmid recombinants LI.23 (Davies et al., 1983), 52A
(Drayna et al., 1984) and DP31 (Page et al., 1982) were all in pBR322 .
2.3. General methods
2.3.1. Restriction enzyme digestion
Digests were set up with the appropriate amount of
DMA, a tenfold concentrate of buffer specified by the manufacturer, 4mH spermidine trihydrochloride, 100 ug/ml nuclease-free BSA, enzyme and double distilled water to
41 bring the DMA concentration to between 0.1 and 0.2mg/ml.
For bacteriophage DMA, 1 unit of enzyme was added per ug
of DMA whilst for human, rodent or somatic cell hybrid
DMA, 2.5 units per ug were added. Digests were performed
in 1.5ml snap-shut eppendorf tubes at the temperature
specified by the supplier. After an incubation time of 1-
2hrs per ug of DMA the digest was placed on ice and a
small sample electrophoresed on an agarose gel to
determine the extent of reaction. Completed digests were
electrophoresed immediately or following phenol and
chloroform/isoamyl alcohol extraction, ethanol
precipitated and re-suspended in 10mM Tris-HCl pH7.5, ImM
EDTA, (TE buffer), at a concentration of 0.4mg/ml.
2.3.2. Ethanol precipitation of nucleic acids
DMA was routinely precipitated by the addition of 0.1 volume 3H sodium acetate plus 2.5 volumes of ethanol.
Bacteriophage and plasmid DMAs were chilled for 2hrs- overnight at -20°C and pelleted by centrifugation for lOmins in an eppendorf inicrofuge (Eppendorf 5412). The supernatant was removed and the pellet washed in 70% ethanol. After re-spinning, the supernatant was removed and the pellet dried under vacuum. High molecular weight
DMAs were lifted directly with a closed pasteur pipette and rinsed in 70% ethanol. The precipitate was allowed to dry in air before redissolving in an appropriate volume of
TE buffer, as were dried pellets.
42 2.3.3. A-garose gel electrophoresis
Agarose gels for molecular weight analysis and/or
Southern transfer were electrophoresed in E-buffer (40mM
Tris, 20mM sodium acetate, 2mM EDTA, pll7.7) containing
0.5ug/ml ethidium bromide. Gels were made by dissolving
the appropriate weight of agarose powder in E-buffer and
0.5ug/ml ethidium bromide and boiling the solution. The
solution v/as allowed to cool before pouring into a gel
mould. After the gel had set, the slot formers were
removed. Samples were loaded in 2% w/v Ficoll plus an
appropriate amount of orange G as a tracking dye.
Electrophoresis v/as carried out in an electric field of 2-
5v/cm until the tracking dyes reached the bottom of the
gel. DMA fragments were visualized by illumination with
short-wave UV light and gels photographed on Polaroid film
using a Polaroid MP-4 camera.
2.3.4. Autoradiography
Samples were autoradiographed at -70°C using
intensifying screens. Fuji film was sensitized by pre- flashing to an O D ^ j-q °f 0.1-0.3.
2.3.5. Gel filtration
Gel filtration was used to remove unincorporated radiolabelled nucleotides from nick translation reactions.
A pasteur pipette was plugged with polymer wool. A matrix of Sephadex G-50 was poured into this pipette and rinsed with 3 column volumes of 2 x SSC (SSC is 0.15M NaCl, 0.15M
43 Ma citrate). The sample was applied to the column in lOOul volume and the column washed with more 2 x SSC. Fractions were collected in 1.5ml eppendorf tubes, monitored and the excluded peak of radiolabelled DMA pooled.
2.3.6. Determination of radioactivity 32 P radioactivity was measured by Cerenkov counting at an efficiency of approximately 30%. 3 H radioactivity was measured by counting in toluene- based scintillant (0.5% w/v 2,5 diphenyloxazole [PPO],
0.03% w/v l,4-di-2-[5 phenyloxazolyl] benzene in toluene) diluted 4:1 in Triton X-100.
2.4. Specific methods
2.4.1. Preparing high molecular weight DNA from tissue
culture cells
Tissue culture cells were centrifuged for 5mins at
3000rpm in a Sorvall SA600 rotor and carefully resuspended in phosphate-buffered saline (PBSA - 0.15M NaCl, 3mM KC1,
4mM Na HPO^, , 1.5mM KH PO^) . Cells were washed a further time in PBSA and finally resuspended in 0.15M NaCl, 0.5% w/v sodium dodecyl sulphate (SDS) and lOOug/ml Proteinase
K. Cells were incubated overnight at 37°C then allowed to cool and extracted twice with phenol saturated with TE buffer and once with chloroform/isoamyl alcohol (24:1)
44 before ethanol precipitation and re-dissolution overnight o at 4 C. DMA concentrations were estimated by 2 GO measurements using a Cecil CE595 dual beam
spectrophotometer.
2.4.2. Small scale preparation of bacteriophage DNA
All host bacteria were harvested from an exponential
culture in L-broth (5g/l yeast extract, 10g/l tryptone,
5g/l NaCl, pH7.2) supplemented with lOmM MgSO^ and 0.4% w/v maltose and stored in lOmM MgSO^ at 4°C. Suitable
dilutions of the 'X-library' bacteriophage mix were pre adsorbed onto approximately 10^ bacteria and plated on tryptone plates (12g/l tryptone, 5g/l NaCl, llg/l agarose)
in a lawn of tryptone top agarose (12g/l tryptone, 5g/l
NaCl, 7g/l agarose) and incubated overnight at 37°C.
Single plaques were picked with a pasteur pipette into 1ml of TM buffer (lOmM Tris-HCl, pH7.5, 2mM MgCl^) and o 8 incubated for 4hrs-overnight at 4 C. Approximately 10 cells were preadsorbed v/ith 0.05-0.lml of this solution and plated in a lawn of tryptone top agarose on tryptone agarose plates. After overnight incubation at 37°C the plates which were almost confluently lysed were overlaid with 5ml of TM and the phage allowed to elute for l-2hrs at 4°C. The phage solution was transferred to 13ml centrifuge tubes and bacterial debris removed by centrifugation at 5500rpm in the Sorvall SAGOO rotor for lOmins at 4°C. The supernatant was digested with RNAase A and DNAase I at concentrations of lug/ml for 30mins at
45 o 37 C. Bacteriophage particles were precipitated by addition of an equal volume of a solution containing 20% w/v polyethylene glycol (PEG), 2M WaCl and incubation at
4°C for 2hrs-overnight. The precipitate was recovered by centrifugation at 6000rpm for 20mins at 4°C and removing the supernatant by aspiration and air-drying. Phage were resuspended in 0.5ml of TM by vortexing and debris removed by centrifugation at 5500rpm for 2mins at 4°C. The supernatant was transferred to a 1.5ml eppendorf tube to which 5ul of 10% w/v SDG and 5ul of 0.5M EDTA (pH8.0) were added. The solution was incubated at 68°C for 15mins to rupture the phage particles, allowed to cool and extracted once with phenol saturated with TE and once with chloroform/isoamyl alcohol (24:1). DMA was ethanol precipitated from the supernatant and resuspended in 50ul
TE. lOul of this solution was removed to a fresh eppendorf tube with 1.2ul restriction enzyme buffer and 1-2 units of enzyme. Fragments were analysed by agarose gel electrophoresis.
2.4.3. Large scale preparation of bacteriophage DMA
The method used to obtain larger amounts of bacteriophage DMA is essentially that described by
Blattner et al., (1977). A phage solution of a single plaque picked into 1ml of TM buffer was preadsorbed onto 8 approximately 2.5 x 10" cells m a lOmM MgSO^ suspension at 37°C for 15mins. These were then inoculated into 200ml
46 of L-broth supplemented with lOmM MgSO^ and 0.4% w/v
maltose and shaken vigorously overnight at 37°C.
Concomitant growth of bacteria and bacteriophage results
in lysis of the culture which was completed by addition of
5ml of chloroform. Bacterial debris was removed by centrifugation in 250ml GSA bottles in the Sorvall GSA
rotor at 8000rpm for 15mins at 4°C. 2ml MgC^# 7g NaCl and
25g PEG were added to the supernatant, dissolved and left overnight at 4°C for the phage to precipitate. PEG precipitates were recovered by centrifugation in the GSA
rotor at 7500rpm for 20mins at 4°C. The precipitate was air dried and taken up in 2ml of TM buffer. This high titre phage suspension was carefully layered onto a caesium chloride step gradient of 2.5ml of 1.7g/ml, 3ml of
1.5g/ml and 4ml of 1.3g/ml CsCl in a 13ml cellulose nitrate centrifuge tube. Phage were banded by ultracentrifugation in the Sorvall SW-40 swing out rotor at 32,000rpm for 2hrs at 20°C. Phage bands were extracted using a syringe and dialysed overnight at 4°C against TM buffer. The phage DNA v/as extracted by proteinase K digestion for 3hrs at 55°C followed by phenol and chloroform extraction, ethanol precipitation and re dissolution in TE buffer.
2.4.4. Labelling DMA by nick translation
The procedure for radio-labelling recombinant DNA probes v/as essentially that described by Rigby et al.
(1977). Trittiated deoxyribonucleotides came in 50%
47 ethanol and were dried down by vacuum and re-constituted 3 in 5ul of 11^0 by vortexing. II nick translation included 3 3 3 0.5ug of DNA, 50uCi each of H-dATP, H-dCTP and H-dTTP,
5mM MgC^/ 50mM Tris-HCl pH7.5, lOmM yS -mercaptoethanol,
4uM dGTP/ 5 units of endonuclease free E.coli DNA
polymerase I and 5-50pg of DNAase I in a 50ul volume. The
amount of DNAase I to be added was determined by sizing
reaction products on an alkaline agarose (denaturing) gel
and using the amount that gave a mean fragment size of
approximately lkbp. Incubation was for 150mins at 15°C
after which the reaction was quenched by gel filtration. 32 P nick translations were essentially similar except 3 2 25uCi of P-dCTP (>400Ci/mmol) was used as label and 4uM
dATP, dGTP and dTTP were used in the reaction.
Alternatively reactions used the Amersham nick translation kit. lOul of solution I and Sul of solution II were added 32 to 0.5ug DNA and 25uCi P-dCTP and water to a final volume of 50ul. Incubation times and gel filtration were
i d e n t i c a l .
2.4.5. Screening lambda recombinants
In situ hybridization to phage plaques v/as performed after the method of Benton and Davies (1977). The ' X- 2 library' was plated on 400cm bio-assay dishes in tryptone top agarose and plates v/ere chilled to 4°C. Sterile nitrocellulose filters were labelled with pencil marks, soaked in 3 x SSC for 30mins and blotted dry. Filters were
48 laid onto the surface of the plates, the first for 30secs,
the second for lmin, the third for 2mins and the fourth
for 4mins. Filters were then laid phage side uppermost for
3mins on a pad of Whatman 3MM paper soaked in 0.5M NaOH,
1.5M NaCl to denature phage DNA. Filters were then
submerged twice (30 secs each) in 0.5M Tris-HCl pH7.5,
1.5M NaCl and blotted dry. Filters were wrapped between
sheets of Whatman 3MM paper and baked for 2hrs at 30°C.
Filters were soaked for 30mins in 3 x SSC at 65°C and
overnight with shaking in 3 x SSC, 10 x Denhardt's
solution at 65°C. 1 x Denhardt's solution is 0.02% w/v
Ficoll, 0.02% w/v polyvinyl pyrrolidone and 0.02% w/v BSA
(Denhardt, 1966). The final pre-hybridization v/ash was for
2hrs in 3 x SSC, 0.1% w/v SDS, 0.1% w/v sodium
pyrophosphate, 10% w/v dextran sulphate, lOOug/ml
sonicated denatured herring sperm DNA, 5 x Denhardt's
solution at 65°C with shaking. The same solution was used
for hybridizations. Filters were transferred to hybridization chambers designed by A. Jeffreys, Leicester
University, to which was added 5-10ml of hybridization 32 . buffer and P-labelled probe DNA denatured by immersing
in boiling water for 5mins. Hybridization chambers were
left shaking overnight in a 65°C v/ater bath after which
filters were removed and washed 5 times in 3 x SSC, 0.1%
SDS, 0.1% sodium pyrophosphate for 5mins each at 65°C with
shaking. Filters were then transferred to 1 x SSC, 0.1%
SDS, 0.1% sodium pyrophosphate for two washes at 65°C of
30mins each. Filters were then blotted dry, covered in
49 clingfilm and transferred to X-ray cassettes for
autoradiography.
Phage which reacted positively with probe DMA were
picked with a pasteur pipette and propagated as plate
lysates. 2ul of each lysate were spotted on an ordered
array of 5 x 4 spots on a bacterial lawn and grown
overnight. These plates were then rescreened with the same
probe DMAs.
2.4.6. Southern transfers
Genomic or recombinant DMA for Southern transfers was
fractionated on 0.8% agarose gels. A hybridization marker
of lOOpg of Ilindlll digested lambda DMA was
electrophoresed in parallel. In addition visible size
markers of lug of Hindlll digested lambda DMA was run on
the same gel. This gives fragment markers of 23.1, 9.4,
6.6, 4.4, 2.3, 2.0 and 0.6 kbp. The lambda 'arms' of
recombinant phage hybridize to the largest four fragments.
After electrophoresis gels were photographed and denatured
for 2hrs in 0.5M NaOM, 1.5M MaCl. Gels were then
neutralized in 0.5M Tris-HCl pH5.5, 3M MaCl for 2-4hrs
with several changes. DMA was blotted overnight in 20 x
SSC onto a nitrocellulose filter by capillary action using
the apparatus described by Southern (1975). The
nitrocellulose filter was washed for 20mins in 2 x SSC,
blotted dry and baked for 2hrs at 80°C. Filters were pre hybridized and hybridized as described in section 2.4.5.
50 with an additional post-hybridization wash in 0.1 x SSc,
0.1% SDS, 0.1% sodium pyrophosphate for 30mins at 65°c
with shaking as necessary.
2.4.7. Competing repetitive sequences in Southern blot
hybridizations
The method used to reveal the cognate band in a total
genomic Southern blot probed with a cloned fragment
including a repeated sequence is based on that of Davies
et al.(1933B). Total human DMA was sheared to an average
length of 2kbp by sonication. The sheared DMA was mixed 32 with p-labelled probe DMA to a final concentration of
O.lmg/ml in 3 x SSC. Driver and probe DMA v/ere denatured together by boiling for 5mins and then transferred
immediately to 65°C for an incubation of 2-4hrs. The DMA was then transferred to hybridization chambers to which hybridization buffer constituents had been added to give a final volume of 5-10ml. Hybridization and post hybridization washes were as described in 2.4.5.
2.4.8. Preparation of metaphase spreads
The method for preparing slides of human metaphase chromosomes was essentially that of Moorhead et al.
(1960). 10ml of blood were collected in a heparinized tube and the red cell layer allowed to settle. 0.3-0.4ml of the white cell layer was added to 10ml of RPMI-1640 medium supplemented with 10% v/v foetal calf serum and 0.1ml of a solution of phytohaemagglutinin. 2-3 drops of the red cell
51 layer were added before careful mixing and incubation overnight at 4°C. Cultures were then transferred to 37°C and incubated for 72hrs, mixing daily by inversion. 0.15ml of a 25ug/ml solution of colchicine was added 90mins before removal from the 37°C incubator. Cells were harvested by centrifugation in a Sorvall bench centrifuge at 1500rpm for 5mins. Cells were carefully resuspended with constant agitation in 0.075M KC1 prewarrned to 37°C and then incubated for 5mins at 37°C. Osmotically swollen cells were recovered by centrifugation at 1500rpm for
5mins and fixed v/ith 10ml of freshly prepared pre-chilled
(on ice) fixative (1 part glacial acetic acid to 3 parts methanol) by careful resuspension with constant agitation.
After leaving at 4°C for at least 30mins, cells were pelleted by centrifugation at 1500rpm for 5mins and washed twice in freshly prepared fixative before resuspending finally in 0.5-2ml. This suspension was spread on slides which had been cleaned by soaking overnight in chromic acid and rinsing under the tap for at least 30mins. 1 or 2 drops were applied to the centre of the wet slides and allowed to spread before drying on a hot plate. Spreading was monitored by phase contrast microscopy of dried s l i d e s .
2.4.9. G-Banding of metaphase chromosomes
Two methods of banding were employed, the first using trypsin/Giemsa (Seabright, 1971) and the second
52 Lipsol/Leishmann' s (Malcolm et al., 1982). Trypsinization
of slides was achieved by soaking slides for 30secs-
2.5mins in a 1:14 dilution of Trypsin in 0.9% w/v NaCl at
4°C. Slides were then rinsed thoroughly in 0.9% w/v NaCl,
air dried and stained in coplin jars with 5% w/v Giemsa in
phosphate buffer pH6.8 for 40mins. Well banded metaphase
spreads were recorded by noting microscope vernier
positions and photographed. The alternative and preferred
banding procedure was to immerse slides for 5-60secs in a
1:100 v/v dilution of Lipsol detergent in water, rinse off
with 0.9% NaCl and stain by covering with 5ml of
Leishmann's stain diluted 1:5 in phosphate buffer, pH6.8.
Leishmann's stain was made by dissolving 1.5g of powder in
a litre of acetone-free methanol, stirring for 2hrs and
filtering through 2 layers of V7hatman 3MM filter paper.
Slides were stained for 2.5mins and then rinsed with
phosphate buffer, pII6.8 and dried on a hot plate. Well
banded spreads were recorded and photographed.
2.4.10. 'In situ' hybridization
The area of spreads was marked with a diamond pencil
and photographed slides were destained by successive dehydration in 50%, 75%, 95% and 100% ethanol for 2mins
each and allowed to dry. Slides were then floated in trays in a 37°C waterbath and allowed to warm up before placing
0.2ml of a solution of RNAase A (O.lmg/ml in 2 x SSC) onto each slide and covering with 24 x 50mm coverslips previously cleaned in 0.1M IIC1 and distilled water. After
53 an incubation of lhr the coverslips were removed and the
slides rinsed three times in 2 x SSC for 2mins each before
dehydration in 50%, 75%, 95% and 100% ethanol for 2mins
each. Slides were then air dried and returned to trays in
a 70°C waterbath and allowed to warm up. 0.2ml of
denaturing solution (70% v/v forrnamide/2 x SSC) was then
placed onto each slide under acid cleaned 24 x 50mm
coverslips. After 2mins of incubation slides were again
rinsed three times in 2 x SSC, dehydrated and air dried.
Trittiated, nick translated probe was prepared as
described in 2.4.4. Pooled excluded fractions were ethanol
precipitated, re-dissolved in 20ul of formamide hybridization buffer and counted. Formamide hybridization buffer is 50% v/v formamide/2 x SSC/10% w/v dextran
sulphate/5 x Denhardt's solution/0.lmg/ml sheared, denatured herring sperm DNA/0.1% w / v sodium pyrophosphate.
By correlating the recovered radioactivity with the
initial excluded fractions, the labelled DMA was diluted to a concentration of 0.1-10ug/ml in formamide hybridization buffer. lOul of this dilution was then spotted onto slides and covered with acid cleaned 22 x
22mm coverslips ensuring that no air bubbles became trapped underneath. The edges of the coverslip were sealed with Holdtite adhesive and slides were incubated on trays in a 42°C waterbath overnight. The adhesive was removed with tweezers and coverslips were soaked off in 2 x
SSC/0.1% sodium pyrophosphate. Slides were washed 5 times
54 in 2 x SSC/0.1% sodium pyrophosphate for 2mins each and 3
times in 0.1 x SSC/0.1% sodium pyrophosphate for 2mins
each at room temperature. Slides were then transferred to
0.1 x SSC/0.1% sodium pyrophosphate at 55°C for two 30mins
washes and finally washed extensively in 0.1 x SSC/0.1%
sodium pyrophosphate overnight with stirring at 4°C.
Slides were dehydrated, air dried and coated v/ith a
subbing solution of 0.1% w/v gelatin, 0.01% w/v chromalum
by dipping for 30secs. After allowing to dry, slides were
coated with a layer of emulsion (Ilford K2 or L4 or Kodak
MTB-2) diluted 1+1 with water at 50°C by dipping for
approximately 20secs. Emulsion was wiped from the back of
the slides before they were dried vertically in test tube
racks in the dark room for l-2hrs. Autoradiography was
performed in light-tight boxes with dessicant at 4°C for
10-22 days. Slides were developed in Ilford Phenisol diluted 1+4 with water for 5mins, rinsed in water and
fixed in Ilford Hypam diluted 1+4 with water for 3mins.
After rinsing under the tap for 30mins, slides were air dried and stained with Leishmann's stain diluted 1+4 v/ith phosphate buffer, pH6.8, for 3mins.
2.4.11. Scoring hybridized spreads
Recorded spreads were compared after autoradiography to the photograph of the G-banded spread before hybridization. The position of grains on the autoradiograph v/as marked directly onto the G-banded karyotype of each spread and totals for each chromosome
55 computed by summation over all recorded spreads. This
observed value was compared to an expected value computed
by multiplying the total number of chromosomes over all
scored chromosomes by the diploid percentage DMA content
of each individual chromosome. For a random distribution
of grains the deviation of each observed chromosomal grain
value from the expected (0-E) will be normally distributed
about a mean of zero. Consequently the significance of
each result can be estimated by comparing the number of
standard deviations each 0-E value takes to the
standardized normal distribution.
Grain distribution was also displayed
diagrammatically by using chromosome ideograms to display
composite distribution. The position of each grain as
marked on the G-banded karyotype was marked on a single
karyotype ideogram. In this way the total number of grains
over individual G-bands or interbands and hence the intra-
chromosomal distribution was also determined.
2.4.12. Pedigree analysis
Linkage between markers was estimated by the
sequential or lods method (Morton, 1956). Recombination
fractions for any pair of loci were estimated by
interpolation after summing the lod scores over all the families determined from a standard lod score table (Race and Sanger, 1975). Additionally, lod scores were computed by using the program LIPED (Ott, 1974).
56 CHAPTER 3
Results
57 3.1. Screening recombinant bacteriophage 'in situ'
The majority of non-repetitive human DMA is organized
in discrete blocks interspersed with repetitive elements.
These repeated DMA sequences are only weakly conserved between species and consequently we predicted that
recombinant phage from the genomic library enriched for
the X would react with a total human repetitive DMA probe
and only weakly with a mouse probe. Therefore
reassociation of a repetitive DMA probe from a human x mouse somatic cell hybrid whose sole human chromosome was
the X with recombinant phage could be attributed to repetitive sequences present on the human X chromosome.
Similarly, experiments utilizing radiolabelled somatic cell hybrid DMAs containing different amounts of human X chromosomal DMA should delineate recombinant phage specific to a particular portion of the X chromosome.
A few hundred plaques from the X enriched library were transferred to nitrocellulose by the Benton-Davis 32 procedure and incubated with P-labelled total genomic
DMAs. Under these conditions, only the repetitive DMA in the probe is in sufficient concentration to allow autoradiographic detection of reassociated DMA. To ensure the reaction was species specific, an initial comparison was made between human and mouse DMAs as probe. IIorl9X is a somatic cell hybrid containing the X as its sole human chromosome (Goodfellow et al., 19H0) and is derived from the mouse cell line IR Radiolabelled DMA from male human blood and from these cells was hybridized to duplicate
Benton-Davis filters (see Materials and Methods). Fig. 1.
shows the autoradiographs after a second screening of
recombinants picked from the first screening for giving
1., a weak signal with Horl9X DMA and with mouse, 2., a
weak signal with Horl9X DMA and none with mouse, 3., a
strong signal with Horl9X DMA and a weak signal with mouse
and 4., a strong signal with Horl9X DMA and none with
mouse. The spots on the autoradiographs corresponding to
recombinants DXR1.5 and DXR4.4 are blank as these
recombinants gave a zero titre on this plate. There were
no recombinants that gave a signal with mouse but not
human DMA as the probe. The probes in each secondary
screen were in A; human DMA, B; mouse DMA and C; Horl9X
DMA. The result shows that none of these recombinants
reacts as strongly with horl9X DMA as with human DMA. This
can be attributed to two factors, firstly that the X
chromosome concentration in each probe is different; the hu ma n iprobe has one X chromosome in 46 hu ma n chromosomes whilst the IIorl9X probe has one X chromosome in a background of approximately 60 mouse chromosomes (Davies
et al . , 1981). Secondly, the concentration of individual
repeat families will be higher in the human probe due to
the presence of autosomal copies of each repeat. The
result additionally shows that IIorl9X DMA always hybridizes to recombinants the human DMA reacts with whilst the mouse probe does not. This confirms the species
59 c
Fig. 1 Recombinants from the 'X - enriched* library- screened with total genomic DNA from : A - Human blood; B
- mouse fibroblasts; C - Horl9X fibroblasts
60 specificity of many repetitive sequences and implies a good representation of X chromosomal DMA sequences in the
library. That conserved repeats also exist is demonstrated by the recombinants which cross react with mouse DMA. The pattern is confirmed by taking the individual recombinants and hybridizing to Southern (1975) filters of restriction endonuclease digested hybrid DMAs. Autoradiographs show an
intense smear over the human, Horl9X and mouse DMAs for
the recombinants DXR1.3. and DXR1.4. but only on the human
and Morl9X tracks with the recombinants DXR2.3. and
DXR2.5. (Results not shown). The human x mouse somatic cell hybrid MCP-6 retains
the translocation chromosome t(S;X )(6qter-p21::Xql3-qter)
as its only human chromosome (C-oodfellow et al., 1902). Me
reasoned that a comparative screening between total DMA
from this hybrid and that from Horl9X should delineate
repeats on the X chromosome from Xpter-ql3 on the basis of
their reaction with IIorl9X and not MCP-6 DMA. Duplicate . . . 39 filters were hybridized overnight to ^P-labelled total hy br id DM a at concentrations adjusted to compensate for
the different ploidies of the hybrids. Filters were washed
to a stringency of 1 x SSC at 65°C and exposed to Fuji X-
ray film with an intensifying film at -70°C for two weeks.
Fig. 2. shows the autoradiographs which demonstrate no
visible differences. Attempts to superimpose negatives
failed to demonstrate any differences. This result implies
there are no repetitive sequences in the X-enriched
library which are unique to the region Xql3-pter.
61 Fig.2 Two duplicates of one quarter of a 20 x 20cm nitrocellulose filter blotted to a bio-assay dish of several thousand recombinants from the 'X-enriched library'. The probe in each hybridization is total DNA from A - Horl9X;B - MCP-6 cells.
62 3.2 Isolation and characterization of
individual recombinant bacteriophage
To facilitate the identification of X chromosome DMA
sequences it is important to isolate them from any
adjacent reiterated sequences which may have become cloned
within the same recombinant molecule. Consequently we
reasoned that recombinants on Benton-Davis filters which
did not hybridize to total human DMA would contain only
non-repetitive sequence and therefore be most suited to
further characterization. Unfortunately phage picked on
this basis largely turned out to carry DMA derived from
bacteriophage lambda inserted between the left and right
arms of the cloning vector, or no insert at all and an
abnormally large vector arm. As a consequence, the
stochastic approach of picking plaques at random from the
library was adopted. The rationale for this procedure is
that single copy sequence would almost certainly be
contained in the same molecule as repetitive sequence and
so a more representative collection of X chromosome DM.A
sequence is accomplished. Such interspersed non-repetitive
DMA can be detected by restriction enzyme digestion and
re-cloning or by competition in hybridization with a
repetitive probe.
Individual plaques were picked, propagated on a
suitable host strain and the DMA extracted from plate
lysates (see Materials and Methods). The DMA was digested i 63 Fig .3 Polaroid photograph of an a g arose gel with
EcoRI digested DNA f rom s everal d i ffer ent recomb inants
derived from the ' X- enriched ' library e lectrophoresed in
l E I-b u ffe r overnight . Lanes 1 and 14 are lambda DNA
d i gested with HindII I which g ives mo lecular weight marker
bands of 23.1, 9.4 , 6. 6 , 4 . 4 , 2 . 3 , 2. 0 and 0 . 6kbp . The
recombi nant s whose DNA has been f ractio nated a r e from
Lanes 2 - 13; DX l - 12.
• 64 by the restriction endonuclease EcoRI, fragments were
separated by agarose gel electrophoresis and visualised by
ethidium bromide fluorescence. Fig. 3. is a photograph of
a gel of EcoRI digests of the clones DX1-12. The largest molecular weight bands represent the DMA of the vector
phage ('arms') of 21.7kbp and 13.3kbp. The smaller bands
are the human DMA inserts ranging from 1.7kbp (DX9) to
8.4kbp (DX7). The clones DX1 and DX3 have no detectable
DMA insert and a concomitant extended long 'arm'. The
presence of more than one insert in DX9 suggests either that the EcoRI digestion of sorted chromosome was
incomplete or that fragments had religated together prior
to ligation to bacteriophage lambda DMA..
The reiteration frequency of human DMAinsert
sequence was examined by Southern transfer of gels such as
Fig. 3. to nitrocellulose and overnight hybridization to
3 p total human DMA labelled with “P by nick translation (see
Materials and Methods). Fig. 4. shows the autoradiographs of such an experiment. In A the filters have been washed
to a stringency of 3 x SSC at 65°C, in P> to 1 x SSC at
65°C and in C to 0.1 x SSC at 65°C(see Materials and
Methods). The lambda DMA bands are retained on the
autoradiographs to act as markers by inclusion of a trace 32 . amount of P-labelled lambda DMA m the hybridization
reaction. It is known that reassociated repetitive DMA has
a wide melting range (Britten and Kolme, 1968) and by
reducing the salt concentration and hence the Tm of
65 X DX12 3 4 5 6 7 8 9 10 11 12 X
XDX1 2 3 4 5 6 7 8 9 10 11 12 X
Fig.4 - A Southern blot hybridization of DMA from recombinants DX1 - 12 hybridized to both lambda DMA and total human DMA. The filter was washed after hybridization to a stringency of 3xSSC. B - The same filter after washing to a stringency of IxSSC.
6 5 X DX123 4 5 6 7 8 9 10 11 12 X
X DX1 2 3 U 5 6 7 8 9 10 11 12 X
Fig.4 - C The same filter in Fig.4A washed to a stringency of O.lxSSC. D - A duplicate filter hybridized only to lambda DMA.
67
( reannealed Did A (Britten et al. / 1974) it is possible to selectively melt off imperfect hybrids. Thus in Fig. 4 . A
and 8, DX4 and DX7 still reanneal to a significant component of human DMA but at a stringency of 0.1 x SSC at
65°C/ hybrids are not formed, as shown in Fig. 4C. These
recombinants are candidates for non-repetitive DMA
sequences, whilst those such as DX2 or DX5 clearly contain
repetitive DMA sequences. The result of hybridizing lambda
DMA alone to a duplicate filter is displayed in Fig. 4D
and clearly only the high molecular weight bands of the vector arms are detected. The smear of radioactivity at
the bottom of the autoradiograph represents either vector
DMA degraded by nucleases or bacterial DMA reacting with trace amounts in the probe.
In total, 23 clones were examined in this fashion. A summary of the data from these clones is presented in
Table 1. Six of the 23 do not contain any human DMA and
the average insert size of the remaining 22 is
approximately 6.5kbp. In total the cloned DMA examined
amounts to 142kbp.
3.3. Chromosomal localization I. Somatic cell hybrids
Having isolated human DMA sequence from the X
enriched genomic library it is necessary to show that it
is derived from the human X chromosome. The most practical
method to employ is the use of somatic cell hybrids which
68 Table 1
Recombinant DNA Insert (kbp) Hybridization to total human DNA
DX1 no insert, extended long arm DX2 7.2 + DX3 no insert, extended long arm DX4 4.1 DX5 3 .8 + DX6 7.7 + DX7 8.4 — DX8 6.1 + DX9 5.2, 1.7 + DX10 7.9 + D X 1 1 5.9 + D X 1 2 5.9 + DX13 2.2, 1.8 - DX14 8.0 + D X 1 5 6.5 - D X 1 6 lambda B insert D X 1 7 5.5 + DX18 6.2 + D X 1 9 lambda insert DX20 7.3 + D X 2 1 lambda inserts DX22 6.6 + DX2 3 6.7 + DX24 7.2 — DX2 5 5.6 + DX3 3 7.4 + DX48 7.2 + RE 3 lambda B insert
Summary of results obtained from the analysis of
randomly isolated recombinants from the X-enriched
library. Those recombinants marked + retained an
autoradiographic signal after hybridization to total human
DNA and washing the filter to 0.1 x SSC.
i 69 contain the X chromosome as their sole human complement.
Such hybrids have been constructed by Goodfellow by
selecting for the X- li nk ed 1IGRPT enzyme in HAT medium. The
fibroblast cell line Horl9X retains an intact X chromosome
in a tetraploid mouse background (Goodfellow et al./
1930), MCP-6 is a teratocarcinoma cell line and carries
the translocation chromosome t(6;X )(6qter-6p21::Xql3-qter)
in a normal diploid mouse background (Goodfellow et al.,
1932), IUI-5 is a fibroblast line with a single X
chromosome deleted for Xpll-pter in a diploid mouse
background (Solomon et al., 1976) and IR is the mouse
fibroblast line from which Horl9X and IbI-5 were
constructed. Cells were grown in tissue culture and the
DMA extracted for digestion with the restriction
endonuclease EcoRI (see Materials and Methods). Digests
were fractionated by agarose gel electrophoresis and
visualized by ethidium bromide fluorescence. A photograph
of such a gel is shown in Fig. 5. with lambda DMA digested
with Hindlll in lane 1 acting as molecular weight markers.
The intensely fluorescing bands present in the smear of
DMA fragments represent a family of repetitive DMA
sequences organized in tandem arrays, each copy containing
a single EcoRI site (Darling et al., 1932; Mu and
Manuelidis, 1930). Restriction fragments were transferred
to nitrocellulose by the method of Southern (1975) and
filters were hybridized overnight to recombinant DMA 3 2 labelled with P by nick translation (Rigby et al.,
i 70 Fig.5 Polaroid photograph of an agarose gel of EcoRI digested DNAs from : Lanes 2 and 7, human blood; Lanes 3 and 8, Horl9X fibroblasts; Lanes 4 and 9, IWI-5 fibroblasts; Lanes 5 and 10, MCP-6 teratocarcinoma cells;
Lanes 6 and 11, mouse fibroblasts. Lane 1 contains lambda
DNA digested with HindiII electrophoresed in parallel.
71 1976). Filters were washed to 1 x SSC or 0.1 x SSC at 65°C depending on the background hybridization of imperfect hybrids. Clearly, recombinants which hybridized to a single EcoRI band in the human and Horl9X but not the mouse (IR) tracks must be derived from the human X chromosome. Furthermore, those recombinants which hybridized additionally to single bands in the IVTI-5 and
MCP-6 tracks must be derived from Xql3-qter, those which hybridized to IVII-5 but not MCP-6 from Xpll-ql3 and those which hybridized to neither (but did hybridize to HorlPX)
from Xpll-pter.
Some of the chromosomal localizations effected by this method are shown in Figs. 6-8. where a series of autoradiographs are shown. Fig. 6.A shows that the recombinant DX24 hybridizes to a 7.0kbp band (and additionally but less intensely to a 7.8kbp band) in the human track alone. This implies that it is derived from a human autosome rather than the X chromosome. This autosome is most likely to be either chromosome 7 or chromosome 8 as these chromosomes are closest in size and DHA content to the X and are fractionated by the flow sorter into the same group as the X (Davies et al., 1981). It is unlikely to be chromosome 6 as the translocation chromosome in MCP-
6 carries most of chromosome 6. In Fig. 6B the autoradiograph demonstrates that the recombinant RD6 hybridizes to a 5.3kbp band in the human track which is also present in the Horl9X and IbI-5 tracks but absent
from the MCP-6 and IR tracks. This indicates that this
72 A B
1 23456 1 2345 6
9 ♦ * * * ‘ + % *
Fig.6 Southern blot hybridizations of EcoRI digests of DNA from : Lane 2; human blood, Lane 3; Horl9X fibroblasts, Lane 4; IWI-5 fibroblasts, Lane 5; MCP-6 teratocarcinoma cells, Lane 6; mouse fibroblasts to A.
DX24, B RD6.
73 recombinant is human in origin (as it does not hybridize
to mouse DLJA) and derived from the X chromosome in the
region Xpll-Xql3. The higher molecular weight bands evident in the human track represent imperfect duplexes between homologous sequences. These can be melted by increasing the stringency to 0.05 x SSC at 65°C.
Fig. 7.A shows the result of hybridizing the recombinant S21 to the somatic cell hybrid panel. A single
7 .3kbp band is detected in all tracks bar the mouse parent
(IR) and this indicates that the recombinant is derived from a locus on the human X chromosome between Xql3-qter.
The same is true for the recombinant DX4, as Fig. 7.B demonstrates. This hybridizes to a 4.2kbp human fragment
(and an additional 2.2kbp band revealed on longer exposure) present in all tracks bar the mouse.
The overexposed autoradiograph shown in Fig. o.A demonstrates that the recombinant DX7 hybridizes to all the tracks, including the mouse, but in a species specific fashion. This recombinant detects an 3.4kbp band in the human track that is absent from all the hybrid tracks and the mouse but, in addition, detects a 2.2kbp band (and also a 1 .3kbp band revealed on longer exposure) present in all the hybrids and the mouse but not human tracks. This fragment must then be derived from a human autosome and contains sequence which is conserved between mouse and human but contained within diverged EcoRI sites.
In Fig. 8.B, DX6 detects a single 7.7kbp band present
74 A B
Fig.7 Southern blot hybridizations of EcoRI digests of DNA from : Lane 2; human blood, Lane 3; Horl9X fibroblasts, Lane 4; IWI-5 fibroblasts, Lane 5; MCP-6 teratocarcinoma cells, Lane 6; mouse fibroblasts to A S21 and B D X 4 .
75 A B
Fj.g.8 Southern blot hybridizations of EcoRI digests of DNA from : Lane 2; human blood, Lane 3; Horl9X
fibroblasts, Lane 4; IWI-5 fibroblasts, Lane 5; MCP-6
teratocarcinoma cells. Lane 6; mouse fibroblasts to A DX7
and B D X 6 .
76 in all tracks except the mouse and is therefore derived
from Xql3-qter. This 7.7kbp fragment includes a highly reiterated sequence (of unknown origin) as evidenced by the reaction v/ith total human DMA in Fig. 4. and under normal conditions hybridizes to a smear of fragments on a
Southern blot of genomic DMA. This repetitive sequence can be competed out of hybridizing to the filter by hybridizing the nick translated recombinant v/ith total human DMA in solution before and during hybridization to the filter (see Materials and Methods). In this way, the cognate EcoRI fragment representing the single chromosomal locus from which the recombinant is derived can be detected as shown without physical separation from reiterated sequences present on the same molecule.
In most of the Southern blots, a background smear of radioactivity can be seen in the human and hybrid tracks
(although very much fainter in the hybrid). The cause of this smear is interpreted as hybridization of a reiterated sequence present on the same recombinant molecule to many homologous sequences in the genome. The evidence for this interpretation lies in the species specificity of the smear which is rarely evident in the mouse track.
A summary of the chromosomal localizations effected by this method is presented in Table 2. A higher resolution using a hybrid panel with many more derivative
X chromosomes has been possible in the laboratory of
Ropers. Experiments in that laboratory have further localized DX13 to Xq23-qter, RB6 to Xq26-q27, RC3 to Xpll-
77 Table 2
Recombinant Hybridization to Localization
HUMAN H0 RL9X IWI-5 MCP-6
DX4 + + + + X q l 3 - q t e r DX6* + + + + X q l 3 - q t e r DX7 + - - - autosomal D X 1 3 + 4* + X q l 3 - q t e r D X 1 5 + - -- autosomal DX24 + --- autosomal DX48 + + + + X q l 3 - q t e r RB6 + + + + X q l 3 - q t e r RD6 + + + - X p l 1 - q l 3 RC8 + + - - X p l 1 - p t e r LI .28 + - - X p l 1 - p t e r DP 31 + + + + X q l 3 - q t e r S21 + + + 4- X q l 3 - q t e r 52A + + + + Xq l3 - q t e r
Summary- of chromosomal localizations using somatic
cell hybrids.
* DX6 contains repetitive sequence and the
hybridization profile shown represents the cognate EcoRI
fragment and not other homologous sequences. i 78 Xp223, and 821 to Xq213-q26 (V:ieacker et al., 1983). Using
their hybrid panel in this laboratory, DX4 maps to Xq28- qter (results not shown).
In all these experiments hybridization to a single
EcoRI fragment of the same molecular weight as the cloned
fragment (compare, for example, Figs. 7 and 8 to Fig. 3.) was taken as evidence for a single genomic locus. This remains, of course, an assumption, but the assumption is
strengthened by the greater resolution of chromosomal
localization afforded by 'in situ' hybridization.
3.4. Chromosomal localization 11. 1 In situ1 hybridization
h Introduction
The technique of 'in situ' hybridization offers the greatest potential in ma mm al ia n gene mapping ye t it remains shrouded with s u s p i c i o n . It is clear that the reaction of input labelled DMA with chromosomal DMA on microscope slides yields specific DMA hybrids (Join e s ,
1973; Wimber and Steffensen, 1973). The problem is trying to detect a single locus corresponding to extremely small amounts of DMA. One can begin to approach the problem mathematically in the following fashion.
1. Let the input DMA be a trittiated 6kbp DMA sequence
labelled to a specific activity of 4.5 x 10 dpm/ug. This
is hybridized to metaphase chromosomes and exposed to
79 emulsion for 15 days
2. 1 mole of base pairs has a mass of approximately 660g . . . 7 Consequently a specific activity of 4.5 x 10 dpm/ug 7 G —"5 corresponds to 4.5 x 10 x 660 x 10 = 4.9 x 10 dpm/bp
6.023 x 1 0 23
3. By hybridizing 5S RNA to Drosophila salivary chromosomes Szabo et al. (1977) have estimated that only
9% of sites become label led. As the ratio of input to target DMA can reach as 1 ow as 100, it is unlikely that saturation is achieved, but assuming saturation, the amount of radioactivity at each site will be _ p _ ^ 9 x 5 x 103 x 4.9x 10 = 2.2x10' d pm/site 100
As there are two sites on every chromosome, this will be
4.4 x 10 dprn/chromosome
4. lode rn nuclear emulsions have an autoradiographic efficien cy for trittium of approximately 0.12 grains per beta d i s integration (Ada et al., 1966), thus for a 15 day exposure one would expect
15 x 24 x 60 x 0.12 x 4.4 x 10 =0.12 grains/chromosoine
This fig ure must be taken with extreme caution but clearly represen ts an experiment which becomes feasible with some enhancem ent. Such enhancement can be obtained by using double stranded (nick translated) DMA probes in conjunct ion with dextran sulphate in the hybridization reac tion (Wahl et al., 1979; Harper and Saunders, 1931).
Dextran sulphate accelerates the renaturation of a variety of DMAs approximately 10 fold in solution (I’etmur, 1975).
80 This is attributed to the polymer excluding the DMA from its volume and hence causing an increase in the effective
DMA concentration. For a bimolecular reaction, the ten
fold increase in rate corresponds to an increase of approximately three fold in the effective DMA concentration of each component. Thus when one component is immobilized a 3-4 fold increase has been observed with a single stranded probe (Wahl et al.f 1979). However, when the probe is randomly nicked, double stranded DMA (dsDUA), there is up to a hundred fold increase in rate and this dramatic increase is attributed to the ability of the dsDMA probe to form extensive networks (Dahl et al.,
1979). These observations have led to the development of an 'in situ' hybridization protocol that has general applicability to any
Saunders, 1931) which
(Szabo and Wahl, 1932).
The protocol used methods (Harper and Saunders, 1931; Malcolm et al., 1932) and direct experience (see Materials and Methods). The fidelity of the Paddington protocol was proven by the localization of a random genomic sequence derived from a library of DMA enriched for chromosomes 3, 4, and 5 by flow sorting. This recombinant, 4C11, contains two contiguous EcoRI fragments for a total of 6kbp and was nick translated to a specific activity of 4.5 x 10^ dpm/ug using three trittiated deoxyribonucleotides. The labelled
DMA was hybridized overnight to male metaphase spreads at
81 B
Fig.9 Human male metaphase spread A - after G-banding and B - after hybridization to 4C11 and autoradiography.
One of the two chromosome 6s is labelled.
82 a concentration of lOug/ml and exposed to Ilford K2
emulsion at 4°C for 15 days before developing and scoring
(see Materials and Methods). Fig. 9. is an example
autoradiogram, shown next to the same spread after banding
but before hybridization. The high background of grains is
evident, but clearly one chromosome 6 is labelled.
Table 3 compares the observed number of grains to an
expected value assuming random grain distribution. This
value is computed by multiplying the total number of
grains by each individual chromosome's percentage DMA
content of the diploid total (Mendelsohn et al., 1973). In
this instance only chromosome 6 is labelled significantly
over the 9 cells examined. Fig. 10 depicts the
distribution of grains over the 10 chromosomes scored
diagramatica1ly and clearly the recombinant is derived
from a sequence at 6pl2 as 18 of the total 39 grains lie
over this region. This result places into doubt the
fidelity of either the flow sorting from which 4C11 was derived, or the 'in situ' hybridization experiment. That
the latter is not the case is confirmed by hybridization of 4C11 to a somatic cell hybrid DMA panel identical to
those of Figs. 4-6. Fig. 11. shows the result of this
experiment which demonstrates that 4C11 detects EcoRI
fragments in only the human and MCP-6 DMA tracks. As MCP-6 contains the translocation chromosome t(6;X)(6qter- p21 : : Xql 3-qter) and 4C11 hybridizes to neither IIorl9X (X) nor IbI-5 (Xpll-qter) DMAs, this confirms the 'in situ'
0 3 Table 3
4C11 hybridised to 9 male metaphase spreads
Chro mo so me 0 E 0-E 1 15 23.11 - 3.11 2 18 22.52 - 4.52 3 20 18.62 1.38 4 9 17.82 - 8.82 5 9 17.08 - 8.08 6 43 16.12 26.88 7 21 14.78 6.22 8 21 13.61 7.39 9 11 12.65 - 1 .65 10 13 12.43 0.57 11 11 12.70 - 1 .70 12 7 12.54 -5.54 13 11 9.93 1.07 14 12 9.61 2.39 15 11 9.02 1 .98 16 6 8.27 - 2.27 17 6 7.95 - 1 .95 18 5 7.47 - 2.47 19 6 5.76 0.24 20 5 6.40 - 1 .40 21 3 4.38 -1.38 22 2 4.59 - 2.59 X 11 7.20 3.80 Y 1 2.45 - 1 .45
S D = 6 .88247
The deviation of observed (o) from expected (E) values (O-E) is estimated as a probability that the deviation is random by the number of standard deviations of all 0-E values each 0-E value takes. Only in the case of chromosome 6 is the 0-E value greater than one standard deviation. The probability that this value is random is approximately 0.005%.
84 X4CII 18 Chromosomes
6 p 1 2
Fig.10 Distribution of grains over 18 chromosome 6s
following 'in situ' hybridization to 4C11. The vertical
axis represents the composite number of grains and runs
from 0 to 20. The horizontal axis represents a G-banded
chromosome 6 running from the tip of the short arm at the
left to the tip of the long arm at the right. The banding
pattern is from the Paris Conference, 1981 (ISCN, 1981).
> 35 12 3 4 5
Fig.11 Southern blot hybridizations of EcoRI digests of DNA from : Lane 1; human blood, Lane 2; Horl9X fibroblasts, Lane 3; IWI-5 fibroblasts, Lane 4; MCP-6 teratocarcinoma cells, Lane 5; mouse fibroblasts to nick translated 4C11. The open arrows represent molecular weight sizes of 23.1, 9.4, 6.6 and 4.4kbp.
86 result. In this experiment if saturation of available
sites were accomplished, one would have expected 0.12
grains/metaphase chromosome. As the observed average is 1
grain/metaphase chromosome clearly dextran sulphate is
enhancing the rate and the dsDMA probe is facilitating
network formation.
The distribution of repetitive sequences in the human
X chromosome (and indeed throughout the genome) is clearly visualized in single cells by 'in situ' hybridization.
Fig. 12. demonstrates the distribution of two types of middle repetitive sequence. In A the probe is derived from
a 340bp or 630bp reiterated sequence homologous to the
African Green nonkey alpha satellite (Darling et al.,
1932), and in 3 the probe is derived from a 1.9kbp
dispersed repeat with a single Hindlll site (Manuel idis
and Biro, 1982; Crampton, unpublished observations). The
result indicates the remarkably different chromosomal
organizations of these two mainland middle repetitive
sequences. The HindiII 1.9kbp repeat is dispersed
throughout the genome with no obvious preferential sites.
Tnis is similar to the distribution of Alu sequences
revealed on hybridizing a cloned member, BLUR8 (Jelinek et
al., 1930), to metaphase spreads (results not shown). The
alpha RI repeat, however shows a striking constraint to
procentric locations, [particularly on chromosomes 1, 3, 5,
16 and 19, with some interstitial grains. The X chromosome
appears to be somewhat underrepresented for this
particular DMA sequence but a 2kbp BamHI repeat (Yang et Fig.12 Autoradiographs of human metaphase spreads hybridized 'in situ' to cloned human repetitive DNA sequences. In A the probe is alpha RI repetitive sequence and in B a copy of the 1.9kbp Hindlll repeat.
88 al., 1982; Willard et al., 1982) reveals X centromeric grains almost exclusively (results not shown).
B Cytological mapping of the X chromosome
Having proven the fidelity of 'in situ' hybridization with the localization of 4C11 to 6pl2, the X chromosome markers isolated and characterized as described in sections 3.2.-3.3. have been used to construct a high resolution cytological map of the X chromosome. Technical amendments were applied to the protocol with different experiments, principally in adjusting the probe concentration to reduce the rather high background evinced in Fig. 9. In addition, the trypsin-Giemsa banding used prior to hybridization for chromosome identification in the 4C11 experiment was abandoned in favour of Lipsol banding followed by Leishmann's stain (Malcolm. et al.,
1982). This modification resulted in a much greater retention of chromosome morphology thus facilitating direct comparisons of autoradiographs with photographs of banded spreads.
The recombinant RC8 contains a 5.1 kbp human insert and was nick-translated to a specific activity of 2.51 x 7 . . . 10 dpm/ug. The probe was hybridized overnight to slides at a concentration of lug/ml and exposed to Ilford L4 emulsion for 11 days at 4°C. Fig. 13 shows a spread with both X chromosomes labelled at Xp22 with a photograph of the banded chromosomes before hvbridization for
89 Fig.13 Human female metaphase spread A - after G- banding and B - after hybridization to RC8 and autoradiography. The two X chromosomes are labelled and in addition grains reside over chromosomes 7 and 22.
90 RC8 12 Female 6 Male Cells
Fig.14 Composite grain distribution over 18 cells after hybridization 'in situ' to RC8. Each dot represents the chromosomal position of a single grain. The banding pattern is from the Paris Conference, 1981 (ISCN, 1981).
91 Table 4
RC8 hybridised to 6 male and 12 female metaphases
Ch ro mo so me 0 E 0-E 1 3 8.49 - 5.49 2 10 8.28 1.72 3 2 6.85 - 4.85 4 13 6.55 6.45 5 7 6.28 0.72 6 3 5.92 - 2.92 7 7 5.43 1.57 8 2 5.00 - 3 .00 9 2 4.65 - 2.65 10 1 4.57 - 3.57 11 4 4.67 - 0.67 12 8 4.61 3.39 13 4 3.65 0.35 14 1 3.53 - 2.53 15 6 3.32 2.68 16 3 3.04 - 0.04 17 4 2.92 1 .08 18 3 2.75 0.25 19 3 2.12 0.88 20 1 2.35 -1.35 21 2 1.61 0.39 22 2 1.69 0.31 X 12 4.41 7.59 Y 0 0.30 - 0.30
S D = 3 .06817
The deviation of observed (0) from expected (E)
values (0-E) is estimated as a probability that the
deviation is random by the number of standard deviations
of all O-E values each 0-E value takes. Only in the case
of the X chromosome is the 0-E value greater than two
standard deviation. The probability that this value is
random is approximately 0.1%.
> 92 20-i XRC8 11 Chromosomes
pll
Fin.15 Distribution of grains over 11 X chromosomes
following 'in situ' hybridization to RC8 . The vertical
axis shows the composite number of grains. The horizontal
axis shows a G-banded X chromosome. The banding pattern is
from the Paris Conference, 1981 (ISCN, 1981). comparison. Fig. 14 is a diagrammatic representation of the composite number of grains over IB spreads and clearly demonstrates that this sequence is derived from Xp22. The
analysis of this result in Table 4 suggests only the X chromosome is significantly labelled (P<0.1%).
Further confirmation of this localization is accomplished by performing the hybridization at lQug/ml probe concentration. Under these conditions (with this probe), extremely high backgrounds are obtained and a full analysis becomes impossible due to large clusters of grains obscuring many chromosomes. The origin of these clusters is unknown, but presumably they represent extensive networks of radioactive probe. Fig. 15 is a diagrammatic representation of the composite grain di st r i b u t i o n over 11 X chromosomes v/hich remained unobscured by clusters. Of the total of 3.8 grains, 15 reside over Xp22.
Fig. 16 shows a metaphase spread after hybridization to the probe LI.28 and 20 days exposure to Ilford K2 emulsion at 4°C. The unhybridized banded spread is also shown for comparison. The probe contains a human DUA insert of 1.5kbp within the plasmid pBR322 (pavies et al.,
1983), was nick translated to a specific activity of 2.4 x - j 10 dpm/ug and hybridized at a concentration of lOOng/ml to microscope slides. The example autoradiograph) in Fig. 16 has one of the X chromosomes labelled at Xpl1. The composite grain distribution from 40 cells in Fig. 17 and the analysis in Table 5 provide convincing B
Fig.16 Human female metaphase spread A - after G- banding and B - after hybridization to LI.28 and autoradiography. One of the X chromosomes is labelled and in addition a single grain resides over chromosome 2.
95 LI.28 40 Cells
Fig.17 Composite grain distribution over 40 cells after hybridization to LI.28. Each dot represents the chromosomal position of a single grain. The banding pattern is from the Paris Conference, 1981 (ISCbT, 1981) .
96 Table 5
LI.28 hybridised to 40 female metaphases
Chromosome 0 E 0-E 1 9 10.66 - 1.66 2 8 10.39 - 2.39 3 9 8.59 0.41 4 6 8.22 - 2.22 5 10 7.88 2.12 6 8 7.44 0.56 7 6 6.82 - 0.82 8 4 6.28 - 2.28 9 10 5.84 4.16 10 5 5.74 - 0.74 11 7 5.86 1.14 12 7 5.79 1.21 13 5 4.58 0.42 14 1 4.43 - 3.43 15 6 4.16 1.84 16 3 3.82 - 0.82 17 0 3.67 - 3.67 18 2 3.45 - 1 .45 19 3 2.66 0.34 20 1 2.95 - 1 .95 21 0 2.02 - 2.02 22 0 2.12 - 2.12 X 20 6.65 13.35
S D = 3 .40881
The deviation of observed (O) from expected (E) values (0-E) is estimated as a probability that the deviation is random by the number of standard deviations of all 0-E values each 0-E value takes. Only in the case of the X chromosome is the 0-E value greater than one standard deviation. The probability that this value is random is approximately 0.006%.
97 evidence that this sequence is derived from proximal Xpl1
(P < 0 .0061).
5 2 a is a recombinant containing 1.7kbp of human DbA inserted into the plasmid pBR322 and is a subclone, freed of reiterated sequence, of a bacteriophage lambda recombinant isolated from the X-enriched library (Drayna et al., 1934). The recombinant was nich translated to a 7 specific activity of 6.74 x 10 dpm/ug and hybridized overnight to slides at a concentration of 100ng/ml. Fig.
13 is an autoradiograph after 20 days exposure at 4°C and one of the X chromosomes is labelled. The composite grain distribution over 40 cells is presented in Fig. 19. The seven grains over Xq27 suggest that 52A is derived from this region of the X chromosome but the statistical significance of this result is lower than other experiments (P<1.551, Table 6). This corresponds to a lower hybridization efficiency which may be due to diffusion constraints - the chromosome in this region might be highly condensed and consequently less accessible to probe. Alternatively the sequence might be more labile to fixation of cells in acid/methanol or the denaturing of the DdA. Small amounts of D'dA are lost from the chromosome during these treatments (Comings, 1978; Gall and Pardue,
1971 ) .
The recombinant S21 detects a locus in the region ql3 to qter as demonstrated in section 3.3. This was further localized by an 'in situ' hybridization experiment. The Fig. 18 Human female rnetaphase spread A - after G- banding and B - after hybridization to 52A. and autoradiography. One of the two X chromosomes is labelled and in addition grains reside over chromosomes 1, 6 and 7.
99 52 A 40 Cells
Fig.19 Composite grain distribution over 40 cells after hybridization 'in situ' to 52A. Each dot represents the chromosomal position of a single grain. The banding pattern is from the Paris Conference, 1981 (ISCO, 1981).
100 Table 6
52A hybridised to 40 female metaphases
Ch ro mo so me 0 E O-E 1 14 9.18 4.82 2 7 8.95 - 1.95 3 2 7.40 - 5.40 4 10 7.08 2.92 5 6 6.79 - 0.79 6 7 6.41 0.59 7 9 5.88 3.12 8 2 5.41 - 3.41 9 3 5.03 - 2.03 10 10 4.94 5.06 11 3 5.05 - 2.05 12 4 4.98 - 0.98 13 5 3.95 1.05 14 0 3.82 - 3.82 15 1 3.58 - 2.58 16 3 3.29 - 0.29 17 4 3.16 0.84 18 1 2.97 - 1 .97 19 2 2.29 - 0.29 20 2 2.55 - 0.55 21 1 1.74 - 0.74 22 4 1 .82 2.18 X 12 5.73 6.27
S D = 2 .90386
The deviation of observed (0) from expected (E) values (O-E) is estimated as a probability that the deviation is random by the number of standard deviations of all 0-E values each O-E value takes. The probability that the O-E value for the X chromosome is random is approximately 1.55%.
101 Fig.20 Human female metaphase spread A - after G- banding and B - after hybridization to S21 and autoradiography. Only the X chromosome is labelled with g r a i n s .
102 at rn s rm h Prs ofrne 13 (SN 1981). (ISCN, 1931 Conference, Paris represents the dot from Each is n . er 1 tt 2 pa S to situ' 'in n io at iz id br hy after the chromosomal po si ti on of a single grain. The banding banding The grain. single a of on ti si po chromosomal the
F i g .21 Composite grain di st ri bu ti on over 36 cells cells 36 over on ti bu ri st di grain Composite .21 g i F ai ■ i iBrnm S2I 36 Cells 36 103 Table 7
S21 hybridised to 36 female metaphases
Chromosome 0 E 0-E 1 2 6.48 - 4.48 2 6 6.31 - 0.31 3 5 5.22 - 0.22 4 3 5.00 - 2.00 5 3 4.79 - 1 .79 6 1 4.52 - 3.52 7 6 4.14 1 .86 8 4 3.82 0.18 9 3 3.55 - 0.55 10 4 3.49 0.51 11 3 3.56 - 0.56 12 1 3.52 - 2.52 13 1 2.78 - 1.78 14 1 2.69 - 1.69 15 1 2.53 - 1 . 5 3 16 3 2.32 0.68 17 3 2.23 0.77 18 2 2.09 - 0.09 19 0 1.62 - 1 .62 20 3 1 .80 1.20 21 1 1.23 - 0.23 22 3 1.29 1.71 X 20 4.04 15.96
S D = 3 .73982
The deviation of observed (0) from expected (E) values (0-E) is estimated as a probability that the deviation is random by the number of standard deviations of all 0-E values each 0-E value takes. Only in the case of the X chromosome is the 0-E value greater than one standard deviation. The probability that this value is random is approximately 0.001%.
104 DMA was nick translated to a specific activity of 4.94 x 7 10 dpin/ug and hybridized at a concentration of lOOng/ml to
slides which were exposed to Kodak NTB-2 emulsion for 22
days at 4°C. Fig. 20 shows the resultant autoradiograph
with one X chromosome labelled and in Fig. 21 the
composite grain distribution over 36 spreads is displayed
diagramatically. The only chromosome with a significant
number of grains covering it is the X and clearly S21 is derived from a sequence at Xq22-q23 (P<0.001%, Table 7).
A recombinant has been reported which detects homologous loci on both the X and Y chromosomes (Page et
al., 1932). This recombinant, DP31, contains 4.5kbp of human DMA cloned into pBR322. The X chromosome locus has been investigated by 'in situ1 hybridization in this
study. The probe was nick translated to a specific
activity of 2.62 x 10 dpm/ug and hybridized at a
concentration of lOOng/ml to slides which were exposed to
Kodak MTB-2 emulsion for 22 days. Figs. 22-23 represent
the result. Fig. 22 is a hybridized spread showing one of
the X chromosomes labelled at Xql3 and Fig. 23 represents
the composite grain distribution over 37 spreads. The X chromosome remains the only significantly labelled
chromosome (female cells were used and hence the Y locus was not investigated) and it seems unequivocal that the X
locus is at Xql3 (P<0.009%, Table 0).
The paradox of when a single-copy sequence is repeated is raised by the recombinant DX13. This lambda
recombinant contains two human EcoRI fragments of 2.2 and
105 Fig.22 Human female metaphase spread A - after G- banding and B - after hybridization to DP31 and autoradiography. One of the two X chromosomes is labelled and in addition a single grain resides over chromosome 1.
106 DP3I 37 Cells
Fig.23 Composite grain distribution over 37 cells after hybridization 'in situ' to DP31. Each dot represents the chromosomal position of a single grain. The banding pattern is from the Paris Conference, 1981 (ISCN, 1981).
107 Table 8
DP31 hybridised to 37 female metaphases
Chromosome 0 E 0-E 1 12 11 .89 0.11 2 7 11.59 - 4.59 3 15 9.58 5.42 4 8 9.17 - 1.17 5 15 8.79 6.21 6 4 8.29 - 4.29 7 10 7.61 2.39 8 4 7.00 - 3.00 9 6 6.51 - 0.51 10 6 6.40 - 0.40 11 4 6.54 - 2.54 12 11 6.45 4.55 13 2 5.11 - 3.11 14 5 4.94 0.06 15 4 4.64 - 0.64 16 4 4.26 - 0.26 17 2 4.09 - 2.09 18 0 3.84 - 3.84 19 0 2.97 - 2.97 20 0 3.30 - 3.30 21 1 2.25 - 1.25 22 0 2.36 - 2.36 X 25 7.41 17.59
S D = 4 .73838
The deviation of observed (0) from expected (E) values (0-E) is estimated as a probability that the deviation is random by the number of standard deviations of all 0-E values each 0-E value takes. Only in the case of the X chromosome is the 0-E value greater than two standard deviations. The probability that this value is random is approximately 0.009%.
108 Fig.24 Human female metaphase spread A - after G- banding and B - after hybridization to DX13 and autoradiography. The two X chromosomes are labelled and in addition grains reside over chromosome 7.
109 DXI3 14 Cells
1 1
’ : ¥ 7 R ' f S i qi: ‘■y-y. ) ■.i ___
I6 I7 I8 19 20 21 22 Y
Fig.25 Composite grain distribution over 14 cells after hybridization 'in situ' to DX13. Each dot represents the chromosomal position of a single grain. The banding pattern is from Paris Conference, 1981 (ISCN, 1931).
110 Table 9
DX13 hybridised to 14 female metaphases
C h ro mo so me 0 E 0-E 1 6 6.23 - 0.23 2 0 6.07 - 6.07 3 6 5.02 0.98 4 1 4.81 - 3.81 5 3 4.61 - 1.61 6 4 4.35 - 0.35 7 4 3.99 0.01 8 6 3.67 2.33 9 4 3.41 0.59 10 6 3.35 2.65 11 2 3.43 - 1.43 12 2 3.38 - 1 . 3 8 13 2 2.68 - 0.68 14 2 2.59 - 0.59 15 2 2.43 - 0.43 16 3 2.23 0.77 17 1 2.14 - 1 . 1 4 18 2 2.02 - 0.02 19 2 1.55 0.45 20 0 1.73 - 1.73 21 0 1.18 - 1 . 1 8 22 0 1.24 - 1 . 2 4 X 18 3.89 14.11
S D = 3 .48924
The deviation of observed (0) from expected (E) values (0-E) is estimated as a probability that the deviation is random by the number of standard deviations of all 0-E values each 0-E value takes. Only in the case of the X chromosome is the 0-E value greater than one standard deviation. The probability that this value is random is approximately 0.001%.
I l l l.Okbp yet hybridizes to five human EcoRI fragments of
1.3, 2.2, 2.5, 2.7 and 7.4kbp. All four of these fragments
co-segregate with the human chromosome in IIorl9X, IUI-5
and MCP-6 (results not shown) and so all must be derived
from Xql3-qter. This evidence is suggestive of a localized
repeated sequence and further corroboration is provided by
an 'in situ' hybridization experiment. The recombinant was
hybridized to slides overnight at a concentration of
500ng/ml and a specific activity of 2.35 x 10 dpm/ug.
Slides were exposed to Ilford K2 emulsion for 25 days at
4°C and Fig. 24 is an autoradiograph depicting label on
both X chromosomes at Xq23. In Fig.25 the genomic
distribution of this repeated sequence is revealed and
clearly it is derived from a single chromosomal locus at
Xq28 (P<0.001%, Table 9). Such a discrete block of
homologous sequences is reminiscent of many chromosomal
loci including that of the histone genes (Heintz et al.,
1931).
C Summary of 'in situ' hybridization data
The chromosomal loci of six human X chromosome DMA
sequences have been determined to a high resolution by 'in
situ' hybridization. The composite X chromosomal grain distribution of these 'in situ' hybridization experiments
is presented schematically in Fig. 26. The height of each bar represents the composite number of grains over each
band from all the chromosomes scored. The specificity of
the technique is evident from this diagram and clearly the
112 2 XS2I 72 Chromosomes 20i p52A XDXI3 \ 80 Chromosomes 50 Chromosomes s \s s s\ k> io- s s s \ \ q28 Fig.26 Composite X chromosomal grain distributions. 113 Fig.27 Cytological map of the human X chromosome from 'in situ' data. The locus each marKer detects is taken as the G-band or interband with the greatest total number of grains from each experiment. The banding pattern is that from the Paris Conference, 1981 (ISCN, 1981). 114 six recombinants detect discrete loci spread along the chromosome from Xp2 2 ( RCR) to Xg20 (DX13). This then provides the basis for the cytological nap of the human X chromosome displayed in Fig. 27. The map reported here agrees with the chromosomal localization of many of the same probes using somatic cell hybrids (L'ieacker et al., 1983; see section 3.3.). The particular relevance of these six DMA probes is that in addition to detecting discrete cytological loci, they each detect restriction fragment length variants. These variant fragments act as inheritable markers conseguentlv permitting linkage distances between each pair of loci to be computed. 3.5. Restriction fragment length polymorphism A RFLP and heterozygote detection The use of restriction length variants in gene mapping and linkage analysis has become well established and is the subject of many recent reviews (Kurnit and lloehn, 1979; Botstein et al., 1980; Davies, 1981). The identification of RFLP has been approached in different ways. The stochastic approach is to screen a large number of random female DMAs digested v/ith different restriction enzymes v/ith a single probe. Such a screen has revealed RFLP detected by RC8 (Murray et al., 1982), by LI.28 (Davies et al., 1983) and by S21, 52A and DX13 (Drayna et 115 al., 1934). An internediate approach is to screen a snail number of male and female DMAs digested with many enzymes as has been successfully applied for DP31 (Page et al., 1932) or indeed to screen just two females and one male with very many enzymes. This latter approach offers the advantage that only very frequent X-linked polymorphism will be detected and it is this very frequent polymorphism which lends itself most amenable to linkage analysis. An example of the variants detected by the recombinant RC3 is provided in Pig. 23. Female DPAs ''/ere digested with the restriction endonuclease Tag I, blotted . . 32 onto nitrocellulose and nyondized to p nick translated HC3. The autoradiograph demonstrates that this recombinant hybridizes to three TaqI fragments of molecular weights 6.6, 5.3 and 3.2kbp. The largest of these fragments is invariant in all individuals tested but the 5.3kbp band interchanges with the 3.2kbp band. Pales disolay only the 5.3kbp band or the 3.2kbn band and never both (hurray, 1934). Consequently it appears that females with. both bands are heterozygous for the mutation leading to the absence (or creation) of a TaqI site at this chromosoma1 locus and can give birth to male offspring inheriting either fragment. This conclusion has been confirmed by exhaustive inheritance studies of these restriction fragments (Murray, 1934)* Genetic mappingof the X chromosome is greatly facilitated by the fact that the male is effectively 116 1 2 3 4 Fig.28 TaqI fragment variants detected by RC8. Lanes 1 and 2 are female blood DMAs digested to completion with TaqI. Lanes 3 and 4 are male blood DMAs. 117 haploid for the X and consequently there is no disruption through recombination in the chromosomal order of X-linked alleles in father to daughter transmission. Consequently once the restriction fragment variants of a father's X chromosome has been determined the linkage phase of his daughter's X chromosome can be absolutely defined. This permits the accurate definition of sites of crossover in offspring when the RFLV constitution of the grandfather, mother and father are known. Some three generation pedigrees were available for this study in the form of lvmphoblastoid cell lines derived from leukocytes or as short term peripheral blood cultures stimulated with a mitogen. The first step of linkage analysis in such pedigrees is the identification of mothers heterozygous for mutations leading to RFLP. Consequently DNA from each mother was digested with various restriction enzymes, fractionated by agarose gel electrophoresis and transferred to nitrocellulose. Filters were hybridized with the six recombinants known to detect RFLP and also with recombinants which had not been screened for RFLP. The advantage of using the mothers to screen for RFLP is that any heterozygotes identified can instantly be used in linkage analysis. However, no RFLP was detected for the recombinant DX4 with DMAs digested with either TaqI, FspI, Bglll or SstI (results not shown). Fig. 29. shows the result of hybridizing the recombinant S21 to TaqI digested DHAs of seven of the eight mothers screened in this study. This 7.3kbp sequence 118 12 3 4 5 6 7 8 Fig.29 TaqI fragment variants detected by S21. Each lane represents female cell line DNA digested to completion with TaqI. Lane 1 is molecular weight markers of 23.1, 9.4, 6.6 and 4.4kbp; Lane 2 - D3II2; Lane 3 - D 2112 ,* Lane 4 - D2II? ; Lane 5 - BD26II2; Lane 6 - BD26II4 ; Lane 7 - DlII^; Lane 8 - DlII^. 119 hybridizes to two invariant bands of 4kbp and 2.5kbp and detects variant bands of 3.2kbp and 3kbp. Hales inherit only the 3.2kbp or the 3kbp fragment and never both. Consequently, heterozygotes display both fragments and homozygotes only one. In the figure, females D2Il2» D2II_, and BD26II^ are homozygous for the commoner 3.2kbp fragment (p=0.65) and female DlII^ homozygous for the rarer 3kbp (q= 0.35)• Obviously genetic exchange between the two X chromosomes will not be detected in these mothers but females ED26II_ and Dill,, can pass on either 2 o fragment to their offspring and hence recombinants between this locus (q22-23) and an adjacent locus readily identi f ie d . The DMA constitution at six loci was determined for the eight mothers screened in this study. Table 10 displays the results of these experiments. The six loci are represented by the recombinants which detect both their cytological location and genetic constitution and are arranged in the order in which they are located on the chromosome from RC3 at Xp22 to DX13 at Xq23 (see Fig. 27). Mothers who are heterozygous at adjacent loci are clearly defined and by studying inheritance of restriction fragments from these mothers, recombinants between two cytological loci can be determined. For example, the two sisters, females 0311^ and D3II are heterozygous at the L.1.28 and DP31 loci. Thus, once the phase of their chromosomes has been determined (by determining the 120 Table 10 Female Locus RC8 LI .28 DP31 S21 52A D X 1 3 D1 II4 Bl/Bl Cl/Cl E2/E2 J2/J2 Ll/Ll Ml /Ml Dl II Bl/Bl Cl/Cl E2/E2 J2/J1 Ll/Ll Ml/Ml 6 D 2 I I ? B2/B1 C1/C2 E3/E3 Jl/Jl L2/L1 M2 / M2 D 2 1 1 ? B2/B1 C1/C2 E3/E3 Jl/Jl L2/L2 M2/M1 Bl/Bl C2/C1 E3/E2 Jl/Jl L2/L1 Ml/Ml D 3 I I 2 Bl/Bl C2/C1 E3/E2 Jl/Jl Ll/Ll Ml/Ml D 3 I I 4 BD26II2 Bl/Bl C2/C1 E2/E3 J2/J1 L2/L1 M1/M2 BD2611 . Bl/Bl C2/C1 E3/E3 Jl/Jl Ll/Ll M l / M2 4 DNA sequence variants at six X chromosome loci in eight mothers from three generation pedigrees (Dl, D2, D3 and ED26) RC8 TaqI fragments B1 3.2kbp, B2 5.3kbp LI.23 TaqI fragments Cl 12kbp, C2 9kbp DP31 TaqI fragments El 14.6kbp, E2 ll.Bkbp, E3 lO.Gkbp S21 TaqI fragments J1 3.2kbp, J2 3kbp 52A TaqI fragments LI 0.6, 0.7kbp, L2 1.3kbp DX13 Bglll fragments Ml 2.8kbp, M2 5.8kbp 121 fragments on their father's X chromosome), recombinants between these two loci can be detected in their offspring. B Inheritance studies Having identified informative mothers in the three generation pedigrees available, DUA was prepared from all the family members, digested with the appropriate restriction endonuclease (TaqI for probes RCS, LI.28, DP31, S21 and 52A; EglII for probe DX13) and blotted to nitrocellulose. Filters were hybridized to radioactively labelled recombinant probes and hybridized fragments detected by autoradiography. Fig. 30 shows the result of screening pedigree D3 with the probes LI.23 and DP31. LI.23 detects up to two Tag I fragments of 12kbp and 9kbp in females - only one of these fragments is found in males (Davies et al., 1983). The inheritance of these fragments is revealed in the figure - the grandfather (D3I ) has transmitted the 9kbp fragment (C2) to the mothers (D3II? and D3II^) who have inherited the 12kbp alternative (Cl) from the grandmother (D3I^). Consequently the offspring in the third generation can inherit either fragment from their mother - the departure of expected Mendelian ratios of 50% of the offspring inheriting each fragment is not significant in this small a sample. The pattern revealed when DP31 is hybridized to a TaqI digest of pedigree D3 is more complex (Fig. 30B). This recombinant detects a 14.6kbp TaqI fragment of the Y chromosome and either an ll.Bkbp or a lO.Gkbp fragment of 122 A Fig.30 TaqI fragment variants in pedigree D3 detected by A - LI.28, B - DP31. Lane 1-1 , Lane 2-I2# Lane 3-II , Lane 4-111-^, Lane 5-III2, Lane 6-111^, Lane 7-II , Lane 8- II4 , Lane 9-III , Lane 10-III , Lane ll-III . 123 Fig.31 Pedigree D3 showing the inheritance of Tagl variants detected by LI.28 (Cl and C2) and by DP31 (El , E2 and E3) with Duchenne muscular dystrophy. The El var iant is linked to the Y chromosome. Affected offspring are shaded, carriers are marked with a dot. 124 the X. The inheritance of the Y fragment (El) is shown by the transmission from the father II„ to his sons III and S III . The mothers (II0 and 11^) have inherited the ll.Ghbp fragment (E2) from the grandmother I and the lO.Ghbp fragment (E3) from the grandfather I and transmit either X-linked sequence to the third generation. As the X chromosome in I has the fragment C2 at the LI.28 locus and the fragment C3 at the DP31 locus, the phase in both mothers must be C2E3/C1E2. Consecuently, any sons with the chromosomes C2E2 or C1E3 nust be recombinants and those with C2E3 or C1E2 are non-recombinants. The pedigree in Fig. 31 shows that one of the two sons in each family (III and II lr) are recombinants. In addition, III has inherited the C1E2 chromosome from her father II and is also a recombinant having therefore inherited a C2E2 chromosome from her mother. Ill may or maynot be a recombinant depending on her father's X chromosome - if he were C1E3 then she is not a recombinant whereas if he were C1E2 then she must hie a recombinant. Unfortunately the father was not available for testing. The pedigree D3 also segregates for Duchenne muscular dystrophy as shown. The segregation of the DXA at the LI.28 locus with the Duchenne locus is commensurate with the 17cli genetic distance established between the two (Davies et al. , 1933). The RULE for the LI.28 locus is also demonstrated in the pedigree D2. In addition the RFLP detected by RCO is 125 Fig.32 Pedigree D2 showing the inheritance of TaqI fragments detected by LI.28 and RC8 with Duchenne muscular dystrophy. The TaqI variants detected by LI.28 are termed Cl and C2 and by RC8 are termed B1 and B2. Affected offspring are shaded, carriers are marked with a dot. 126 also evident in this pedigree. Fig. 32 shows how the 12kbp (Cl) and 9kbp (C2) Taol fragments detected by LI. 23 segregate with the 3.2kbp (P.l) and 5.3kbp (132) Tacrl fragments detected by RGB. The grandfather's (1^) chromosome is B2C1 thus the phase in both mothers is B2C1/B1C2 and recombinants in the third generation will be B2C2 or B1C1. There are no recombinants in the offspring of IIn (the daughter II^ is homozygous at both loci and thus has inherited trie non-recombinant B1C2 chromosome from the mother) and a single recombinant in the offspring of II_ - III-, who has inherited the B1C2 chromosome from 7 3 her father and therefore the B1C1 recombinant chromosome from her mother. The diagnostic potential of these RFLP markers is revealed by following their segregation with the Duchenne phenotype in this pedigree. The mutant chromosome also carries the D1C2 Tagl fragments in the mothers and the two affected offspring. This observation taken with the linkage data available between DC 8 and DTP and LI.28 and DFD (Davies et al., 1933), allows one to predict that III9 and III7 who also have the R1C2 chromosome from their mothers, will be carriers for DI .D. Table 10 shows that mother DD2GII^ is heterozygous at each of the six loci investigated bar that detected by RC3 . The TaqI fragments detected by DP31 and the Bglll fragments detected by DX13 in the individuals from the pedigree are shown in Fig. 33. DX13 detects six Bglll fragments, four invariant bands of 8.3kbp, 1.7kbp, 1.4kbp and 0.9kbp and two variant bands of 5.Bkbp and 2.Bkbp. 127 12 3 4 5 6 7 B Fig.33 Restriction fragment variants in pedigree BD26. A TaqI fragments detected by DP31. B Bglll fragments detected by DX13. Lane 1-1^, Lane 2-I2, Lane 3-II2, Lane 4 -H 4 1 , Lane 5-111^, Lane 6 -III3# Lane 7-III4 . 128 U u in -J ^ Fig.34 Pedigree BD26 showing the inheritance of restriction fragments detected by LI.28, DP31, S21, 52A and DX13. Variants are termed for; L1.28-C1,C2; DP31- E 2 , E 3 ; S21-J 1 ,J 2 ; 52A-L1,L2; DX13-M1,M2. 129 Only one of the variant bands is ever seen in males. The inheritance of these Bglu fragments along with the TaqI fragments which hybridize to LI.28, DP31, S21 and 52A is shown in Fig. 34. Unfortunately the mother who is heterozygous at all these loci has only a single son but remarkably no recombination can be detected in this son's X chromosome from the LI.23 locus (Xpll) to the DX13 locus (Xq28). The grandfather's chromosome is C1E3J1L1M2 and the mother, II , therefore has the phase ClE3J1L1M2/C2E2J2L2H1 and her son (III ) has inherited the non-recombinant chromosome C2E2J2L2M1. It seems quite extraordinary for there to have been no recombination in the interval Xpll- Xcen-Xqter at a single meiosis and it seems more likely that two crossovers have occurred between adjacent loci. As chiasma interference does not extend across the centromere in humans (ilulten, 1974) and crossover interference does not extend across the centromere in Drosophila (Muller, 191G), the most likely site of a double crossover is between the loci detected by DP31 (Xql3) and LI.28 (Xpl1). 3.6. Genetic distances a genetic map of the X The paucity of informative meioses available for this study prohibit a high lod score for the maximum likelihood estimate of theta between any two loci. However, Table 11 130 Table 11 A - Li.28 v DP-31 Mother No. Recombinants Recombination Traction (theta) 0.01 0.05 0.] 0.15 0.2 0.25 0.3 0.35 0.4 0.45 D3II.2 1/2 -1.402 -0.721 -0.444 -0.292 -0.194 -0.125 -0.076 -0.041 -0.018 -0.004 D3II.4 2/3 -3.101 -1.721 -1.143 -0.815 -0.592 -0.426 -0.298 -0.196 -0.115 -0.050 BD26II.2 0/1 0.297 0.279 0.255 0.230 0.204 0.176 0.146 0.114 0.079 0.041 TOTALS -4.206 -2.163 -1.332 -0.877 -0.582 -0.375 -0.228 -0.158 -0.054 -0.013 B - RC8 v LI.28 Mother No, Recombinants Recombination Fraction (theta) 0.01 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 D2II.2 0/2 0.593 0.553 0.511 0.461 0.403 0.352 0.292 0.228 0.158 0.083 B2II.7 1/3 -1.106 -0,442 -0.188 -0.062 0.010 0.051 0.070 0.073 0.061 0.037 TOTALS -0.513 0.116 0.323 0.399 0.418 0.403 0.362 0.301 0.219 0.120 Lod scores from the inheritance studies in this work. The Lod values from each pedigree were obtained from standard tables or using the computer program LIPED. 131 shows the loci scores at different values of theta between the loci detected by DP31 and LI.28 (derived from pedigrees D3 and BD26) and between those detected by LI.20 and RC8 (derived from pedigree D2). From this data i t can be seen that linkage is suggestive betv/een RCS and LI .28 Q t a recombination fraction of 0.2-0.25 and unlikely betv/een LI.28 and DP31 The sequential method of determining the maximum likelihood estimate permits this data to be combined with similar analyses elsewhere and a more unequivocal estimate achieved. Ouch an analysis has been performed by Drayna in Salt Lake City and the combined data suggest the following genetic distances where Xcj is the locus for the X-l inked blood antigen at Xp223 (Ferguson-Smith et al., 1902). Xg - RCS > 50cM RC8 - LI.20 3 3 cM LI.28 - DP31 > 50c M DP31 - S21 14 c H S21 - 5 2A 4 5cli 52A - DX13 4Gci 1 Th ese geneti distances are still equivocal (Drayna et al., 1904) but comparison with the cyto logical map of Fig. 27 is possibl e. Fig. 35 represents this schematically by interpreting d ifferences between genetic and physical distances as an expansion or contraction of the mitotic metaphase G-bands. The physical locus is taken as the mid- point of the band each marker has been assigned to by 'in situ' hybridizati on. The genetic locus is taken from the 132 f Xg Physical Genetic Fig.35 Physical and genetic maps of the human X chromosome. The relative position of band-interband boundaries to flanking loci is converted to a fraction of the genetic distance separating the loci to generate the genetical representation of the chromosome shown. 133 genetic map of Drayna et al (1984) Clearly the genetic distances between the RC3 locus and the qene for the Xq surface antigen at Xp223 (Ferguson-Smith et al. , 1932) and between the loci for 52A (Xg27) and DX13 (Xq23 ) grossly differ from the comparative length of the metaphase chromos'orne that separates them. This impl ies that recombination is occurring far more f requently in proterminal regions. 134 CHAPTER 4 Discuss ion 135 4.1. DMA seq uence organization of the human X chromosome The results from this study demonstrate the distribution of some repetitive DMA sequences over the X chromosome. The observation that total MCP-S DMA reacts with the same recombinant sequences as total Horl9X DMA suggests that there are no repetitive sequences on the X chromosome Inching in the translocation chromosome t(X;6)(Gqter-6p21::Xql3-qter). This therefore implies that there are no centromeric repetitive sequences specific to the X. However a 2kbp Bamlll repeat sequence has been reported to be highly represented at Xcen (Yang et al., 1902; Millard et al., 1923) and 'in situ* hybridization with this sequence reveals grains almost exclusively at Xcen (Yang et al., 1932; Cates and Hartley, unpublished). The conclusion must be that this repeated sequence was not represented in any of the recombinants probed '-4th somatic cell hybrid DMA. This conclusion is strengthened by analysis of the repeated sequence. Millard et al. (1933) report that the 2kbp repeat is predominantly tandemly repeated at a pericentromeric location on the X chromosome. Southern blot analysis to total EcoRI digests of human DMA indicated that this enzyme primarily cleaves outside of the tandem repeat and consequently most EcoRI fragments containing this repeat migrate as high, molecular weight DMA. As a consequence even though this sequence may make up to as much as 3-103 of X chromosomal DMA it would 136 be expected to make up a significantly smaller proportion of EcoRI digested DMA in the size fragment range 2-14 kbp. This is the target DMA capacity of X gtt'ES "Xb and consequently the expected insert size range in the X- enriched library (and indeed the largest insert size observed was 7.9kbp). As only 1000 or so recombinants were analysed in this comparative screen, it appears likely that the repeat makes up less than one per cent of the cloned sequences present in the library and therefore of the order of one per cent of X chromosomal DMA EcoP.I fragments in the fragment size range 2-14kbp (the library is not completely representative of X chromosome sequences and so this will additionally dilute the BamHI repeat). ' 'hen cloned repetitive DMA sequence probes were used in 'in situ' hybridization experiments, two types of distribution were observed . The Hindi 11 and Alu re'peats are apparently dispersed randomly along the chromosome whilst the ocRI and BamHI repeats are clustered. It is hard to see how the clustered repeats are involved in gene regulation as postulated by Davidson and Britten (1973) but perhaps their centromeric location, similar to the distribution of human satellite DMA sequences (Mosden et al . , 1975 ), indicates a functional role in chromosome identification and segregation at mitosis and meiosis. The sequence conservation of primate alnhoid DMA sequences such as the EcoRI 340bp dimer, the African green monkey component alpha DMA (Manuelidis and Hu, 1978) and the 137 BamHI 2kbp repeat (Yanget al., 1982) as v/ell as their similar chromosomal organization agrees with this hypothes i s. It has been proposed that repeated DMA. sequences occupy different concentrations in Giemsa-dark (G) and Giemsa-1ight (R) bands (Comings, 1979). However, the apparently random distribution of flindlll and Alu sequences argue against this hypothesis. In addition, comparative analysis, of G and R band DMA from Chinese hamster cells demonstrates no differences in their rate of reassociation (Holmquist et al., 1982). The dispersed distribution of HindiII and Alu repeats provides circumstantial evidence for uniform condensation along the chromosome. If it is assumed that the periodicity of interspersion of these repeated sequences is relatively constant, as seems to be the case for intermediate repetitive sequences in general (Schmid and Deininger, 1975), then their uniform distribution implies a uniform distribution of DHA exhibiting this type of interspersion along the metaphase . chromosome. This conclusion must be taken with the caveat that a significant percentage of non-repetitive human D8A occurs in stretches of greater that 20kbp (Schmid and Deininger, 1975). In addition, these non-renetitive sequences are generally absent from highly repetitious 'satellite' !)HA which remains highly condensed throughout the cell cycle (Brutlag, 1980). 138 4.2. The construction of chromosome maps in mammals Considerable attention has been devoted recently to the new methodologies in mammalian gene mapping (Ruddle, 1431). In this study both somatic cell genetics and 'in situ' hybridization have been used in the construction of a cytological map of the human X chromosome. Each of these techniques suffers from disadvantages. The production of the human x rodent hybrid cells retaining specific human chromosomes presents difficulties in the absence of a selective marker on the chromosome such as HGPRT on the X, thymidine kinase on 17 or adenine phosphoribosyltransferase (APRT) on 16. Chromosomes which are not retained by selection are subject to continual loss during growth thus generating considerable mosaicism which must be monitored by enzyme analysis and karyotyping for accurate DHA sequence mapping. In addition, some somatic cell hybrids can be labile to chromosome rearrangement during growth which nay be beyond the resolving power of the microscope. The main drawback to the use of somatic cell hybrids in intra-chromosomal gene map-ping is the relative paucity of derivative chromosomes such as translocations or deletions. This can be circumvented in part by constructing deleted chromosomes through radiation-induced chromosome breakage (Goss and Harris, 1975). However, the characterization of such 139 fragments by karyotyping is a formidable task and their mitotic stability is poor. The most favourable aspect of somatic cell genetics is the relative ease of chromosomal assignment of a gene or DbA segment once suitable hybrids have been constructed, either by Southern blot analysis or by antigenic expression or enzyme assay. Thus for chromosomal assignment the use of somatic cell hybrids remains particularly powerful, but in the absence of a large number of well characterized derivative chromosomes its resolution remains inadequate for the construction of cytological maps. The resolution afforded by 'in situ' hybridization outstretches that of somatic cell genetics and as this work demonstrates, permits a number of discrete loci to be mapped onto a single chromosome to the level whereby genetic distances can be estimated and compared to their physical counterparts. This resolution coupled with the use of normal human chromosomes in a normal cellular environment as target are the great advantages this technique bestows upon gene mapping. The main disadvantages are the technical difficulty of the experiment and the time taken for autoradiography. The technical difficulties are surmounted by good cytogenetic practice and careful probe preparation. Autoradiographic exposure times can be reduced by the use of different 1 ? 5 35 radioactive isotopes such as “ I or 5. Iodine-125 suffers from a much lower resolving power than trittium (Ada et al., 1956) and Sulphur-35 as a higher energy beta 140 emitter causes the production of more than one grain per disintegration, thus again reducing resolution. DLTA probes v/hich are detected by non-isotopic means offer particular amelioration to this technique (Langer-Safer et al., 1932) provided trie sensitivity rivals that of radioisotopic label. The analysis of autoradiographs after hybridization represents the most subjective element to 'in situ' hybridization. This subjective element was minimized in this study in three ways. Firstly slides were banded and photographed before hybridization for unequivocal karyotyping and ensuring a random collection of hybridized spreads. Secondly autoradiographs were scored blind (i.e. ignorant of the probe) and more than one probe was hybridized in each experiment (to different slides) to minimize bias on scoring grains. Finally the statistical significance of each result was estimated by comparing the observed distribution of grains to a purely random distribution. In this fashion, a high resolution of mapping was achieved, relatively free from bias and specific to the DPT probe. The six cytogenetic loci determined in this study represent a significant contribution to a completely mapped human X chromosome. T precise human genetic map must be a prerequisite for comprehending’ each gene's function and mode of expression and thus defining all aspects of normal and abnormal human biology and 141 development. There are over one hundred X-linked Mendelian traits in man (McKusick, 1983) whose linkage to any one or more of these six loci open the way to genetic counselling and prenatal diagnosis. The deficiencies associated with the alpha and beta globin gene clusters (beatheral1 and Clegg., 1982) have become the classical example of the use of linked DU A sequence variants in prenatal diagnosis. Restriction fragment length variants have been reported assoc^ited with mutant genes (Little et al., 1980; Kan and Dozy., 1978) or caused by the mutation giving rise to a mutant phenotype itself (Geever et al., 1981). Two of the loci mapped in this study have proved useful in the mapping of the loci for Duchenne (Murray et al., 1982; Davies et al., 1983) and Becker (Kingston et al., 1983) muscular dystrophies and also in determining the carrier status of females in affected families (Penbrey et al., 1933). Using a cloned fragment of the HGPRT gene (Jolly et al., 1932) to detect BamHI restriction fragment variants and a cloned fragment of the gene for blood clotting factor IX (jaye et al., 1983), the loci detected by 52A and DX13 are shown to lie in the genetic order HGPRT-52A- FIX-PX13 relative to these genes (Dravna et al., 1934). Similarly there is equivocal data on the linkage between the locus detected by 52A and the inheritance of X-linked mental retardation associated with a fragile site at Xq27- 28 (Tariveruian and T’eck, 1982) given that the FIX locus is linked to this locus (Camerino et al., 1933). Although individually the loci mapped in this study 142 may be of limited diagnostic potential for X-linked disorders, the value of such randomly isolated genomic sequences in gene mapping and diagnosis has been made dramatically clear in the case of Huntingdon's chorea (Gusella et a1., 1983). 4.3. Localized recombination A comparison of estimated physical distances between chromosomal loci and their albeit equivocally determined genetic counterparts indicates clear differences in the case of the human X chromosome. The diagrammatical comparison in Fig. 35 indicates that recombination occurs much more frequently than expected given physical distance in proterminal regions and additionally in the vicinity of the centromere on the long arm. The primary caveat to this conclusion must be the equivocal genetic distances in some cases. In addition the assumption is that the condensation of the mitotic metaphase and meiotic pachytene DMA along the chromosome length is uniform. This assumption seems valid from electron microscopical analysis of chromosomes (DuPraw, 1970; Miller and Hamkalo, 1972) and from the uniform distribution of DMA specific stains along the chromosome (Carrano et al., 1979). The equivalence of the mitotic metaphase and meiotic pachytene chromosome seems reasonable from measurements of relative lengths and 143 centromeric indices, as well as banding patterns (Ilulten, 1974; Holm and Rasmussen, 1977). Genetic maps derived from measurements of chiasma distribution have been compiled only for chromosomes 1, 2 and 9 (tiulten et al., 1982, Laurie et al., 1982). The similarity between genetic maps prepared in that fashion with the map of the X in this work is convincing evidence for Janssens' chiasmatype theory. For example, the diagrammatic representation of chromosome 9 in Laurie et al. (1932) bears a striking resemblance to Fig. 35 representing the genetic map of the similarly sized X chromosome. The implication of localized recombination as evinced both by this work and experiments demonstrating chiasma localization is for differential DMA sequence affinity for recombination along the human chromosome's length. This conforms with chiasma localization observed in piany species, including the mouse (Polani, 1972), newts (Callan and Perry, 1977), locusts (Fox, 1973), grasshoppers (Shaw, 1974) and higher plants (Rees and Dale, 1974) and with crossing over localization in Drosophila (Schalet and Lefevre, 1971). This observation is also made at the DMA sequence level when large blocks of cloned genomic sequence are examined. Thus in a region of the short arm of chromosome 11 adjacent to the insulin gene, Lebo et al (1933) found that a tandemly repeated sequence was apparently labile to recombination in a region which exhibited overall less recombination than exoected based. 144 on physical lengths. In the immune response region from the murine major histocompatibility complex on chromosome 17/ all recombination between the I-A and I-E subregions (0. lcn) maps to a region of DMA of at most 9. Okbp (Steinmetz et al., 1982). If one extends the analogy of genetic domains in this chromosome to the human X, the 233cM chromosome would consist of 2,400 of such domains. As the X consists of approximately 1.3 x 10 kbp (Mendelsohn et al., 1973), each genetic domain would therefore correspond to approximately 75kbp, a figure which corresponds remarkably to the size of chromosome domains detected by microscopical and biochemical techniques. The non-random distribution of cross-over sites and chiasmata in Drosophila prompted Mather (1938) to propose that chiasma were formed sequentially. By his model, the first chiasma in the Drosophila oocyte forms near the centromere and subsequent chiasmata form at a distance dependent on an interference effect exhibited by the first. An exhaustive study of Locusta chromosomes caused Fox (1973) to elaborate this model such that a 'chiasma determination mechanism' passed along the chromosome at a constant rate starting from the telomeres and moving to the centromere. Interference is explained by the mechanism requiring a fixed amount of time to become 'recharged' after chiasma determination. Certainly this model explains the genetic map of the human X chromosome but Jones 145 pointed out that the spaced out distribution of chiasma does not exclude non-sequential models (Jones, 1974). In addition, he described a mutant in rye in v/hich chiasma distribution was random and interference abolished. This implies that both chiasma localization and interference are under a single genetic control. h similar mutant in Drosophila, mei-3 52, has also been described (Baker and Carpenter, 1972). Jones concluded that interference and chiasma localization were a consequence of control of recombination rather than control mechanisms in their own right. Electron microscopic analysis and three-dimensional reconstruction of meiotic cells has revealed structures which are strong candidates for the 'machinery' of recombination. Termed recombination nodules they were first described in Drosophila oocytes (Carpenter, 1975) and have subsequently been found in other species including human spermatocytes (Rasmussen and Holm, 1973). Two types of nodule are seen in Drosophila♦ The first type are roughly spherical, appear late in pachytene and in number per nucleus, number per chromosome arm and distribution along each arm correspond './ell to reciprocal exchange events. The second type appear more ellipsoidal in shape and are seen earlier in pachytene. From their uniform distribution it is postulated that the sites of these structures correspond to gene conversion events. From examining the morphology and; distribution of nodules in meiotic mutants Carpenter has proposed that all the 146 properties for chiasma determination and formation lie in the nodule (Carpenter, 1079). Her hypothesis is that the earlier ellipsoidal nodules determine all recombinational events and the subsequent determination of a site for reciprocal exchange results in conversion of an ellipsoidal nodule to a spherical nodule. This structure then performs the enzymatic functions of strand isomerization and resolution. The mutants moi-210 and mei- 41 exhibit reduced exchange and have morphologically abnormal nodules which are reduced in numbers. Carpenter concludes that these mutants synthesize reduced numbers of functional nodules and those that form functionally are abnormal in their ability to respond to (or generate) the constraints regulating chromosomal locations of reciprocal events. In contrast, the mutant mei-9 exhibits hinhly reduced exchanges yet has morphologically normal nodules and a normal number of nodules. Carpenter postulates that in this mutant the nodules are deficient in some enzymatic process of recombination such as strand isomerization in the iieselson-Padding model (Peselson and Padding, 1975). The role that chromosome architecture might '''lay in determining chromosomal regions sensitive to recombination was stressed by Hulten in her demonstration of localized chiasmata in human spermatocytes (llulten, 1974). It is interesting in this respect to note that a region of the X chromosome that is apparently ’prone to meiotic recombination exists between the loci detected by 52A and 147 DX13 (Xq27 - Xqter). The same region also exhibits a physical lability in the manifestation of a fragile site (brookwell and Turner, 1933). From results demonstrating differing sensitivity of crossing over to translocation heterozygosity in different chromosomal regions, Roberts (1972) concluded that the initiation of synapsis was crucial to chiasma determination. The hypothesis that the short stretches of synaptonerna 1 complex formed early in leptotene were initiated by special repeated sequences was put forward by von bettstein (1971). He saw the construction of the synaptonema1 complex as a result of folding of these sequences onto each other to bring the element to a continuous thread. The distribution of such sequences along the chromosome might determine which parts of the chromosome are able to recombine more frequently. Maguire calculated the frequencies of recombination she would expect between chromosomes with limited genetic homology. From higher than expected frequencies of recombinants obtained from crosses in maize between chromosomes of differing extents of homology in genetically marked regions, she proposed that chiasna were determined before synapsis was complete (Maguire, 1977). This indicates that regions labile to recombination are sites at which synapsis is initiated. Rasmussen and Holm (1930) speculated that the distribution of recinrocral exchange events is governed not just by the sites at which synapsis is initiated but also the rate at which synapsis occurs. 148 From the features of the recombination nodules and electron microscopic observations on synaptonemal complex formation they postulate a model delineating four constraints on chiasma determination; 1) A fairly constant total number of nodules are synthesized or assembled during zygotene. 2) A variation in the rate of chromosome pairing and synaptonemal complex formation. 3) A temporal variation in nodule availability during zygotene and pachytene. 4) A mechanism whereby nodules can be selectively removed from the synaptonemal complex during the zygotene- pachytene transition. In this way the sites at which reciprocal exchange events occur is governed both by the DMA and/or chromosome structure and by the recombination nodule. It is manifest that chiasma localization occurs across a wide range of species and is presumably an ancient device in evolution. In addition, changes in chiasma distributions and frequencies can be selected for (Rees and Dale, 1974). These observations have led Carson (1975) to the hypothesis that recombination cannot reach into every corner of the genome and he postulates two types of heritable genetic variability systems, an 'open' variability system and a 'closed' variability system. he proposes that the genetic components of the open variability system operate singly, relatively late in 149 development and are freely made available to the vagaries of natural selection or random drift by chance recombination. They are consequently the basis of easily adaptive differences found within a species. As a contrast, the closed variability system consists of blocks of genes which have come to be inseparably associated in a coadapted chromosome segment. Such a coadaptive arrangement has been termed 'internal balance' (father, 1943) or 'supergenes' (Darlington and Mather, 1949). .Although it is not a necessary consequence, such associations create a genetic climate for the development of mechanisms leading to chiasma localization. Carson sees the closed variability system as evolving early in the history of species and becoming essentially fixed in the form of normally unchangeable heteroses and internally balanced sets of genes which oblige the species to solve its problems in a certain way. As a consequence he proposes a non-gradualistic mode of speciation which can occur within a few generations by means of a population flush-crash-founder cycle. The flush phase is characterized by a release from natural selection which leads to the break up of coadaptations and supergenes of the closed variability system (such population flushes in nature are accompanied often by a release of bizarre genetic variability). A sudden change in the environment leads to a population crash with occasionally the survival of but a single propagule which may well be by chance alone as selection would not have enough time to act. This 150 founder is seen as a highly unusual recombinant coming out of the closed variabilty system which only survives by the absence of competition. The organization of a new closed variability system ensues when the expanding population is once again brought under natural selection. A direct test of Carson's hypothesis has been made available in the grasshopper genus Caledia which exhibits differences in chiasma distribution both between species and between chromosomal taxa (Shaw and Knowles, 1976). Two chromosomal taxa of Caledia cantiva differ by seven pericentric rearrangements, heterochromatin and allozyme differences. The Moreton taxa has interstitial chiasmata whilst the Torresian taxa exhibits a strictly proxima1/distal chiasma pattern. This consequently implies differences in their closed variability systems according to Carson. Coates and Shaw (1992) examined the chiasma distribution in the generation of a cross between the two taxa. They found a repositioning of chiasmata in all chromosomes leading to maximum recombination in interstitial regions thus disrupting the presumed internal balance of coadapted gene combinations and leading to novel genotypes of unpredictable fitness. In an accompanying paper, Shaw et al. (1932) analysed directly the effects of this recombinational repatterning by estimating the relative viabilities of recombinant and non-recombinant chromosomes (identified by neterochromatic markers) through their distribution in offspring from 151 backcrosses of the F^ with both parent taxa and in the F . They found that although the F^ exhibited normal viability and fertility, the F^ were completely inviable and the backcross progeny exhibited a highly reduced viability. In addition, the relative viabilities of chromosomes in backcross progeny corresponded well to the structure of the hybrid zone between the Moreton and Torresian taxa. They concluded that each chromosome of each taxa is internally coadapted in a functional sense and such coadaptation is required for harmonious gene interactions within the same chromosomes during development. This induced their speculation that the recombination system plays an integral and directing influence in maintaining karyotype uniformity. The selective factor which is responsible for the survival or extinction of variants relates primarily not to extraneous environmental factors, but to the acceptability of the variant by the maternal genome which discriminately activates only those zygotic genomes which correspond to its own pattern of chromosomal organization. Rees and Dale (1074) considered the effect of recombinational repatterning in selected lines of Festuca and bo1iurn populations in less drastic terms. They acknowledged the selective disadvantage of repatterned chromosomes and saw this as an overall dissipation and depletion of the heritable variation both in the potential and the free state. The observation of localized recombination in the human X chromosome provides circumstantial evidence for 152 the existence of coadapted genetic components locked in epistases. The inference from other species is clearly, however, that this pattern of recombination may have been preserved through the evolution of the species to maintain the internal balance of genetic components v/ithin the chromosome and also as a species barrier. One might infer that the v/ithin species variation in recombination pattern in Galedia (Shaw and Knowles, 1976) might also exist in homo Sapiens and this would explain the differences in recombination fraction observed between the same loci in different populations (Weitkamp, 1974). 153 CHAPTER 5 Summary and 'Perspectives 154 This research project has investigated the pattern of genetic recombination in the human X chromosome. It provides convincing evidence for Janssens' chiasnatype theory when comparisons are made to the chiasma distribution along male autosomes. In addition, it suggests non-uniformity of recombination frequency along the X chromosome. This may be an accidental result of chromosome structure or a consequence of the existence of regions inaccessible to crossing over. In this respect it would be interesting to examine the distribution of coding sequences along the 1C chromosome. A random distribution would support the hypothesis that genetic components may not recombine freely in all parts of the genome. The construction of a genetic map of the ''/hole human X chromosome permits linkage analysis between any of the markers used and any X-linked genotype. It has been estimated that the majority of loci must lie within 20cX of the marker loci analysed in this study. The delineation of aberrant X-linked genotypes to a small region of the chromosome permits diagnosis of such X-linhed disorders through the determination of the genetic constitution of the surrounding chromosome. In addition, analysis of RITA transcripts from this region in normal and affected individuals may allow the isolation of the particular D’TA an d Rd A segments responsible for the affected condition. This would then facilitate the indroduction of a functional allele into a mutant background once the technology for such 'gene therapy' exists. 155 CHAPTER 5 References 15 6 Ada, G . L . , Humphrey, J.H., Askonas, B . A. , HcDevitt, II. Q. and Hossal, G.J.V. (19G6) Exp. Cell Res. £1, 557-572 Bak, A. Leth, Zeuthen, J. and Crick, F.II.C. (1977) Proc. Uatl. Acad. Sci. USA 74, 1595-1599 Baker, B.S. and Carpenter, A.T.C. (1972) Genetics 71_, 255- 28 G Baker, B.S., Carpenter, A.T.C., Esposito, 14.3. , Esposito, R.C. and Sandler, L . (1975) Ann. Rev. Genet. 10, 53-134 Benton, U.D. and Davis, R.U. (1977) Science 195,180-182 Benyajati, C. and T'orcel, A. (1976) Cell 9, 393-407 Blattner, F.R., Williams, B.G., Blechl, A.E., Denniston- Thompson, K., Fiiber, II. E. , Furlong, L.-A., Grunv/ald, D.J., Kiefer, D.O., Moore, D.D., Sheldon, H.L. and Smithies, O. (1977) Science 196, 161-169 Botstein, D., White, R . 7j . , Scolnick, M. and Davis, R.W. (1980) Arn. J. Hum. Genet. _32, 314-3 31 Boyer, H. and Roulland-Dussoix, D. (1969) J. Mol. Biol. 41, 459-472 Britten, R.J. and Kohne, D.E. (1968) Science 151, 529-540 Britten, R.J., Graham, D.E. and Ueufeld, R.R. (1974) Methods in Enzvmology 29E, 363-^06 Brookv/ell, R. and Turner, G. (1993) Hum. Genet. 63, 77 Brown, S. and Zonary, D. (1955) Genetics 40, 850-873 D r u f l a g , D . L. (1980) Ann. Rev. Genet. 14, 121-144 C a l l a n , H .G . and Perrv, P.E. (1977) Phil. Trans B o n d . B 277, 277-233 Camerino, G., Mattei, H . G . , Uattei, J . F . , Java, I'. and I'and el, J.-L. ( 1923) Mature 306, 701-704 Carpenter, A.T.C. (1975) Proc. Uatl. Acad. Sci. USA 72, 3186-3189 Carpenter, A.T.C. (1979) Chromosoma 7H5, 259-292 Carrano, A.V., Gray, J.U., Langlois, B.G., Burkhart- Schultz, K.J. and Van Dilla, U.A. (1979) proc. Uatl. Acad. Sci. 76, 1382-1384 Carson, U.L. (1975) Am. Uatur. 109, 93-92 157 Casadaban, M.J. and Cohen, S.M. (1930) J. Mol. Biol. 133, 179-201 Coates, D.J. and Shaw, D.D. (1932) Chromosoma 3^5, 509-531 Comings, D.H. (1972) Adv. Hum. Genet. 3, 237-243 Comings, D . F.. (1973) Ann. Rev. Genet. 12, 25-46 Creighton, H .B. and HcClintock, E. (1931) Proc. Hatl. Acad. Sci. USA 17, 492-497 Darling, S.M., Crarapton, J . H. and Uilliamson, R. (1932) J. Mol. Biol. 154, 51-53 Darlington, C.9. (1931) Biol. Rev. 6, 221-234 Darlington, C.D. and feather, K. (1949) The Elements of Genetics. /Alien and Unwin, London Davidson, E.1I. and Britten, R.J. ( 197 3 ) 9. Rev. Biol. 43, 565-513 Davies, K.E. (1931) Hum. Genet. 53, 351-357 Davies, K.R., Young, R.D., Elies, R.G., Hill, H . H. and ’•Jill iamson, R. (1931) Mature 293, 374-375 Davies, K.E., Pearson, P.L., Harper, P.S., i'urrav, J.f ., O'Brien, T., Sarfarazi, ! ■. and Williamson, R. ( 1933 ) Duel. Acids Res. 11, 2303-2312 Davies, K.E., Hartley, p .A ., Hurray, J.H., Harper, P.S., Casey, G., Taylor, P. and Williamson R.(1933R) Cold Spring Harbor Banbury Reports Denhardt, D.T. (1966) Biochem. Biophvs. Res. Cominun. 23, 641-545 ~ " Doolittle, b.F. and Sapienza, C. (1930) Mature 234, 601- GO 3 Drayna, D., Davies, H . , 'hartley, D., Handel, J.-L., Carnerino, G., hi 11 iamson, R. and Tihite, R. (1984) Proc. Hatl. .Acad. Sci. USA, manuscript submitted DuPraw, E.J. (1963) Cell and Molecular Biology. Academic Press, Dew York DuPraw, E.J. (1970) DMA and Chromosomes. Holt, Rinehart and Winston, Dew York Ferguson-Smith, M.A., Sanger, R., Tippett, P., Aitken, D . A . , and Boyd, E. ( 1932) Cytogenet. Cell Genet. 3_2, 273- 27 4 153 Finch, J.T. and Nlug, A. (1976) Proc. Natl. Acad. Sci. USA 73, 1397-1901 Fox, D.P. (1973) Chromosoma 43, 289-323 Gall, J.G. (1963) Nature 198, 36-38 Gall, J.G. and Pardue, M.L. (1969) Proc. Natl. /Acad. Sci. USA S3, 373-383 Gall, J.G. and Pardue, M.L. (1971) Methods in Enzymology 21, 470-480 Geever, P.F., I/ilson, L.B., Nallaseth, F.S., Milner, P.F., Bittner, M. and Wilson, J.T. (1981) Proc. Natl. Acad. Sci. USA 70, 5031-5035 Good fellow, P., Banting, G., Levy, R., Povev, S. and Mcl-iichael, A. (1930) Somat. Cell Genet. _6, 777-733 Good fellow, P., Banting, G. Trowsdale, J., Chambers, S. and Solomon, E. ( 1932) Proc. Natl. Acad. Sci. IJS/A 79, 1190-1194 Gosden, J.R., Mitchell, A.R., Buckland, R.A., Clayton, R.P. and Evans, li.J. ( 1975 ) Exp. Cell Res. 9_2, 143-158 Goss, S.J. and Harris, II. ( 1975) Nature 255, 680-684 Gusella, J.F., Keys, C., Varsanyi-Breiner, A., Kao, F-T., Jones, C., Puck, T.T. and iiousman, D. (1930) Proc. Natl. Sci. USA 77, 2829-2833 Gusella, J.F., Uexler, N.S., Conneally, P . r •., Naylor, S.L., Anderson, II. A., Tanzi, R . E. , Watkins, P.C., Ottina, K., Wallace, M.R., Sakaguchi, h. Y , , Young, A.B., Shoulson. I., Bonilla, E. and Martin, J.B. (1933) Nature 306, 234- 2 38 Haldane, J.B.S. (1931) Cytologia _3, 54-65 Hand, R. (1978) Cell 1_5, 317-325 Harper, M.E. and Saunders, G.F. (1981) Chromosona 8J3, 431- 439 Harper, IN E . , Barrera-Saldana, II. A. and Saunders, G.F. (1982) /Am. J. Hum. Genet. _3_1, 227-234 Ueintz, N., Zernik, M. and Boeder, R.G. (1981) Cell 24, 661-668 Holm, P.B. and Rasmussen, S.W. (1977) Carlsberg. Res. Commun . 4 2 , 2 8 3-323 159 Holmquist, G., Gray, [ . , Parker, T. and Jordan, J. (1982) Cel1 21' 121-129 Houck, C.M., Rinehart, F.P. and Schmid, C.b. (1979) J. ko1. Biol. 132, 289-306 Hozier, J.C., Renz, M. and Mehls, P. ( 1977) Chromosome 62, 301-318 Hulten, M . (1974) Hereditas jREp 55-73 Hulten, r.. , Palmer, R.U. and Laurie, D . A.. (1982) Ann. dura. Genet. 4_6, 167-175 ISCU (1981) .An International System for Human Cytogenetic 1'Ionenclature - High Resolution Banding. Cytogenet. Cell Genet. 3_1, 1-23 Igo-Remenes, T. and Zachau, II.G. (1977) Cold Spring Harbor Symp. Quant. Biol. 42:, 109-118 Igo-Kerrienes, T . , Slorz, V. and Zachau, H.G. ( 1982) Ann. Rev. Biochem. ^5^, 89-121 Jackson, D.A., Symons, R.H. and Berg, P. (1972) Proc. Hatl. Acad. Sci. USA 69, 2904-2909 Janssens, F.A. (1909) Cellule 2_5, 387-411 Jaye, 11. de la Salle, H., Schanber, F., Balland, A., Hohli, V., Fideli, A., Tolstoshev, P. and Lecocg, J.-P. (1983) Mucl. Acids Res. 11, 2325-2335 Jelinek, W.R., Toomey, T.P., Leinv/and, L., Duncan, (3. II., Biro, P.A., Choudary, P.V., Veissman, 5.4., Rubin, C.H., Houck, C.H., Deininger, P.L. and Schmid, C.b. (1980) Proc. Hatl. Acad. Sci. USA 11_, 1393-1402 Jelinek, I /.R. and Schmid, C.b. ( 1982) Ann. Rev. Biochem. 51, 813-844 Jepperson, P.G.LT. and Lankier, A.T. (1979) Fuel. Acids Res. 1_, '19-67 Jolly, D.J., Fsty, A.C., Bernard, U.U. and Friedmann, T. (1982) Proc. Hatl. Acad. Sci. USA 7_9, 5038-5041 Jones, G.H. (1974) nature 250, 147-143 Jones, K.b. (1973) In Hew Techniques in Biophysics and Cell Biology, Pain, R.H. and Smith, B.J., eds., pp29-56, J. Wiley and Sons, London Ran, Y.W. and Dozy, A.H. (1978) Lancet ii, 910-911 160 Kavenoff, R. and Zimin, B.H. ( 197 3) Chromosoma 4U, 1-20 Kingston, II. M., Thomas, W.S.T., Pearson, P.L., Sarfarazi, M. and Harper, P.S. (1933) J. Med. Genet. 20, 255-253 K u r n i t , D.M. and lloehn, H. (1979) A n n . Rev. Genet. 13, 2 3 5 - 2 5 3 L a e m m l i , U .K ., C h e n g , S .M ., Zidolnh, K . U., P a u l s o n , J .R ., Brown, J.A. and Baumbach, L’.R. (1977) Cold Spring Harbor Symp. Quant. Biol. 4_2, 351-355 Langer-Safer, P.R., Levine, M. and bard, D.C. (1982) Proc. Hat1. Acad. Sci. USA 79, 4331-4385 Latt, S.A. (1931) .Ann. Rev. Genet. _1_5, 11-55 Laurie, D .A., Palmer, R.U. and Hulten, H .A. (1932) Ann. Hun. Genet. 4j3, 233-244 Lebo, R.V., Chakravarti, A., Buetow, K.H., Cheung, M.-C., Gann, h., Cordell, B. and Goodman, II. (1983) Proc. Uatl. Acad. Sci. USA 80, 4308-4312 Lefevre, G., Jr. (1971) Genetics _67, 497-513 Littl e, P . F . R.. , Addison, G., barling, S., Uilliamson, R., Camba, L. and Lodell, B. (1980) Mature 235, 144-147 Lyon, I:. F. ( 1976) Genet. Res. _28, 291-300 IIcKusick, V . A. (19:30) J. Heredity 7_1, 370-391 McKusick , V.A. (1933) Uendelian Inheritance in Kan : Catalogs of /autosomal Dominant, Autosomal Recessive and K - linked Phenotypes, 6th edn.. Johns Hopkins University Press, Baltimore Maguire, K.P. (1977) Phil. Trans. R. Soc. Lond. B 277, 245-258 Malcolm, S., Barton, P., Murphy, C., Ferguson-Smith, K.A., Bentley, D.L. and Babbitts, T.H. (1982) Proc. Uatl. Acad. Sci. USA 19_, 4-957-4961 Maniatis, T., Hardison, R.C., Lacy, E., Lauer, J., O'Connell, C., Quon, D., Sim, G.K. and Ufstradiatis, A. (1978) Cell 15, 637-701 Maniatis, T ., Frit s ch , H .F . and S a m b r o o k , J . (1932) Mole c u l a r cloning (A laboratory nianua 1 ) . Cold Spring Harbor Laboratory Press, Hew York Manuel idis, L. and 'u, J.C.( 1973) Mature 276, 92-94 161 Manuelidis, L . and Biro, P.A. (1982) Nucl. Acids Res. 10, 3221-3239 Barsden, II. and Laemmli, U.K . (1979) Cell Mather, X . (1933) Biol. Rev. 1_3, 252-292 lather, X . (1943) Biol. Rev. 1-8, 3 2-64 Maynard-Smith, J. (1978) The Evolution of Sex, Cambridge University Press I lend elsohn, H.L., Mayall, B.H., Bogart, H., tioore, D.U., r II and Perry, B .H. (1973) Science 179, 1126-112':) Meselson, M . S . and Radding, C.M. (197 5) Proc. !: a 11. Acad. Sci. USA 72, 35:3-361 Miller, 0 . I. . , Jr. and Llamkalo, r . a (1972) Int. Rev. Cytol. 33 , 1-23 M o o r h e a d , P . S . , Howell, P.C., Heilman, . J . , Bat imps , D . H . and Hunge rford, D.A. (1960) Exn. Cell. R e s . 20, 613-616 Morgan, T.II. (1911) J. Exp. Zool. 11, 305-414 Morgan, T.H. and Cattell, E. (1912) J. Exn. Zool. _13, 79- 101 Morton, M . E. ( 19 56) <\n o. Hum. Genet. 1_, 277-231 Moses, M .J . ( 195 8 ) J Biophysic. Biochem. Cytol. 4, 6 3 3- 633 Muller, il.J. (1916) Am Uatur. 50, 193-221 Murray, J.M., Davies, K.E., Harper, P.S., Meredith, L,. , Mueller, C.R. and hi 11 ifimson, R. (1982) Mature 300, 69-71 Hurray, J.M. (1984) PhD thesis, University of Cardiff, Males Mabholz, U., Higgiano, V. and Bodmer, MM (1969) Mature 223, 358-363 Ueuffer, M.G. and Coe, E.U., Jr. (1976) In Handbook of Diochemical and Molecular Biology,3rd edn., G.D. Pasoan, Ed. Cleveland CRC Press Orgel, L.E. and Crick, F.H.C. (1930) Mature, 234, 604-607 Ott, J. (19 74) Am. J . Hum. Genet. _26, 588-597 Oudet, P., Gross-Bellard, M. and Chambon, P. (1975) Cell 162 4, 201-300 Page, D., deMart invi 1 le, B. , Barker, D., Hyman, A., White, R.L., Francke, U. and Botstein, D. (1902) Proc. ITatl. Acad. Sci. USA 79, 5352-5356 Paulson, J.R. and Laemmli, U.K. (1977) Cell _12, 317-029 Pembrey, M.E., Davies, K.E., b'inter, R.M., Elies, R.G., Williamson, R., Fazzoni, T.A. and Calker, C. (1933) Arch. Dis. Child, in the press Polani, P.E. (1972) Chromosoma 3_6, 343-374 Race, R.R. and Sanger, R. (1975) Blood Groups in Man, 6th edn. . Blackv/ell .Scientific Publications, Oxford Rasmussen, S.w. and Holm, P.B. (1973) Carlsberg. Res. Commun. 4_3, 27 5-3 27 Rasmussen, S.l . and Holm P . i3. (1930) Ilereditas 93, 137-215 Rees, H. and Dale, P.J. (1974) Chromosoma 47_, 335-351 Ricciuti, F.C. and Ruddle, F.II. ( 1973 ) Genetics 74, 661- 573 Rich, A. (1932) Cold Soring Harbor Symp. Quant. Biol. 47, 1-12 Rigby, P.J.W., Dieckrnann, M., Rhodes, C. and Berg, P. (1977) J. Mol. Biol. 113, 237-251 Ringertz, H.R. and Savage, R.E. (1975) Cell Hybrids. Academic Press, Iiew York Roberts, P .A. (1972) Genetics 71, 401-415 Ruddle, F.H. (1931) Mature 294, 115-120 Schalet, .A. and Lefevre, G., Jr. (1976) In ' kshburner and E. Hovitski (Eds.), The Genetics and iology of Drosophila. Academic Press, London, ppS43-902 Schmid, C.W. and Deininger, P.L. ( 197 5 ) Cell o_, 345-352 Seabright, I:. (1971) Lancet ii, 971-972 S h a f i t - 3 a g a r d o, Maio, J.J. and Drown, F.L. (1932) duel. Acids Res. IT), 3175-3193 Uhaw, D.D. (1974) Chromosoma 4^5, 36 5-374 Shaw, D.D. and Knovdes, G.R. (1975) Chromosome 59, 103-127 163 Shaw, D.D., Uilkinson, P. and Coates, D.J. (1982) Chromosoma 3G, 533-549 Solomon, E., Bobrov;, 15., Good fellow. P.N., Bodmer, MT.E., Swallow, D.M., Povey, S. and doe 1, I'.. (1975) Sornat. Cell Genet. 43, 241-248 Southern, E.M. (1975) J. Mol. Biol. 98, 503-517 Stadler, J., Larson, A., Engel, J.D., Dolan, I'., Groudine, M. and Ueintraub, II. (1900) Cell 2!0, 451-450 Steinmetz, H., Hinard,R., Horvath, S., EcHicholas, J., Srelinger, J., Uahe, C., Long, H., Mach, B. and Hood, L. (1982) Mature 300, 35-42 Stern, C. (1931) Biol. Sbl. _51, 547-587 Sturtevant, A.II. (1913) J. Exp. Zool. _14, 43-59 Sutton, \:.S. (1903) Biol. Bull. (Moods Hole) 4, 231-243 Szabo, P., Elder, R., Steffenson, D.M. and Uhlenbeck, O.C. (1977) J. Mol. Biol. 115, 539-553 Szabo, P. and Hard, D.C. (1982) Trends Diochem. Sci. 7, 425-427 “ Tariverdian,G. and Meek, B . (1982) Hum. Genet. G2, 95-109 Taylor, J.H. (1958) Sci. Am. 198, 36-47 Taylor, J.H. (1963 ) In J.II. Taylor (Ed.) Molecular Genetics. Academic Press, Hew York, pp65-113 Taylor, J.H. (1965) J. Cell. Biol. 25, 57-57 Tease, C. and Jones, G.H. (1978) Chromosoma 59, 163-178 Tiemeier, D ., Enquist, E. and Leder, P. (1976) Mature 263, 526-527 Trent, J.M., Olson, S. and Lawn, R.H. ( 1982) Proc. Mratl. Acad. Sci. USA 22/ 7309-7013 vorT 'ettstein, D. (1971) Proc. Natl. Acad. Sci. MTS A 53, 351-855 Mahl, G.H. , Stern, M:. and Stark, G.R. ( 1979) Proc. Hatl. Acad. Sci. (JSA 7_6, 3683-3637 Heatherall, D.J. and Clegg, J.3. (1982) Cell 29, 7-9 Heismann, A. (1887) Mature Lond. 36, 607-609 154 Weiss, IRC. and Green, II. (1967) Proc. llatl. Acad. Sci. USA 59, 1104-1111 Weitkamp, L.R. (1974) Cytogenet. Cell Genet. 1^3, 179-132 I Jes tergaard, II. and von Wettstein, D. ( 1972 ) Ann. Rev. Genet. j5, 71-110 Wetrnur, J.G. ( 1975 ) Biopolymers L4, 2517-2524 White, K.J.D. (1973) /Animal cytology and evolution, 3rd edn.. Cambridge University °ress 1. ieacker,P ., Davies, K.E., Cooke, II.J., Pearson, P.L., ’./illiamson, R. , Southern, E., Zimmer, J., and Ropers,II.-II. (1933) Am. J. Hum. Genet., in the press Willard, H.F., Smith, K.D. and Sutherland, J. (1982) duel. Acids Res. 11, 2017-2033 Winder, P.E. and Steffensen, D.U (1973) Ann. Rev. Genet. 1_, 205-223 hood, W.B. ( 1936) J. Hoi. Riol. IjS, 113-133 W;u, J.C. and Ilanuelidis, L. (1930) J. Uol. Biol. 112, 3 5 3- 3 36 Yang, T.P., Hansen, S.R., Oishi, R.W., Ryder, D.A. and Hamkalo, B.A. (1982) Proc. Uatl. Acad. Sci. USA 7_9, 6593- 5597 165 ► 166 167