Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Molecular genetic analysis of human 8p inversion duplication chromosomes Nelson, Tanya N. 1998

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1998-34598X.pdf [ 13.43MB ]
Metadata
JSON: 831-1.0088696.json
JSON-LD: 831-1.0088696-ld.json
RDF/XML (Pretty): 831-1.0088696-rdf.xml
RDF/JSON: 831-1.0088696-rdf.json
Turtle: 831-1.0088696-turtle.txt
N-Triples: 831-1.0088696-rdf-ntriples.txt
Original Record: 831-1.0088696-source.json
Full Text
831-1.0088696-fulltext.txt
Citation
831-1.0088696.ris

Full Text

M O L E C U L A R GENETIC A N A L Y S I S OF H U M A N 8p INVERSION DUPLICATION C H R O M O S O M E S  by  TANYA N . NELSON B.Sc. The University of British Columbia, 1993 A THESIS SUBMITTED IN P A R T I A L F U L F I L M E N T OF THE REQUIREMENTS FOR THE D E G R E E OF DOCTOR OF PHILOSOPHY in THE F A C U L T Y OF G R A D U A T E STUDIES (Medical Genetics Programme) We accept this thesis as conforming to the required standard  t h e  TjNjA^Rsrf Y o f r k f n s H  C o l u m b i a  September 1998 ©Tanya N . Nelson, 1998  In presenting this thesis degree  in partial fulfilment  of the  requirements  for an advanced  at the University of British Columbia, I agree that the Library shall make it  freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes department  or  by  his  or  her  representatives.  may be granted It  is  by the head of my  understood  that  copying  or  publication of this thesis for financial gain shall not be allowed without my written permission.  Department of U r e C M r jv>_  CcnJ£T\U,  The University of British Columbia Vancouver, Canada  Date ^ C ^ y j i & E t L  DE-6 (2/88)  3-9)  }  W f t  Abstract  Inversion duplications of the short arm of chromosome 8 (8p) with common morphology have been described in over 60 mentally retarded individuals. These aberrant chromosomes contain material derived from both maternal chromosomes, separated by a single copy region at a common center of symmetry, with deletion of distal subtelomeric sequences.  A mechanism  mediated by inverted repetitive elements may explain the recurrence of these similar aberrant chromosomes in unrelated patients. A megasatellite repeated sequence, localized to chromosome 8p23, was investigated as a candidate for the proposed repetitive elements. Cosmid clones isolated from a single chromosome 8 library contained three classes of megasatellite. Megasatellite-containing Y A C clones map to two locations on chromosome 8p consistent with the flanking positions predicted by the known extent of the single copy region. Although the megasatellite sequences span 16 kb, all cosmids containing the megasatellite cross hybridize outside of this region. A B A C clone, that lacks the megasatellite but contains these crosshybridizing sequences within which the megasatellite is embedded, was used as a FISH probe to metaphase chromosomes. Hybridization occurred at multiple locations throughout the genome, including 8p23. Therefore, the 8p copies of the megasatellite are each embedded within a region of at least 160 kb that is itself reiterated throughout the genome. These results suggest that the megasatellite, embedded within a large reiterated region of the genome, may be involved in the generation of inversion duplication (8p) chromosomes by providing a site for anomalous interchromosomal recombination. Investigation of a patient where the single copy region could not be detected suggests that, in rare cases, other mechanisms may be involved.  Table of Contents  Abstract  ii  Table of Contents  iii  List of Tables  vi  List of Figures  vii  Acknowledgements  ix  Chapter 1: Introduction  1  1.1 Human Chromosome Aberrations  1  1.2 Translocations  1  1.3 Duplications, Deletions, Inversions  3  1.4 Inversion Duplications  7  1.5 Inversion Duplication 8p (inv dup(8p))  9  1.6 Project Objectives  22  Chapter 2: Materials and Methods  23  2.1 Polymerase Chain Reaction  23  2.2 Polymorphic STS Genotyping  28  2.3 Polyacrylamide Gel Electrophoresis (PAGE)  29  2.4 Cosmid D N A Preparation and Isolation  30  2.5 Bacterial Artificial Chromosome (BAC) D N A Isolation and Preparation  33  2.6 Yeast Artificial Chromosome (YAC) D N A Preparation and Isolation  35  2.7 Restriction Enzyme Digestion  37  2.8 Agarose Gel Electrophoresis  39  2.9 D N A Imaging for Agarose Gels  41  2.10 Southern Transfer  41  2.11 Radioactive Labeling of D N A  42  2.12 Pre-Hybridization Treatment of Probe  44  2.13 Hybridization, Post-Hybridization Washes, Autoradiography  45  2.14 Extraction of D N A from Agarose Gel  46  2.15 Subcloning of Cosmid Fragments  47  2.16 Transformation of Ligated D N A into Competent E.coli Cells  49  2.17 D N A Sequencing  50  2.18 Cosmid Termini Isolation  53  2.19 B A C Termini Isolation by Bubble PCR  55  2.20 Preparation of D N A from Whole Blood  58  2.21 Epstein Barr Transformed Cell Line  59  2.22 Metaphase Chromosome Preparation  60  2.23 Slide Preparation and Storage  61  2.24 Fluorescence In Situ Hybridization (FISH)  61  Chapter 3: Results  65  3.1 Polymorphic STS Genotyping  65  3.2 Refinement of the Map of the Region Predicted to Contain the Distal Element  68  3.3 Isolation of an 8p Repetitive Element  74  3.4 Subcloning of EcoRI Fragments from Cosmids 153G8 and 39A7  82  3.5 Sequencing of Clones 153G8E3.35 and 39A7E3.8  86  3.6 Isolation of B A C s Containing Megasatellite Sequences  89  3.7 Megasatellite Content in Chromosome 8 Y A C s  91  3.8 STS Content Analysis of Megasatellite Containing Clones  96  3.9 Analysis of the LRS on Chromosome 8p  97  3.10 Lack of Megasatellite in BACs 223B23, 67B6, and 286D2  104  3.11 Mapping the LRS by FISH using B A C 223B23  105  3.12 Characterization of Five Additional Patients  105  3.13 Mapping the Megasatellite to Patient Chromosomes Using FISH  117  Chapter 4: Discussion  120  4.1 Refinement of the Location of the Distal Repetitive Element Predicted by the Hypothesis  120  4.2 Refinement of the Y A C Contig Map in the Region Predicted to Contain the Distal Repetitive Element  121  V  4.3 A Candidate Novel Repetitive Element  122  4.4 The Organization of the Megasatellite on 8p  125  4.5 The Evolution of 8p Megasatellite  129  4.6 Analysis of Patients by STS Genotyping and FISH of the 8p Megasatellite  130  4.7 FISH as a Tool for Determining the Presence of a Single Copy Region  135  4.8 A Single Copy Region May Not be Found in Rare Cases of Inv Dup(8p)  136  4.9 Genomic Organization of 8p23.1  137  4.10 Conclusions  139  4.11 Future Research  140  Bibliography  142  Appendix 1: Sequence Comparisons  153  Appendix 2: BACs Isolated from the Research Genetics BAC library  160  vi  List of Tables  Table 1: A comprehensive list of reported cases of inversion duplication (8p) patients  11  Table 2: STS marker information  24  Table 3: Summary of polymorphic STS genotyping of inv dup(8p) patients 1-9  67  Table 4: Y A C information  70  Table 5: Summary of Y A C STS content data  71  Table 6: Restriction enzyme fragment sizes characteristic of the 8p megasatellite sequences  83  Table 7: Hybridization of randomly cloned EcoRI fragments to megasatellite containing cosmids  85  Table 8: Summary of polymorphic STS genotyping data for patients 10-14  113  List of Figures  Figure 1: Proposed mechanisms for generation of recurrent rearrangements  4  Figure 2: Ideogram of normal chromosome 8 and inversion duplication (8p) chromosome  10  Figure 3: Linkage maps of the region of chromosome 8p of interest  19  Figure 4: A proposed mechanism for the generation of inversion duplication(8p) (Mitchell etal, 1991) Figure 5: A proposed mechanism for the generation of inversion duplication(8p)  20  (Weleber et al, 1976, Dill et al, 1987, Floridia et al, 1996)  21  Figure 6: Construction of T3 end deletion clones from cosmid clones  54  Figure 7: Bubble P C R protocol for B A C end isolation  57  Figure 8: Genotyping of 9 families with STS polymorphic marker D8S503  66  Figure 9: STS content mapping in Y A C s  72  Figure 10: Examples of STS content analysis by PCR  73  Figure 11: Homology of STS markers to 4p megasatellite sequence  77  Figure 12: Hybridization of megasatellite PCR products to EcoRI digested cosmid 153G8  78  Figure 13: Restriction map of cosmid 153G8  79  Figure 14: Characterization of megasatellite containing cosmids clones  80  Figure 15: Cross hybridization of megasatellite containing cosmids  84  Figure 16: Cross hybridization of flanking sequences of type I, II, and III megasatellites  87  Figure 17: Construction of deletion clones from 153G8E3.35 and 39A7E3.8  88  Figure 18: Schematic representation of sequenced fragments  89  Figure 19: Localization of megasatellites by Y A C / B A C content analysis  93  Figure 20: Hybridization of total human D N A to clones containing the megasatellite  94  viii  Figure 21: Analysis of Y A C clones by hybridization of megasatellite sequence  95  Figure 22: D8S1819 and D8S1935 content analysis by hybridization  99  Figure 23: Analysis of STS content of Y A C s and B A C by hybridization  100  Figure 24: Analysis of homology between Y A C s by hybridization of B A C 87B23  101  Figure 25: Analysis of Y A C s by hybridization of B A C 87B23 T7 and SP6 ends  102  Figure 26: Analysis of cosmid clones isolated by chromosome walking from B A C 87B23  106  Figure 27: Analysis of B A C clones isolated by chromosome walking from B A C 87B23  107  Figure 28: Analysis of Y A C s and B A C by hybridization of B A C clones isolated by chromosome walking  108  Figure 29: Analysis of megasatellite content in BACs isolated by chromosome walking  109  Figure 30: FISH analysis with B A C 223B23  110  Figure 31: Examples of genotyping of patients 10-14  112  Figure 32: Hybridization of megasatellite to patient metaphase chromosomes  118  Figure 33: Most likely locations of megasatellite based on Y A C / B A C content analysis  127  Figure 34: A proposed mechanism for the generation of patient TPs' inversion duplication (8p) chromosome  132  Acknowledgements  I would like to express my gratitude to my research supervisors, Drs. Fred Dill and Stephen Wood, for their support, their guidance, and for the opportunity to learn from them. To my supervisory committee, Drs. Sylvie Langlois, Wendy Robinson, and Muriel Harris, thank you for your encouragement, insight, and for always being available. Special thanks to Mike Schertzer for his knowledge and unflagging patience. Thank you to Stanya Jurenka for sharing her clinical knowledge and for all of her kind words. Thank you to Dr. Dagmar Kalousek, the members of her laboratory, and the members of the diagnostic cytogenetic laboratories at B.C. Children's Hospital and the Royal Columbian Hospital, for their time and assistance. To my family and friends, thank you for listening and for believing in me. And finally, thank you to Colin O'Connor for always reminding me of what is most important.  1  Chapter 1: Introduction  Chapter 1: Introduction  1.1 Human Chromosome Aberrations Chromosome abnormalities are seen in approximately 1% of live-born children (Jacobs, 1990). These aberrations may be numerical, resulting from the gain or loss of an entire chromosome, or may be structural, involving rearrangement of single or multiple chromosomes. The gain or loss of a chromosome is generally believed to occur by the process of nondisjunction, the failure of chromosomes to disjoin. The mechanism leading to this nondisjunction is still under investigation; however, there may be chromosome specific mechanisms involved (Abruzzo and Hassold, 1995). Conversely, structural abnormalities must arise by intrachromosomal or interchromosomal interactions leading to changes in chromosome morphology, rather than a failure to undergo proper disjunction.  1.2 Translocations The most common de novo structural abnormality is a reciprocal translocation (Warburton 1984), a balanced rearrangement involving exchange of material between two non-homologous chromosomes. The mechanism by which reciprocal translocations arise is unclear. However, analysis of translocations ascertained through fetal wastage indicates that the location of the chromosome breakpoints appears to be random (Boue et al, 1985). In some cases of X;autosome reciprocal translocation, a few base pairs of homology are found at the sites of rearrangement, but otherwise there are no remarkable features of the genome at these locations (Bodrug et al, 1987, Giacalone and Francke, 1992). These homologous sequence motifs have  Chapter 1: Introduction  2  not been found at all sequenced X;autosome translocation breakpoints (Bodrug et al, 1991), therefore, sequence homology may not be required for chromosomal translocation. A Robertsonian translocation chromosome results from the rearrangement of two acrocentric chromosomes such that the short (p) arms are lost and the long (q) arms join to form the derivative chromosome. The mechanism generating these chromosomes is not known. Ribosomal R N A genes and several families of repetitive (satellite) D N A are present on the p arms of all acrocentric chromosomes (Choo, 1988, Choo, 1990, Trowell, 1993). Most Robertsonian translocation chromosomes are dicentric, with breakpoints in the p arm between the p-satellite and a-satellite D N A (Cheung et al., 1990, Wolff and Schwartz, 1992, Sullivan et al., 1996).). More specifically, these translocations most often involve chromosomes 14 and 21 (14q21q), and chromosomes 13 and 14 (13ql4q) (Choo et al, 1988, Therman et al, 1989) with breakpoints occurring between two chromosome 14 specific subfamilies of satellite III sequences, and distal to the satellite I sequences of chromosomes 13 and 21 (Earle et al, 1992, Kalitsis et al, 1993, Han et al, 1994, Sullivan et al., 1996). The breakpoints in other nonhomologous Robertsonian translocations are more variable (Page et al, 1996, Sullivan et al, 1996). Because there is a difference in occurrence rate and breakpoint variability between the 'common' non-homologous Robertsonian translocations and 'rarer' non-homologous Robertsonian translocations, it has been suggested that the mechanisms involved in the generation of these chromosomes differ (Page et al, 1996). The proposed mechanisms are based on interaction and recombination at homologous repetitive p arm sequences leading to the formation of Robertsonian chromosomes (Ferguson-Smith, 1967, Therman, 1980, Guichaoua et al., 1986, Choo et al., 1988). The (13ql4q) and (14q21q) chromosomes may arise by preferential  Chapter 1: Introduction  3  interaction of homologous repetitive sequences predicted to be present on chromosome 14 in opposite direction to those on chromosomes 13 and 21 (Choo et al., 1988, Therman et al, 1989, Sullivan et al., 1996). Interactions between other combinations of non-homologous acrocentric chromosomes may occur at smaller regions of homology, where recombination may be less likely to occur, and therefore, less frequently lead to Robertsonian translocations (Page et al., 1996). Although satellite D N A is implicated in the formation of at least some of these chromosomes, it is unclear whether interactions at these repetitive sequences initiate recombination or whether ribosomal R N A interactions are responsible for the initial interaction of the acrocentric chromosomes, followed by recombination at more proximal sequences. (Schmickel and Knoller, 1977, Sullivan et al., 1996).  1.3 Duplications, Deletions, Inversions Many examples of these types of rearrangements have been reported. However, certain regions of the genome appear to be more likely to undergo rearrangement. In many cases of recurrent rearrangement, such as the duplication in Charcot-Marie-Tooth type 1A (CMT1 A) (Lupski et al., 1991), sequences flanking the regions involved in the rearrangement are highly homologous, and evidence of recombination between these repetitive sequences can be shown on the aberrant chromosome. This has led to speculation that the underlying mechanism leading to recurrent rearrangements is similar, and is based on inter- and intra-chromosomal interactions at repetitive sequences. The outcome of these interactions would be dependent on the orientation of the repetitive sequences relative to one another, and on whether the interaction, and subsequent recombination, is between homologues, between sister chromatids, or within a single chromatid (figure 1) (Robinson et al, 1998).  Chapter 1: Introduction  Figure 1: Proposed mechanisms for generation of recurrent chromosome rearrangements. Modified from Robinson et al., 1998. Illustrations are schematic representations of chromosomal interactions. Triangles represent repeats. Dotted lines at regions involved in pairing represent the 3-dimensional aspect of pairing. A l l mechanisms rely on mispairing at repetitive elements, followed by recombination and resolution. Pairing may occur between direct repeats (a, b, c) or inverted repeats (d, e), and may occur between homologues (a, e), between sister chromatids (b), or within a single chromatid (c, d). Products may be duplications (a, b), deletions (a, b, c, e), inversions (d) and inversion duplications (e).  Chapter 1: Introduction  5  Misalignment of homologues or sister chromatids, at repetitive sequences present in direct orientation, followed by recombination at these sequences, would lead to duplication and deletion chromosomes (figure la, lb). However, if the misalignment occurred within a single chromatid (figure lc), a deletion chromosome, but not the reciprocal duplication chromosome, would be recovered. These mechanisms have been implicated in the formation of the CharcotMarie-Tooth type 1A (CMT1 A) and hereditary neuropathy with liability to pressure palsies (HNPP) chromosomes. C M T 1 A arises from a duplication of a 1.5 Mb region of 17pl 1.2-pl2 flanked by homologous repetitive D N A (CMT1A-REP) whereas HNPP arises from a deletion of the same region (Lupski et ai, 1991, Raeymaekers et ai, 1991). Initially, it was suggested that these chromosomes may be the reciprocal events of an interchromosomal recombination (Chance et al., 1994). However, a recent study suggests that two distinct sex-dependent mechanisms are involved in the formation of these chromosomes (Lopes et al., 1998). Paternal duplications and deletions result from interchromosomal recombination at misaligned CMT1 A-REP sequences but maternal duplications and deletions arise from intrachromosomal recombination, between sister-chromatids or within a single chromatid, at misaligned CMT1 A-REP sequences (Lopes et al., 1998). Other examples of deletion resulting from intra- or inter-chromosomal recombination at misaligned repetitive sequences include deletions of 7ql 1.23 implicated in William syndrome (Urban et a l , 1996), deletions of the steroid sulfatase gene on distal chromosome X p (Yen et al., 1990), and some deletions of the a-globin gene cluster on chromosome 16p 13.1 (Nicholls et al., 1987). Misalignment within a single chromatid, at repetitive sequences present in inverse orientation, followed by recombination at these sequences, would lead to an inversion of the region between  Chapter 1: Introduction  6  the repetitive sequences (figure Id). This mechanism has been implicated in the cause of 13% of Hunter Syndrome (HS) cases. In this instance recombination has occurred between the iduronate-2-sulphatase (IDS) gene on Xq28 (Wilson et al, 1993, Malmgren et al., 1995) and an IDS-related region (IDS-2) present 90 kb telomeric and in the inverse orientation to the IDS gene (Bondeson et al., 1995). The result is an inversion of the region causing an interruption of the IDS gene. Further, the recombination event may occur within the same highly homologous region of IDS and IDS-2 in all patients (Bondeson et al, 1995), implying that these regions may be prone to recombination. This mechanism is also proposed to account for 50% of cases of severe haemophilia A . Recombination between the factor VIII gene located at Xq28 and an upstream associated gene present in inverse orientation leads to interruption of the factor VIII gene resulting in severe haemophilia A (Lakich et al, 1993). It has been suggested that the proximity of factor VIII to the telomere may decrease steric constraints on intrachromosomal pairing (Lakich et al., 1993). It is interesting that both of these examples of inversion resulting from intrachromosomal recombination within a single chromatid occur in the same region of the genome (Xq28). Perhaps there are features of this region of the genome that predispose to this type of interaction. Proximal chromosome 15q is prone to a variety of recurrent rearrangements. Prader-Willi syndrome (PWS) and Angelman syndrome (AS) result from interstitial deletion of this region (Robinson et al, 1991, Mascari et al., 1992, Zackowski et al, 1993, Saitoh et al, 1994). This region is also prone to interstitial duplication (Clayton-Smith et al, 1993, Mutirangura et al, 1993, Abeliovich et al, 1995) and to the formation of isodicentric inversion duplication supernumerary chromosomes (discussed below) (Cheng et al, 1994, Leana-Cox et al, 1994, Robinson et al, 1997). Although the breakpoints of these rearrangements differ, both proximal  Chapter 1: Introduction  7  and distal breakpoints cluster near a series of repeated sequences (Robinson et ah, 1997, Buiting et ah, 1992, Amos-Landgraf et al, 1994). It has been suggested that these rearrangements may arise due to inter- or intra-chromosomal misalignment and recombination at these sequences (Robinson WP, personal communication). The difference in breakpoints may reflect differences in interactions between these sequences dependent on location and orientation (Robinson et al., 1998).  1.4 Inversion Duplications  Misalignment of homologues at repetitive sequences present in inverse orientation, followed by recombination at these sequences, would lead to a dicentric chromosome and an acentric fragment (figure le). Resolution of this event through breakage of the dicentric chromosome could lead to an inverted duplication deficiency chromosome and the reciprocal deletion chromosome. The existence of repetitive sequences at sites prone to this sort of rearrangement has not been shown. Inversion duplication chromosomes have been described as mirror duplication chromosomes because standard G- or R-banding techniques give a mirror image appearance to the regions involved in the duplications. This is due to a duplication of a portion of the chromosome present in the opposite orientation to the original segment so that the bands are reflected out from a center of symmetry. Inversion duplications may be interstitial, or terminal with a deletion of distal sequences, and invariably result in an unbalanced karyotype. A n example of an inversion duplication with distal deletion is shown in figure 2. The inversion duplication 8p (inv dup(8p)) chromosome in this example has a duplication from band 8pl2 to 8p23.1 and a deletion from band 8p23.1 to the telomere. The duplicated material is present at the end of the aberrant  Chapter 1: Introduction  8  chromosome in opposite orientation to the original segment, with a center of symmetry at 8p23.1. In some examples, the inversion duplication chromosome is present as a supernumerary chromosome, also known as supernumerary pseudodicentric chromosomes. Most of these chromosomes are derived from chromosomes 15 (Cheng et al., 1994, Robinson et al., 1993, Wandstrat et al., 1998) or 22 (Mears et al., 1994) and present as bisatellited marker chromosomes comprised of the p arm material plus a small segment of q arm material. These chromosomes may be asymmetrical, are usually dicentric, and are usually found in unrelated patients. The chromosome 15 supernumerary inversion duplication (inv dup(15p)) chromosomes containing the PWS/AS region can be classified into at least two categories based on similar breakpoint locations. These supernumerary inv dup(15p) chromosomes arise by an interchromosomal maternal event (Wandstrat et al., 1998), that must be followed by nondisjunction (Schrek et al, 1977). It has been suggested that variable interactions at several lowcopy repeat sequences present in this region could lead to the formation of different classes of inv dup(15p) chromosomes (Wandstrat et al., 1998). Alternatively, U-type recombination between homologues (Van Dyke, 1988) may be involved. Inversion duplication chromosomes are rare in live-born individuals. Van Dyke (1988) observed that of the 20 inversion duplication cases, excluding patients with supernumerary chromosomes, found in the literature at that time, there was an over-representation of those involving 4q, 9p, and 8p. The number of reported cases of inversion duplication 8p has increased to over 60 (table 1), however, this same increase in number has not been reported for 4q, 9p, or any other chromosome location. The center of symmetry of the inversion duplication 8p chromosomes is similar in all patients. In the majority of cases the center of symmetry is  Chapter 1: Introduction  9  within 8p22 or 8p23 (figure 2), and the duplication extends to at least band 8p21.2. This identity of center of symmetry in unrelated cases is also seen in a few rarer cases of inversion duplication, including X p inversion duplications (Tuck-Muller et al., 1993, Telvi et al, 1996), 7q inversion duplications (Stetten et al., 1997) and chromosome 14 inversion duplications (North et al., 1995). For these rarer examples the number of cases is extremely small, usually only two, and it is therefore difficult to determine whether this is an ascertainment bias based on viability of the rearrangement or represents some underlying feature of the genome at these locations. However, at least for chromosome 8p, there appears to be a propensity to form this type of rearrangement. This led to investigation of whether other features of the inv dup(8p) chromosomes were shared, and whether this may reflect a common mechanism of origin.  1.5 Inversion Duplication 8p (inv dup(8p))  The inversion duplication 8p chromosome has been described above. The majority of these chromosomes have a duplication involving at least bands 8pl2 to 8p23, and a center of symmetry at 8p23.1 or 8p22 (figure 2). In 1987, Dill et al. reported the deletion of bands distal to the center of symmetry, for the first time describing these chromosomes more correctly as inversion duplication deficiency chromosomes. Dicentric inv dup(8p) chromosomes have also been reported (Floridia et al., 1996). These chromosomes have a single q arm, an inverted duplication of the region from the centromere to band 8p23.1, a deletion of distal bands, and centromere associated sequences, such as alphoid D N A , present near the telomere of the p arm. These dicentric chromosomes represent the largest duplication ascertained for inv dup(8p), the smallest being from band 8p21.2 to band 8p23.1. A l l other inv dup(8p) chromosomes vary in the amount of duplication, within this range.  Chapter 1: Introduction  10  12  21.1 21.2 21.3 23.3 23.2  22  23.1  23.1  22  21.3 21.2 21.1 12  11.22  5  22  21.3 21.2 21.1  11.2  11.2  11.1 11.1 11.21  11.1 11.1 11.21  1.22 •  11.23  11.23  12  12  13  13  21.1  21.1  21.2  21.2  21.3  22.1 22.2 22.3  23  24.1 24.2 24.3  s n normal 8  21.3  22.1 22.2 22.3  23  24.1 24.2 24.3  inv dup(8p)  Figure 2: Ideogram of normal chromosome 8 and inversion duplication (8p) chromosome.  Chapter 1: Introduction  11  Table 1: A comprehensive list of reported cases of inversion duplication (8p) patients. The reported region of duplication is listed, with the center of symmetry being the second band location. Report of a distal deletion is indicated by 'yes', report of an undetected distal deletion is indicated by 'no'. The origin of the chromosome is indicated (origin) as well as whether the parental D N A from which the chromosome was derived is known (DNA origin). d.n. = de novo, rec. = recombinant chromosome, n.k. = not known, mat. = maternal origin Reference (et al.) & Year  Case  Weleber 76 Taylor 77 Taylor 77 Rethore 77 Rethore 77 Hongell 78 Mattei 80 Mattei 80 Poloni 81 Poloni 81 Poloni 81 Jensen 82 Jensen 82 Fryns 85 /Kleczkowska 87 Kleczkowska 87 Kleczkowska 87 Dill 87/Henderson 92 Nevin 90 Gorinati 91/Minelli 93/ Floridia 96 Feldman 93 Feldman 93 Feldman 93 Feldman 93 Feldman 93 Feldman 93 Feldman 93 Feldman 93 Feldman 93 Feldman 93 Minelli 93/Floridia 96  1 1 2 1 2 1 1 2 1 2 3 1 2 1/3 1 2 . G.S. 1 1/3/14 1 2 3 4 5 6 7 8 9 10 1/10  Duplicated Region 8pll.2^p23.1 8pl2-+8p23 8p21^8p23 8pll^8p22 8pll->8p22 8pl2^8p23? 8pl2^8p23.2 8pll-»8p23.2 8pll^8p23.2 8pll^8p23.1 8pll^8p23.1 8p21.2^8p23.1 8p21.1-+8p23.3 8p21.1^8p22 8p21.1^8p22 8p21.1^8p22 8pl2^8p23.1 8pl2-*8p23.1 8p21.1^8p23.1 8p21^8p23 8p21^8p23 8pll.23^8p23.1 8pll.2^8p23 8p21.1^8p23.1 8pll.2^8p23.1 8pll.2^8p23.1 8pl2^8p23 8pl2^8p23 8pl2^8p23.1 8pl2^8p22/ 8pll.2-^p23.1  Origin d.n. d.n. d.n. d.n. n.k. d.n.. n.k. n.k. n.k. n.k. n.k. d.n. d.n. d.n. d.n. d.n. d.n. d.n. d.n. d.n. rec. d.n. d.n. d.n. d.n. d.n. d.n. d.n. d.n. d.n.  Distal Deletion? n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. yes n.k. yes  DNA Origin n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k.  n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k.  n.k. mat. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. mat.  yes  Chapter 1: Introduction  Table 1 continued Reference (et al.) & Year  Case  Minelli 93/Floridia 96  2/2  Minelli 93/Floridia 96  4/7  Minelli 93/Floridia 96  5/8  Minelli 93/Floridia 96  6/9  Minelli 93/Floridia 96  7/16  Minelli 93/Floridia 96  8/5  Minelli 93/Floridia 96  9/1  Barber 94 Mitchell 94 Engelen 94/ de Die-Smulders 95 Engelen 94/ de Die-Smulders 95 Engelen 94/ de Die-Smulders 95 Engelen 94/ de Die-Smulders 95 Redha 94 Hoo 95 de Die-Smulders 95 de Die-Smulders 95 de Die-Smulders 95 Guo 95 Guo 95 Guo 95 Guo 95 Guo 95 Guo 95 Guo 95 Floridia 96 Floridia 96 Floridia 96 Floridia 96 Floridia 96  1 1 1/1 2/2 3/6 4/7 1 3 3 4 5 1 2 3 4 5 6 7 3 4 6 11 12 a  Duplicated Region 8pl2^8p22/ 8cen->8p23.1 8pl2^8p22/ 8pll.2-»8p23.1 8pl2^8p22/ 8pll.2-»8p23.1 8pl2^8p22/ 8pll.2->8p23.1 8pl2^8p22 /8p21^8p23.1 8pl2-+8p22/ 8cen^8p23.1 8pl2-*8p22/ 8cen^8p23.1 8pll.23-»8p23.1 8pl2^8p23.1 8pl2^8p23.1 8p21.1->8p22 8pl2^8p23.1 8pl2^8p23.1 8p21-*8pter 8p21.2->8p23.2 8p21.1->8p22 8pll.2->8p23.1 8pl2-+8p23.1 8pl2-*8p23.1 8pl2^8p23 8pl2^8p23 8pll.2->8p23.2 8p21.3->8p23.3 8pll.2->8p23.1 8pll.2-»8p23.1 8cen^8p23.1 8cen^8p23.1 8cen->8p23.1 8pll.2->8p23.1 8pll.2->8p23.1  Origin d.n.  Distal Deletion? yes  DNA Origin mat.  d.n.  yes  mat.  d.n.  yes  mat.  d.n.  yes  mat.  d.n.  yes  mat.  d.n.  yes  mat.  d.n.  yes  mat.  d.n. d.n. d.n. d.n. n.k. n.k. n.k. d.n. d.n. d.n. d.n. d.n. d.n. n.k. d.n. n.k. n.k. d.n. d.n. d.n. d.n. d.n. d.n.  yes yes yes yes yes yes n.k. n.k. n.k. n.k. n.k. yes yes yes n.k. n.k. n.k. n.k. yes yes yes yes yes  mat. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. n.k. mat. mat. mat. mat. mat.  Chapter 1: Introduction  Table 1 continued Reference (et al.) & Year  Case  Duplicated Origin Distal DNA Region Deletion? Origin 13 d.n. Floridia 96 yes mat. 8pll.2->8p23.1 Floridia 96 15 8p21^8p23.1 d.n. yes mat. Nelson, this thesis TP d.n. yes 8pll.2-»p23.1 mat. Nelson, this thesis DC 8pl2^p23.?2 d.n. no n.k. Nelson, this thesis SW d.n. 8p21.3->p23.?2 no n.k. Nelson, this thesis MD 8p+ d.n. yes mat. Patient was diagnosed as 47,XX, inv dup(8)(p23.3p21.3)?, +r(?). The abnorma banding pattern on the derivative 8p is similar to other cases (Guo et al., 1995) Patient carries an inv dup(8p) chromosome with satellites, see section 3.1.2 b  a  b  Chapter 1: Introduction  14  The phenotype of patients who have an inv dup(8p) chromosome is variable. This has been attributed to three potential causes: variation in the extent of duplicated material, variation in genetic background, and variation in the extent of deleted material. However, deletion of distal bands may not have a significant role in the phenotype of these patients as deletions of terminal (8p) have only a mild phenotypic effect (Fryns et al., 1989). In an attempt to define common features of the phenotype for inv dup(8p) patients, Kleczkowska et al. (1987) studied patients with inverted duplications of bands 8p21.1 —»8p22. Some common features seen in these patients include severe mental retardation, skeletal problems, difficulties sucking and feeding as infants, and a characteristic facies with features including: quadrangular head, prominent forehead, malformed and posteriorly rotated ears, thin lips with everted lower lip, broad nasal bridge with prominent nose and anteverted nostrils.  1.5.1 Inversion Duplication 8p is Sporadic To date, none of the inv dup(8p) chromosomes has been inherited from a parent carrying the same aberrant chromosome. This is not surprising as it is rare for individuals with severe mental retardation to reproduce. Feldman et al. (1993) reported an inv dup(8p) patient whose mother carried a balanced paracentric inversion. The child's inv dup chromosome is believed to have arisen as a result of recombination in a meiotic inversion loop. A l l other confirmed cases of inv dup(8p), where origin has been investigated, have been de novo (table 1). In 1994, Dhooge et al. reported a case that, based on cytogenetic analysis, may be an inversion duplication or a direct duplication, segregating in a family. The mother and the two children carrying the aberrant chromosome have mild mental retardation. Based on molecular and fluorescence in situ  Chapter 1: Introduction  15  hybridization (FISH) evidence presented in this thesis, this rearrangement appears to be a direct duplication.  1.5.2 Center of Symmetry of Inv Dup(8p) In order to define inversion duplication (8p) as a recurrent rearrangement, common features of the inv dup(8p) chromosome must be seen in unrelated patients. Although the amount of duplication varies from patient to patient, the cytogenetic center of symmetry is reported to be in band 8p22 or band 8p23.1 in the majority of patients (see table 1), with rare cases in band 8p23.2 or 8p23.3 (Mattei et al, 1980, Poloni et al, 1981, Jensen et al, 1982, Redha et al, 1994, Hoo et al, 1995, Guo et al, 1995). However, a number of reports of inv dup(8p) chromosomes with a center of symmetry at 8p22 were made prior to either the routine use of high resolution chromosome banding techniques (Rethore et al, 1977), or before it was recognized that there is a deletion of D N A distal to the center of symmetry (Kleczkowska et al, 1987, Fryns et al, 1985). In some cases, re-examination of inv dup(8p) chromosomes has led to redefinition of the center of symmetry to band 8p23.1 (Gorinati et al, 1991 and Minelli et al, 1993 re-examined in Floridia et al, 1996). Therefore, it can be argued that in some, i f not all, instances where the center of symmetry is reported in band 8p22, the correct interpretation should be 8p23.1. Thus, with rare exceptions, the majority of inv dup(8p) chromosomes have a common center,of symmetry in band 8p23.1.  1.5.3 Molecular Features of Inversion Duplication 8p As well as the common phenotypic features and similarity in the cytogenetic location of the center of symmetry in unrelated inv dup(8p) patients, there are a number of common molecular  Chapter 1: Introduction  16  features of these chromosomes. A deletion of distal material on the inv dup(8p) chromosome was first shown at the molecular level by Dill et al. (1987) by densitometric Southern blot analysis, using a probe for the D8S7 locus mapping to 8p23 (Wood et al, 1986). The patient D N A was monosomic for the marker, which indicated transmission of only one parental allele at the D8S7 locus, and lead to the conclusion that the aberrant inv dup(8p) chromosome must be a duplication-deficiency chromosome. This duplication-deficiency was later confirmed by FISH analysis (Henderson et al, 1992). Approximately half of the reported cases have subsequently been tested for deletion of distal sequences, using either molecular techniques or FISH analysis (Barber et al, 1994, Mitchell et al, 1994, de Die-Smulders et al, 1995, Guo et al, 1995, Floridia et al, 1996). In all informative cases, where the center of symmetry is in 8p23.1 or 8p22, terminal deletions have been found (see table 1). The most proximal marker deleted from inv dup(8p) chromosomes is D8S349, located within band 8p23.1 (see figure 3) (Floridia et al, 1996). The inv dup(8p) chromosome is derived from both maternal chromosomes in 17/17 cases investigated (Minelli et al, 1993, Feldman et al, 1993, Floridia et al, 1996). Therefore, these chromosomes must arise by a maternal interchromosomal recombination event. Using a combination of genotyping and/or FISH analysis, Floridia et al. (1996) defined a region located at the center of symmetry called the single copy region. In all informative cases, this region is found in band 8p23.1, delimited at the distal end by D8S349 (deleted from the aberrant chromosome) and at the proximal end by D8S552 (duplicated on the aberrant chromosome) (see figure 3 for locations). The sequence tag site (STS) markers located within this region are present in single copy on the aberrant chromosome, and are flanked by regions of duplication. Therefore, molecular evidence suggests that these chromosomes are asymmetrical.  Chapter I: Introduction  17  In summary, the majority of informative cases of individuals with an inv dup(8p), the aberrant chromosome has a center of symmetry located at 8p23.1, a deletion of genetic material distal to the center of symmetry, a region of single copy located at the center of symmetry, and is derived from a maternal interchromosomal event. The finding of shared features of the aberrant chromosome in unrelated patients suggests that these features may reflect a common mechanism of origin.  1.5.4 Mechanism of Formation of inv dup(8p)  Two mechanisms proposed to account for the common features of the inv dup(8p) chromosomes are presented here: The first mechanism (Weleber et al., 1976) is the end-to-end fusion of two broken chromosomes 8 forming a dicentric chromosome which would break at anaphase to create the inversion duplication chromosome and the reciprocal deletion chromosome. Based on the observation in inv dup(8p) chromosomes of a distal deletion, common cytogenetic 'center of symmetry', and region of single copy, a refinement of this mechanism has been made (Dill et al, 1987, Floridia et al., 1996). A mechanism involving inverted repetitive sequences, analogous to that proposed for other recurrent structural rearrangements, could apply to inv dup(8p). Misalignment of homologues, at these inverted sequences, followed by recombination, would result in a dicentric chromosome (figure 4). As the chromosomes proceed through normal anaphase, the centromeres of the dicentric chromosome may be pulled to opposite poles, causing the dicentric to break at a random point between, or at, the centromeres. It is widely accepted that telomere repair acts to heal broken ends (Morin 1991, Flint et al, 1994) and therefore, could repair the broken ends and stabilize the aberrant products. Of the resulting daughter cells, one  Chapter 1: Introduction  18  would contain an inversion duplication chromosome, the other, a deletion chromosome. The amount of duplication on the inversion duplication chromosome would vary, depending on the location of the break. Furthermore, if the paired repeat sequences are located relatively far apart, then a region of single copy, flanked by the duplicated regions, would be detected on the aberrant chromosome (Floridia et ah, 1996). This mechanism is particularly attractive as it requires only the naturally occurring secondary consequences of an initial aberrant pairing event, and is directly testable at the molecular level. The second involves a U-type exchange within a paracentric inversion loop during meiosis (Mitchell et al, 1994) (figure 5). In a paracentric inversion with a distal break point, a U-type exchange within the meiotic inversion loop would leave the telomeres intact while deleting subtelomeric sequences. This mechanism requires that a parent carry the inversion and could lead to familial cases of inv dup(8p). When both parents are karyotypically normal, however, only a premeiotic event creating a new inversion would allow this type of mechanism to occur. Since this would require two rare events, inversion and U-type exchange, and since this exchange would not lead to a region of single copy at the center of symmetry of the aberrant chromosome, it is unlikely that this mechanism would account for the majority of cases reported. A l l parents of inv dup(8p) patients, with the exception of the mother of the Feldman et al. (1993) case, are karyotypically normal.  Chapter 1: Introduction  19  TJ  O  -s o  -2o s  a o -  a  ON —^  SH C  D  VO  jg  TJ  H-J  oo c oo -a Q T3 s .a ON Ov  T3  OO  HO  00  Q  CJ  >»cS a  VO  TJ G  00  ca  oo Q  oo Q T3  C4H  o  v o  cn o o o o Q  Q  J  0 0 0 0 Q  HOO o o o o  O V O m c s o o o o o o o o Q Q  o o o o OOOO Q Q  OOCN  t — " I t - CN - H OO—• o o OO OOOO 0 0 OOQOOQ  OV — 0 0  v o o  o o o o  o o o o  Q Q  Q  I CN  O v  c s  o o 0 0  CM OO 0 0  o o v > CSIOVO m i n c N o o o o o o o o o o o o  Q  Q  Q Q Q  ^/c^^^^^t _  VO >n o o 0 0  CN  r-  (N OO 0 0 Q  Q  Q  o o  F»  v o o o 0 0  o o 0 0  CN - H CN  c o  3 OO  0 0  o CN OO o o  yr,  ovvo  — O500O0 OO 0 0 0 0 o o 0 0 Q Q Q Q  c c3  Chapter 1: Introduction  20  d  o  C D  ca .2 > Cl ^ 8, cu CL ID CU  a o  t H  <U 0 0 t H  g C D  CU  "t2  ^  <U  O  0 0  O  c3 o -° o cu C U  d  43  I  oo  ca Od o _0 ca CD t H  ce >. CO  CU  O O s  3  T3 o  0 0s  d o B .2 g cu o §  s  cu °  —  .a §f -d d OH  5" cu o •5 > d CU o a, cu a o cu c o cu > d CU t H  t H  t H  s i o cu d > o OX) +-'  • i-H •t!  -a  ca cu ca .3 -*-»  1-s 0 3  CU  U  -d  t  H  OH  o s  t H  d  0 0  °  cu t: « o ca o -a o ca d ca d o £ fc! U ± 3 ca .2 < d £ .a ca cu o cu -d HH.  r H  d y  O  ca cu cu cn 2 -3 • ! = CU O CH O c u cu cu i oo cn O cu o o d cu C8  d  d o  t H  CU  fl  > d .a  &  •-H  CL,  =5  c  «*  0 0  d  CT" 53 O cu cu  C+H  0 0  t -  21  Chapter 1: Introduction  c o a o  CO  OO SH  <D  B > o o ID a CJ < « g k ts CJ -CJ o o 8) s 6 29 > a ^ -S -a CJ  OO  -  a o  u  a,  3  •a >  .3  ca m  a  CJ — ,  cd ca  g  ° o gC oo « ^ 'in cd  —i  .  o a  o a ca  SH  CJ  O  3 5 = 001 §  CJ  o •r ^ ca C8 § o o CJ ca a 3 o TJ 1  O  S3 O  a '« i in o CJ o > .g a C*H  O  o o ^ ^  Irs  W) o fl  °  SH  OH  -CJ r e CJ  CJ  S-i ^ co •<  CJ  »  cS H  "s * o 5• ca  cj ° cj a o a _ ca o T3 CJ  SH  CJ  co  111  Chapter 1: Introduction  22  1.6 Project Objectives The objective of this thesis was to investigate inversion duplication (8p) chromosomes at the molecular level to determine whether a mechanism mediated by inverted repeats could apply. To accomplish this objective, molecular genetic analysis was carried out on inversion duplication (8p) patients, with the intent of isolating proximal and distal 'breakpoint' sequences and examining these sequences for similarity. The breakpoint is defined as the point at which the event has occurred, in this case, by the transition from one maternal chromosome 8 homologue to the second maternal chromosome 8 homologue. Should a mechanism mediated by repetitive sequences apply, this transition point would be located either distal or proximal to the single copy region, depending on where the proposed recombination event occurred. Further, should repetitive sequences be found at these transition points, these sequences should be inverted with respect to one another. Genotyping of inv dup(8p) patients for genetically and physically mapped STS markers allowed the distal breakpoint to be localized to a region flanked by STS markers D8S349 and D8S503 (figure 3). The proximal breakpoint location is defined by STS marker D8S265 and D8S552 (Floridia et al, 1996). As a first step towards isolating the breakpoint sequences, finescale physical and genetic mapping of these regions was required. Mapping efforts were focused on the distal region. A novel repetitive element identified by Kogi et al, 1997 and Gondo et al, 1996 on chromosome 4 was isolated and mapped to the regions of 8p containing the proximal and distal breakpoints. Therefore, this element is a candidate for the proposed repetitive element involved in the formation of inv dup(8p).  Chapter 2: Materials and Methods  23  Chapter 2: Materials and Methods  2.1 Polymerase Chain Reaction  The annealing temperature of each primer of the sequence tag site (STS) primer pair was calculated by the equation (4(C+G) + 2(A+T) - 4). The annealing step of PCR was carried out at the lowest calculated annealing temperature of the pair, allowing specific hybridization of both primers. Sequence and annealing temperature of primer pairs used in this thesis are listed in table 2. 100 ng of sample D N A was amplified in a 25 pi reaction volume containing 2.5 m M M g C l , 2  0.2mM of each nucleotide (dATP, dTTP, dCTP, dGTP), 50 m M Tris-Cl pH 8.3, 0.05% Nonidet40, 0.05% Tween-20, 0.4 m M each primer pair and 0.5U Taq Polymerase. A positive control of 100 ng total human D N A and a negative control of Milli Q filtered distilled water were included with each set of reactions. Samples were overlaid with 25 pi of paraffin oil to prevent evaporation. A n automated thermal cycler was used to amplify the D N A . Forty cycles of denaturation (1 min at 95° C), annealing (30 seconds at the appropriate annealing temperature) and elongation (1 min at 72° C) were carried out, followed by a 10 minute elongation step at 72° C. 5 ul of stop buffer was added prior to gel electrophoresis. Stop Buffer 0.25% xylene cyanol 0.25% bromophenol blue 40% sucrose (w/v) 60 m M E D T A  21  CN as Os CN  m  ON  Os  CN as as  m  o  \Os as  as  as as  T3  -o  Cu KS  (D  1(2 .  o  ca «  CN  ca a.  lo  ea Xl c eo  v  oo  o  ©  a. io  E  X  i-i  e> e>  O  o ea x ea a. ea  a  CD  |X u  '"Si o  IO  \Os \f~  oo  \ ~>  ON  3  eB  IX) c  ,o  on  3  CD  a  CL»  X  as as 3  3  T3  X  XJ •3CJ\  -a  o  CD  |XJ  o  o o  i3!  IO  oo oo  CD  CD  o Ko  loo  ^1  o c  o  o  r  u c 3 -Q •O o o k. N w  1  oo CN  CN  OO  CN CN  i o  CN  in  O O o  s  \as oo  OO  CN  Ko  loo  ho  CN  Ko  a. -a  ts  O  o < a < CJ u < o < u O u <H <a a u << CJ u < u < < a H <a a a < o  o  cu  s  •S  cu  CD  s-  C/3  <N  O  C/3 CN  5  H  H  cd  H  o CN  II  <  00  Q  H H H U <  u <  O a o a <u < << a aa H O  < < <  < < u u <  < u a <u < < H H  <  < a < uc  u u a u H o <U < <H  <  a < u o H  < u a a < o o < H H O u a  < O  H < O H O < O  H H U H U U <  as  m o >/->  ir-i vo  <  00 oo  Q  H  ao OO Q  H O O U H H H O  a < < u u  u a a H H  H U U H H H < < H O H H <  oo  Q  a a < < o H H H H H  au < o a <a a  CN C/3  H H  oo  Q  < <  < <  u H  a <a o HH a <H oa <o < << <o u u u o u a< < o u 00 00  Q  <a u < H u a H H a u < U a a a uo u o H < Ha a H H u u < < CJ o u  o< H cj H a u UH H <  < <<  <a  < a < <a <a a a o <u H a H a < ud H H U  oo  Q  u  < O  a H a < a H H U  u < ; u u a u f— CN  vo  <  ro on  < <  U  C/3  H  00  u a x 00  no  CN VO  product size (bp) VO VO CN  SHGC-1030 AGACCTTCCAGCGATAACCTTTC  Gyapay et al, 1994, Weissenbach et al, 1992 Gerken, 1993  0.82  0.73  no  no  yes  yes  yes  CN VO CN VO vo in © m VO O vo  VO © o VO CO CN in  SHGC4-1739 CCTGTGCACTCCTTCCTGAATTG  TGAGAGGTCTGAGTGACATCCG CCAGGTGAGTTTATCAATTCCTGAG  GGCTCCCTGTTCTTTATCAGGTTG ACTTTTCTCTGGCTATTATGGACTC  D8S439  CGTGCTGAGAATGAGACC  TCACTGAGGGACTTGGC  TCTCTGGCCAAATCAGCC  CTCATTAGGGAAGAAAATACGCTG  D8S277  D8S1819  rsAVA13  SHGC4-1656 CAGAAGCATGAGCAATGCAGATG GTGTGTGAAAACTACAGTGTGATG  CCATGAAAAGACAAGGGAAAGAAAC  SHGC-1436 GTAAGGAACAGGGCGGAATTGAG  115-133  148-180  207-223  no  VO in  AGCCCTTCTGTAAGTCACCCCC  no  CN VO  Weissenbach, unpublished (1996)  Dib etal, 1996  Vocero-Akbani et al, 1996  Myers, unpublished (1995)  Goold et al, unpublished (1994)  Goold et al, unpublished (1994)  Goold et al, unpublished (1994)  no  Goold et al, unpublished (1994)  Myers, unpublished (1994)  References  CN VO  SHGC4-1596 AGTGTTCATGGACTCCTGATGTG AGTTCAACGTCAGAAAGTCGAAG  Max. Hetero.  in ©  no  SHGC4-1135 CACGGCATTTCAGACACTTTTGAC TGGAGACAAACACCACAATGATAG  TCAAACTTCAGTGCATGTTTCTACC  polymorphic  r" o  primer sequences (5'-3')  Locus Symbol  vs  CN  on ON TJ  ON  •a  TJ CU X!  X!  3 C O  CO ON ON  CO C3N ON  TJ  TJ XI  Xi NO ON Os  3  3  3  C O  , •§  < 1 >  '"Si  1/5  o .  TJ  I'  C O  '"Si  >jO ON ON  to  3  3  o  ON ON  X)  |xj  C  ON ON  |X>  o  o  5  CN  CN  00  o  a. u o  o c  o c  o c  OJ ,  CO  o  P" o 1  u c 3 Xi TS ^O <U  oo  «o  I/O  00 CN  ON  co  U U O <  u <u <a <o H < HH < <H < HH  -S3 O  a <  u a cu  < <  s a*  u < a u H < t— u U o  <u a CN  o o  CU  CN  3  vi O 3 X S u  °  "  H H O H O U  S  CZ)  IT) CO ON 00 OO  Q  < u < HH < Oa o a< o H U <  o <  u < < H u o H<-> H  H H  ua H a U H H a H o H <  CN  >n CN 00  a a <  < o< H < H <  a < a o H a < a a u H a H a a u H a a < < :u u aa < H  o NO 00 0O  Q  wo uo  uo uo  I CO UO  N"  NO CN CN  CN NO  < <  u  < < <  < H < O O H <  a < u H O <  a U< H O  tu o u H H <  H H < H O <  CO ON ON  u H H o < < u ao a H H aH a H o a  00 00  00 00  <  u u < u H < H U < u au a u a< a  \0\ ro  < U < a < o a H < H a < OH < Ou < H < a a < c a oo < O H H O u U u U  ON  wo  ON  H  o o H H O H < H U H U H H H O  a a o H < a < a  a H a < < o a < a H H <  < <  <  u u a <  >n  NO  o  UO  o  < a < < H  H U  uo  00 00  Q  CN 00 00  Q  CO 00 00  a  2>  o o o  o  E o  CD  (5 .  ON  4•—» CQ  CD JD  00  16  ca  c<3  |5  ON ON  <D  o  c i-  ca  ~5  o o  a. P- o 1  u  3  -5  ca  a  3 XI •a ^ o <u  U  3  ca  "5  N  a. -S  u o o (< <  ts c  5  a u u <  u < u H u < a  <  <u  3  H  a u H < Ot— u  er <u T3 <D  CU  E  a  CY>  < a < a o a <  O H H  ts o  •S  a < o u < a  o o CN CU  <  CJ  <  a < < H H  a < t<  H U  <c  o < o H a o < H H H  < vo  CA  O  o -J  6  3 £>  CL.  00  <c  a a a H CJ  a H a < < < CJ o u H u u a w —i <  a o a H u u a < : u u H  O <  u O H O <  u u <  Chapter 2: Materials and Methods  28  2.2 Polymorphic STS Genotyping  2.2.1 Labeling of STS Primer The annealing temperatures of the primer pairs were calculated, and the primer with the lower annealing temperature of the pair was chosen for y-end labeling. This decreases visualization of non-specific PCR products. Labeling was carried out in 10 pl reaction containing 50 pmoles of D N A primer, I X Kinase Forward Reaction Buffer (Gibco/BRL), 1U T4 Kinase (Gibco/BRL) and 25 pCi y-dATP. Labeling was carried out at 37° C for 60 minutes followed by 5 minutes at 95° C to heat inactivate the enzyme. The primer was stored at -20° C.  2.2.2 Polymerase Chain Reaction PCR was carried out on 100 ng of sample D N A , prepared from whole blood, using standard reactions (see section 2.1) with the exception that only 30 cycles were completed and that 0.1-0.2 pl end-labeled primer was added to each reaction tube. Upon completion of PCR, samples were frozen at -20° C. Just prior to polyacrylamide gel electrophoresis (PAGE), P A G E stop buffer was thawed and 9 pl was added to each frozen sample. Samples were boiled for 5 minutes and placed on ice. The genotype of individual 134702 (Dausset et al, 1990) was included in each set of reactions as a positive control. P A G E Stop Buffer l O m M NaOH 95% formamide 0.05% bromophenol blue 0.05% xylene cyanol  Chapter 2: Materials and Methods  29  2.3 Polvacrylamide Gel Electrophoresis (PAGE)  2.3.1 Preparation of Gel Apparatus and Casting Tray The reservoir plate, front plate and two 0.4mm spacers were cleaned with 95% ethanol and low-lint Kimwipes. The reservoir plate was silanized and placed silene side up on a level surface. The spacers were placed along either side of the reservoir plate and covered with the front plate. The plates were held together using sealers, allowing approximately 1 cm of glass to protrude from the bottom of the sealers. The P A G E casting tray was lined with a casting tray sponge and a piece of Whatmann paper cut to the dimensions of the sponge.  2.3.2 Preparation of the Polyacrylamide Gel 5.2 g of urea was dissolved in 27 ml d H 0 , 7.2 ml 5X T B E and 6 ml Long Ranger (Mandel 2  Scientific Corp.). To 13 ml of this solution was added 56 pi T E M E D (Promega) and 200 pi 25% (w/v) ammonium persulfate (APS), swirled to mix, and poured quickly into P A G E casting tray. A seal was formed by placing the bottom of the P A G E gel apparatus onto the seal solution and allowing it to move between the glass plates by capillary action. To the remaining solution was added 26 pi T E M E D and 106 pi of 25% APS, swirled to mix, and drawn up into a 25 ml plastic pipet using an automated pipeter. The gel solution was carefully poured from the pipet between the two glass plates until filled. The comb was placed flat side down between the plates and the remaining gel solution was poured over the comb to seal the area from air. The gel was allowed to polymerize for 1 hour, the casting tray was removed, and the gel placed in the P A G E running apparatus containing 0.6X TBE. The comb was removed, the reservoir filled with 0.6X TBE, the  Chapter 2: Materials and Methods  30  electrodes placed, and the gel allowed to run at 3000 V , 150 mA, 55 W for 20-30 minutes to prewarm the apparatus.  2.3.3 PAGE Electrophoresis and Autoradiography The space made for the comb was cleaned with 0.6X T B E and all air bubbles were removed. The comb was placed and the samples were loaded. The gel was subjected to electrophoresis for the appropriate length of time (for allele sizes of approximately 130 bp this was 5 minutes after the first marker dye had run off the gel). The gel was removed from the gel apparatus onto Whatmann paper, dried in a gel dryer, placed on film for the appropriate length of time to obtain clear signal, and developed.  2.4 Cosmid DNA Preparation and Isolation  2.4.1 Human Chromosome 8 Cosmid Library A human chromosome 8 specific cosmid library, LA08NC01, representing 4 genome equivalents, was constructed at Los Alamos National Laboratory, New Mexico (Wood et ah, 1992). Chromosome 8 was isolated from a human x hamster hybrid cell line (UV20HL21-27), containing human chromosomes 4, 8, and 21 (Fuscoe et al., 1986), by fluorescence-activated flow sorting (Deaven et al, 1986). The chromosome 8 D N A was extracted, subjected to partial digestion by Sau3Al, dephosphorylated and ligated into BamHI dephosphorylated sCos-1 vector (Evans et al., 1989).  The D N A was transfected into E.coli D H a M C R cells and plated onto agar  plates containing kanamycin (50 pg/ml). Kanamycin resistant E.coli colonies were transferred into 96-well microtitre plates containing L broth and grown overnight at 37° C. Glycerol was added to a final concentration of 40% (Rose et al., 1990). Four genome equivalents, represented  Chapter 2: Materials and Methods  31  by 20,160 colonies, were arranged into 210 microtitre plates and stored at -70° C. Hybridization with both human and Chinese hamster D N A showed 85% human specificity. Nine loci mapping to different areas of chromosome 8 were used to screen the library. In all cases isolation of 1-4 cosmids was achieved (Wood et al., 1992). Agar Plates 4 g bactopeptone 2 g yeast extract 2 g NaCl 0.4 g Dextrose 4.8 g agar 400 ml d H 0 2  2.4.2 Preparation of High Density Chromosome 8 Cosmid Library Filters High density cosmid library filters were prepared by Ashley Howard using the robotic Biomek 1000 (Beckman Instruments Inc.). Hybond N+ membranes (Amersham) overlaid on agar plates containing kanamycin (50 pg/ml) were inoculated, by staggering, so that the clones from sixteen 96 well plates could be placed on a single filter (1,536 clones). Inoculated filters were grown overnight at 37° C. The D N A was lysed, denatured, and fixed to the membrane by the protocol of Grunstein and Hogness (1975). This high density procedure arrays the entire library on thirteen filters labeled A through M .  2.4.3 Isolation of Cosmid DNA Cosmids were streaked, from glycerol stocks stored at -70° C, onto agar plates containing kanamycin (50pg/ml) and grown overnight at 37° C. Individual colonies were inoculated into 5 ml L broth and grown overnight, with shaking, at 37° C. Cosmid D N A was prepared by the methods of Birnboim and Doly (1979) and Ish-Horowicz and Burke (1981), with some  Chapter 2: Materials and Methods  32  modifications. 1.5 ml broth was centrifuged in 1.5 ml eppendorf tubes at 13,000 rpm for 30 seconds to pellet the E. coli cells. The supernatant was completely removed, the pellet carefully resuspended in 100 pl ice cold Solution I and left for 3-5 minutes at room temperature. 200 pl of freshly made Solution II was added, mixed by inversion, and the tube placed on ice for 3-5 minutes. 150 pl of ice cold Solution III was added, mixed by inversion, and the tube was replaced on ice for 3-5 minutes. Samples were centrifuged for 5 minutes at 13,000 rpm, and 400 pl of solution was removed to a clean 1.5 ml eppendorf tube to which 800 pl of ice cold 95% ethanol or 400 pl room temperature isopropanol was added to precipitate the D N A . Samples were centrifuged at 13,000 rpm for 7-15 minutes, the alcohol removed and the pellet washed with 1ml ice cold 70% ethanol. Pellets were air dried and resuspended in 50 pl of TE(pH8.0)/RNase(25 pg/100 pl) followed by incubation at 37° C for 30 minutes to digest any R N A contamination. 1 pl of sample in 5 pl of stop buffer was run into 0.8% agarose, by electrophoresis, to determine the D N A concentration. L Broth 4 g bactopeptone (tryptone) 2 g yeast extract 2 g NaCl 0.4 g D-glucose 400 ml d H 0 2  Solution I 50mM glucose 25 m M Tris Cl (pH 8.0) 10 m M E D T A (pH 8.0) dH 0 2  Solution II 0.2 N NaOH 1% SDS  Chapter 2: Materials and Methods  33  dH 0 2  Solution III 3 M potassium acetate 5 M acetic acid dH 0 2  IX TE l O m M T r i s - C l (pH 8.0) 1 m M EDTA  2.5 Bacterial Artificial Chromosome (BAC) DNA Isolation and Preparation  2.5.1 Human Chromosome 8p Enriched BAC library • A human chromosome 8p enriched B A C library was prepared by Michael Schertzer. InterAlu PCR allows the amplification of D N A between two Alu repeats separated by a distance compatible with PCR amplification (Nelson et al., 1989). The cell line 1HL12 (Wagner et al., 1991 ) retains only the short arm of human chromosome 8. InXox-Alu PCR using the ALE-1 and ALE-3 primers (table 2) (Cole et al, 1991) was performed on this cell line using standard PCR conditions (see section 2.1). The PCR products were used as hybridization probes (see sections 2.12, 2.13) (Sambrook et al., 1989) to high density filters of the Research Genetics B A C library consisting of 27,698 clones spotted in duplicate (55,296 total clones) onto 5 high density filters. Strong positives were picked and inoculated into 34 - 96 well plates (3,237 clones in total) containing L broth. Clones were grown overnight at 37° C. Glycerol was added to a final concentration of 40% (Rose et al., 1990) and clones were frozen at -70° C.  Chapter 2: Materials and Methods  34  2.5.2 Human Chromosome 8p Enriched High Density Filters. High density B A C library filters were prepared by Michael Schertzer as in section 2.4.2 for cosmid high density filters, with the exception that chloramphenicol (37.5pg/ml) was used in the place of kanamycin. This high density procedure arrays the entire library (3,237 clones) on 2 filters labeled A and B.  2.5.3 Isolation of BAC DNA Bacterial artificial chromosomes were streaked from stocks stored at -70° C onto agar plates containing chloramphenicol (37.5 pg/ml) and grown overnight at 37° C. Individual colonies were inoculated into L broth containing chloramphenicol (30 ng/ml) and grown overnight, with shaking, at 37°C. B A C s were isolated by alkaline lysis, as in section 2.4.3, with the following modifications. After the addition of Solution III and centrifugation for 10 minutes the supernatant was carefully poured into a clean 1.5 ml eppendorf to avoid shearing of the DNA. A n equal volume of room temperature isopropanol was added and the D N A was pelleted by centrifugation at room temperature for 15 minutes. D N A was washed with 70% ethanol, dried, and 45 pl of I X TE(pH8.0)/RNase(25 pg/100 pl) was added. The pellet was carefully scraped from the side of the eppendorf tube and allowed to dissolve at room temperature for 20-30 minutes. The bottom of the eppendorf tube was gently tapped to resuspend the D N A , briefly centrifuged, and placed at 37° C for 45 minutes to allow any contaminating R N A to be digested. Samples were stored at -20° C.  Chapter 2: Materials and Methods  3 5  2.6 Yeast Artificial Chromosome (YAC) DNA Preparation and Isolation  2.6.1 Centre d'Etude du Polymorphisme Humain (CEPH) Mega YACs A chromosome 8 enriched subset of Y A C s from the CEPH Mega-YAC library (BellanneChantelot et al., 1992) was isolated by hybridization of Y A C Inttv-Alu PCR products onto a chromosome 8 specific somatic cell hybrid panel (Chumakov et al., 1992). Yeast lines containing Y A C s found positive for chromosome 8p markers, summarized at the MIT/Whitehead Institute database (http://www.genome.wit.mit.edu), were individually purchased from Research Genetics, Inc.  2.6.2 Isolation of Y A C DNA Cells containing Y A C s were streaked from glycerol stocks stored at -70° C (Rose et al., 1990) onto A H C minimal media agar plates (Brownstein et al., 1989) and allowed to grow for 3 days at 30° C.  2.6.2.1 Protocol 1: Y A C DNA Preparation Using Glass Beads A single red colony was inoculated into 5 mis of Y P D in a 15 ml tube and grown for 24 hours, with shaking, at 30° C. Since yeast cells sediment at the bottom of the 15 ml tube, 4 mis of Y P D were drawn off and the remaining 1 ml, containing the majority of the yeast cells, was placed in a 1.5 ml eppendorf tube. Cells were pelleted by centrifugation, washed with 500 pi distilled water, and resuspended in 500 pi GDIS. The cells were added to 1.5 ml eppendorf tubes containing 0.35 g of acid washed glass beads and 200 pi of 25:24:1 phenol:chloroform:isoamyl alcohol. Samples were vortexed for 2 A minutes, 200 pi of distilled water was added, mixed, and the l  samples were centrifuged at 13,000 rpm for 2 minutes. The upper aqueous layer, containing the  Chapter 2: Materials and Methods  36  D N A , was removed to a new 1.5 ml eppendorf tube to which an equal volume of room temperature isopropanol was added to precipitate D N A . D N A was collected by centrifugation at 13,000 rpm for 15 minutes. Pellets were washed with 1ml 70% ethanol, dried, and resuspended in 50 pi TE. GDIS 2% Triton X-100 1% SDS lOOmM NaCl 10mMTris-ClpH8.0 ImM EDTA  Acid washed glass beads Approximately 10 g of glass beads were washed with 5 ml of 0.1N HC1, rinsed three times with 5 ml of distilled water and excess water was removed with a pipetman. Beads were autoclaved for 20 minutes and dried.  2.6.2.2 Protocol 2: Minipreparation of Y A C D N A A single red colony was inoculated into 5 ml of Y P D containing ampicillin (lOOng/ml) in a 15 ml tube and allowed to grow for 2 days at 30°C, with shaking. 1.5 ml of the sample was centrifuged in a 1.5 ml eppendorf tube at 13,000 rpm for 2 minutes. The supernatant was completely removed, the pellet resuspended in 240 pi of lysis solution and incubated at 37° C for 1 hour. The sample was pelleted by centrifugation at 13,000 rpm for 2 minutes, the supernatant removed, the pellet resuspended in a mixture of 100 pi Y T E plus 10 pi 10% SDS and incubated at 65° C for 20 minutes. 40 pi of alkaline lysis solution III was added, mixed, and the sample was placed on ice for 30 minutes. The cellular debris was separated from the D N A by centrifugation for 3 minutes at 13,000 rpm. 150 pi of the supernatant was removed to a fresh 1.5 ml eppendorf tube and two volumes of 95% ethanol were added to precipitate the D N A . The  Chapter 2: Materials and Methods  37  D N A was pelleted by a 15 minute centrifugation at 13,000 rpm, washed with 70% ethanol, dried, resuspended in 100 pl of I X TE/RNase(25pg/100 pl), and incubated for 1 hour at 37° C to degrade any contaminating RNA. Sorbitol Solution 0.9 M Sorbitol 0.1 M ethylene diamine tetra-acetic acid (EDTA), pH8.0 100 m M tris (hydroxymethyl) aminomethane (Tris-Cl), pH8.0 Lysis Solution 200 pl sorbitol solution 20 pl 1:25 sorbitol solution: P-mercaptoethanol 20 pl sorbitol solution containing a few grains of lyticase or zymolase Y T E Solution 50 m M Tris-Cl, pH8.0 20 m M E D T A , pH8.0 A H C Minimal Media (ura-, trp-) Plates 0.67 g yeast nitrogen base w/o amino acids 1 g acid hydrolysed casein 2 g dextrose 1.86 mg adenine hemisulfate or adenine hydrochloride 1.2 g agar 100 m l d H 0 Adjust to pH 5.8 with concentrated HC1 (7 pl) 2  Y P D Media 4 g tryptone (peptone) 2 g yeast extract 4 g dextrose 200 ml d H 0 2  2.7 Restriction Enzyme Digestion  2.7.1 Standard Restriction Enzyme Digestion  Digestion of approximately 1 pg of D N A was carried out in either 20 pl or 25 pl final volumes, in 1.5 ml eppendorf tubes. Final concentrations of: I X Enzyme React buffer, I X B S A  Chapter 2: Materials and Methods  38  and 4U of enzyme, were used. Digestion was carried out for 1 Vi hours at 37° C, except for BssHll which was incubated at 50° C. Digestion was stopped by the addition of 5 pl of stop buffer.  2.7.2 Double Restriction Enzyme Digestion When D N A was digested by two enzymes, equal amounts of each enzyme were used, and the most appropriate buffer to allow complete digestion of the D N A was chosen. When the enzymes required incompatible buffers for optimal digestion, the D N A was digested using standard conditions with one enzyme. After completion of the first digestion, the total volume was increased to 100 pl with distilled water, an equal volume of TE-equilibrated phenol was added, and the solution mixed by vortexing. Samples were centrifuged for 2 minutes, the upper aqueous layer removed to a clean eppendorf tube, an equal volume of Sevag's Solution added, vortexed, centrifuged, and the upper aqueous layer removed to a clean eppendorf tube. A 1/10 volume of 3 M ammonium acetate was added, 2 volumes of 95% ethanol were added and the D N A was allowed to precipitate at -70° C for 30 minutes. Samples were centrifuged for 15 minutes at 13,000 rpm to pellet D N A , washed with 70% ethanol, dried, resuspended in 16 pl I X TE and digested with the second restriction enzyme. Digestion was stopped by the addition of 5 pl of stop buffer. Sevag's Solution 24:1 chloroform/isoamyl alcohol 10X B S A bovine serum albumin fraction V lmg/ml Restriction Enzymes and buffers £coRI(Gibco/BRL), 10X REACT®3 Sstl (Gibco/BRL), 10X REACT®2  Chapter 2: Materials and Methods  39  BgUl (Gibco/BRL), 10X REACT®3 M i l l (NEB), 10X N E B Buffer for BssHll digests BamHI (Gibco/ BRL), 10X REACT®3 Haelll (Gibco/BRL), 10X REACT®2 Kpnl (Gibco/BRL), 10X REACT®4 AforI(NEB), 10XNEB3 Ndel (Gibco/BRL), 10X REACT®3 EcoRL/Pvul(Gibco/BRL), 10XREACT®7 Hindlll (Gibco/BRL), 10X REACT®2 Sacll (Gibco/BRL), 10X REACT®2  2.8 Agarose Gel Electrophoresis 2.8.1 Standard Gel Electrophoresis Agarose gels were prepared by mixing appropriate amounts of agarose (Sigma) with I X TBE to achieve the desired concentration of gel (w/v). The agarose was dissolved by heating in a microwave oven, cooled until warm, ethidium bromide (EtBr) was added to a final concentration of 0.5 pg/ml, the gel was poured in casting trays and allowed to solidify. Electrophoresis was carried out in I X TBE. Digested D N A was run into agarose gels of the appropriate concentration based on expected fragment sizes. D N A digested with enzymes which cut infrequently, BssHU, Kpnl, and Ndel, were run into 0.4% agarose stabilized by a bottom layer of 1% agarose gel containing no EtBr. Sstl digests were run into 0.66% agarose gels, and digests using frequently cutting enzymes, £coRI, EcoYWPvul, BamHI, and BgM, were run into 0.8% agarose gels. In all cases 5 pg A, D N A digested with Hindlll and Sacll was run alongside D N A samples as a molecular weight marker. Agarose gels containing restriction enzyme digested D N A were run overnight at 30 volts.  40  Chapter 2: Materials and Methods  PCR products were run into 2-3% agarose gels prepared in the above manner. Samples were run into agarose gels for 2-4 hours at 80-150 V in order to achieve the desired separation. In all cases a molecular weight marker of 2.5 pg (j)X174 D N A digested with Haelll (Gibco/BRL) was run alongside PCR samples. 10X T B E 54 g Tris base 27.5 g boric acid 10mMEDTA,pH8.0 d H 0 to 500 ml 2  X D N A marker 40 pi X D N A (500 ng/pl) 16 pi 1 0 X B S A 16 pi 10XREACT®2 88 pi M i l l i Q filtered d H 0 2  5 pi 5 pi  Hindlll Sstll/Sacll  40 pi of stop buffer added after 1 '/ hours at 37° C 2  2.8.2 Pulsed Field Gel Electrophoresis (PFGE) A hexagonal array pulsed field gel apparatus was used. Three litres of I X T B E were prepared from autoclaved 10X TBE. 150 ml of this solution was removed to make a standard 1% agarose gel containing no EtBr. Approximately 2 litres of the remaining I X T B E were poured into the PFG apparatus, the pump and cooler were turned on, and the I X T B E was cooled to 12°C. The 1% agarose gel, prepared as in section 2.8.1, was cast in the PFG casting tray and allowed to cool. The comb was removed and a small slice of Low Range PFG marker(NEB) or Lambda ladder PFG marker (NEB) in agarose was loaded into the two lanes that would flank the samples. The casting tray was placed into the PFG apparatus, the volume of I X T B E was adjusted with the remaining solution to just cover the gel. Samples were digested with NotI and loaded into the  Chapter 2: Materials and Methods  41  gel. The run conditions were as follows: 36 second pulses alternating from North/South to East/West, at 200 V for 20 hours. The gel was removed from the casting tray and stained in a 0.5 pg /ml EtBr solution for 30 minutes.  2.9 DNA Imaging for Agarose Gels Agarose gels were placed on a FOTO/Convertible (FOTODYNE) Ultraviolet radiation box to visualize the D N A containing intercalated EtBr. The gels were photographed using the Scion Gel Imaging System via a C C D camera linked to a Power Macintosh computer.  2.10 Southern Transfer Agarose gels were photographed alongside a ruler, trimmed to remove excess agarose and blotted onto nylon membranes by the method of Southern (Southern, 1975). Those gels containing D N A fragments of 15 kb or more were first treated with 0.25 N Hydrochloric Acid for 30 minutes to depurinate the D N A . Gels were placed in 1.5 M NaCl/0.5 M NaOH for 30 minutes to denature the D N A . The solution was removed and replaced with neutralization solution (1 M Tris/1.5 M NaCl) for 30 minutes, followed by rinsing with distilled water. A Southern blot apparatus was assembled. Approximately 500 ml of 10X SSC was placed in a pyrex dish, a piece of glass was placed across the top of the dish and two sheets of 3M Whatmann paper wet with 10X SSC were placed on the glass to form a wick. Air bubbles were removed by rolling a pipet over the wick. The agarose gel was placed on the wick, a piece of Hybond N+ nylon membrane (Amersham) cut to the gel dimensions was briefly wet with distilled water and placed over the gel, ensuring that the top of the membrane was flush with the wells of the agarose gel. A piece of Whatmann paper cut to the gel dimensions was wet with  Chapter 2: Materials and Methods  42  distilled water and placed on top of the membrane, followed by a similar piece of dry Whatmann paper. The area surrounding the gel was covered with plastic and paper towel was placed over the gel to draw the solution up through the gel and allow the D N A to bind to the charged nylon membrane. The apparatus was left overnight, disassembled the next morning, the membrane placed between two pieces of dry Whatmann paper and placed at 80° C for 2 hours to cross link the D N A to the membrane. 20X SSC 175.3 g NaCl 88.2 g sodium citrate d H 0 to 1 litre 2  2.11 Radioactive Labeling of D N A  2.11.1 Random Primer Labeling  The random prime labeling method of Feinberg and Vogelstein (1984 a, b) was used. 15 pi of 1 ng/pl probe was boiled for 5 minutes, placed on ice for 5 minutes and mixed with 2.5 pi ccdATP (NEN Dupont 10 pCi/pl), 5 pi 5X Oligolabeling buffer without dATP (OLB-A) 32  containing random hexamer D N A , 2.5 pi 1 OX B S A , and 1U Klenow enzyme (Pharmacia or Gibco/BRL). Labeling was allowed to occur at room temperature overnight. 5X O L B - A Mix solutions A : B : C to a ration of 100:250:150 Solution O: 1.25 M Tris-Cl pH8.0 0.125 M M g C l Solution A : 1 pi Solution O 18 pi P-mercaptoethanol 5 pi each dGTP, dCTP, dTTP (each dNTP is 0.1 M in 3 m M Tris pH7.0, 0.2 m M EDTA) Solution B: 2MHepespH6.6 Solution C: Hexadeoxyribonucleotides suspended in I X TE to 90 OD/ml 2  43  Chapter 2: Materials and Methods  2.11.2 Klenow Primer Extension Labeling 15 pl of 1 ng/pl probe plus 0.5 pl of each primer pair was boiled for 5 minutes, placed on ice for 5 minutes, and mixed with l p l a-dATP (NEN Dupont lOuCi/pl), 5 pl 5X K E L buffer, 2.5 32  pl 10X B S A , 1U Klenow enzyme (Pharmacia or Gibco/BRL). Samples were placed at 37° C for 1 Vi hours to allow labeling by Klenow primer extension to occur. 5X K E L buffer O L B - A with the exception that Solution C is d H 0 2  2.11.3 y-end Labeling D N A was labeled in a 10 pl reaction containing 50 pmoles of D N A primer, I X Kinase Forward Reaction Buffer (Gibco/BRL), 1U T4 Kinase (Gibco/BRL) and 25 p C i y-dATP. Labeling was carried out at 37° C for 60 minutes followed by 5 minutes at 95° C to heat inactivate the enzyme. The primer was stored at -20° C.  2.11.4 Nick Translation of Plasmid DNA 1 pg of plasmid D N A (usually pBluescript) was labeled by nick translation using the Gibco/BRL nick translation kit. To 1 pg of D N A was added 5 pl Solution A l , 5 pl S-aATP 35  (NEN/Dupont), and d H 0 to 45 pl. 5 pl of Solution C (DNA Polymerase/DNasel) was added 2  and labeling was allowed to proceed for 1 hour at 14° C. 5 pl of Solution D (300 m M Na EDTA, 2  pH8.0) was added to stop the reaction. The labeled plasmid was diluted with d H 0 to a final 2  concentration of 5 ng/pl. Solution A l 0.2 m M dCTP, dGTP, dTTP 500 m M Tris-HCl, pH7.8 50 m M M g C l 100 m M 2-mercaptoethanol 2  Chapter 2: Materials and Methods  44  2.12 Pre-Hybridization Treatment of Probe  2.12.1 Removal of Unincorporated Nucleotides 25 pl of nick translation stop buffer (NTSB) was added to the probe upon completion of labeling. The probe was centrifuged at 1,000 rpm through a Sephadex G-25 spin column. The Sephadex G-25 beads capture unincorporated nucleotides while allowing nucleotides incorporated into D N A to remain in solution and pass through the column. The solution containing the probe was brought back up to 50 pl by the addition of M i l l i Q filtered distilled water. NTSB 50mM E D T A 20mM NaCl 0.1% SDS 500 mg/ml salmon sperm D N A  2.12.2 High Complexity Probes Probes containing little or no common repetitive D N A were boiled for 5 minutes, placed on ice for 5 minutes and used for hybridization. These included PCR products, the megasatellite D N A and the subcloned fragments of the cosmids 153G8 and 39A7  2.12.3 Low Complexity Probes Probes containing common repetitive D N A were mixed with 5.5 pl 25X SSC and 15 pl human placental D N A (10.6 mg/ml) (Sigma), boiled for 5 minutes to denature the D N A and allowed to preanneal at 65° C for 1 hour before being used for hybridization. These probes included B A C inserts and end sequences known to contain repetitive D N A .  Chapter 2: Materials and Methods  45  2.12.4 Probes to High Density Filters Probes being used to screen high density filters were mixed with 10 pi of plasmid (5 ng/pl) labeled with S - a A T P by nick translation, to allow background visualization of high density 35  filter grid, boiled for 5 minutes to denature the D N A , placed on ice for 5 minutes and used for hybridization.  2.13 Hybridization, Post-Hybridization Washes, Autoradiography Hybridizations were carried out in heat-sealable hybridization bags containing the probe and enough hybridization buffer to wet the filter plus 5 ml. Hybridization was allowed to occur overnight at 65° C followed by a 5 minute wash in I X SSC/1% SDS at room temperature and a 50 minute stringent wash in 0.2X SSC/1% SDS at 65° C. Filters were partially air dried, wrapped in plastic wrap and placed on film on a lightening plus screen at -70° C or at room temperature for the appropriate length of time required for a clear signal. When screening libraries, positive signals were read off the high density filters and the grid locations were matched with the corresponding clone address. If necessary, blots were stripped of signal by pouring boiling 0.5% SDS over them and allowing the solution to cool to room temperature, rinsing with distilled water and allowing them to dry. Hybridization Solution 150 ml 2 0 X S S C 50 ml 5X Denhardt's Solution (filter sterilized) 7.5 ml 20% SDS 50 ml 100 mg/ml salmon sperm D N A 242.5 ml distilled water 100X Denhardt's Solution 4% Ficoll 400 4% polyvinyl pyrrolidone 4% (w/v) bovine serum albumin (BSA)  Chapter 2: Materials and Methods  46  2.14 Extraction of DNA from Agarose Gel  2.14.1 Qiagen Gel Extraction In order to subclone restriction enzyme fragments and PCR products the D N A must be free of agarose contaminants. The QIAEXII Agarose Gel Extraction Kit (Qiagen) was used to extract D N A that was to be used for subcloning experiments. The D N A band was excised from the agarose gel and placed in an eppendorf tube containing Buffer QX1, which solubilizes agarose and creates a high salt environment, and Q I A E X II silica gel particles, which adsorb nucleic acids in the presence of high salt. The solution was incubated at 50° C for 10 minutes with intermittent vortexing to keep the Q I A E X II in suspension. The tube was centrifuged at 13,000 rpm for 30 seconds and the supernatant removed. The pellet was washed once with Buffer QXI to remove residual agarose and twice with Buffer PE to remove salt contamination. The pellet was air-dried followed by a 5 minute incubation in 20 pl distilled water for elution of the D N A from the QIAEXII particles occurs. The QIAEXII particles were pelleted by centrifugation and the supernatant, containing the D N A , was removed to a clean eppendorf tube and stored at -20° C.  2.14.2 Low Melting Point Gel Extraction The band of interest was excised from the agarose gel and placed in the well of a low melting point (LMP) agarose gel (Sigma). The D N A was run into the L M P gel in fresh I X TBE at 80 V until the band had migrated approximately 5 cm into the gel. This removes contaminating  Chapter 2: Materials and Methods  47  plasmid D N A . The band was extracted, placed in I X TE to a concentration of 1 ng/pl, boiled for 5 minutes to liberate the D N A from the agarose and stored at 4° C.  2.14.3 DEAE Protocol A cut in the agarose gel was made just below the band of interest. The gel was pried apart using clean forceps and a small piece of D E A E paper (Schleicher & Schuell) was inserted. D E A E paper carries a functional group diethylaminoethyl in its protonated form that has a binding capacity of 20 pg/cm for D N A , RNA, and protein. The gel was run at 100 V for 2  approximately 5 minutes until the band had just contacted the D E A E paper. The D E A E paper was removed and placed in an 1.5 ml eppendorf tube containing 500 pi of NaCl, creating a high salt environment. The band was eluted from the D E A E paper at 65° C for 45 minutes. The paper was checked under U V illumination for any residual D N A , removed, and two volumes of 95% ethanol were added the tube. The sample was placed at -70° C, to facilitate precipitation, until frozen. The D N A was collected by centrifugation at 13,000 rpm for 15 minutes. The pellet was washed in 1 ml 70%) ethanol, air dried and resuspended in the appropriate amount of I X TE.  2.15 Subcloning of Cosmid Fragments Cosmid fragments were subcloned into either pBluescriptll KS (Stratagene) vector or a modified pBluescript vector. The modified vector has the Spel and Xbal restriction sites replaced with an AscI restriction site allowing i&sHII fragments to be cloned (DeBella et a l , unpublished, 1998). Two selectable markers are contained in pBluescript: the ampicillin resistance gene, which allows selectable growth on ampicillin of those clones containing the vector, and the coding region for the amino-terminal region of the P-galactosidase gene (lacZ).  Chapter 2: Materials and Methods  48  Contained within the cloning region of the lacZ gene is a polycloning site which does not disrupt the reading frame, and therefore the gene product, unless a D N A fragment has been cloned into it. When the vector is present within a host such as DH5cc Escherichia coli, which encodes the carboxy-terminal portion of the P-galactosidase gene, the two fragments associate to form an enzymatically active protein (Ullman et al., 1967). Clones containing the active protein appear blue when grown in the presence of 5-bromo-4-chloro-3-indolyl-P-D-galactoside (X-gal) and isopropyl-thiogalactoside (IPTG). Insertion of D N A into the polycloning site of the vector, disrupting the transcription of the amino-terminal portion of the gene, results in white colonies.  2.15.1 Preparation of pBluescript Vector The pBluescriptll K S or the pBluescriptll KSAsc was cut with the appropriate restriction enzyme to achieve compatible ends with the D N A fragment to be subcloned. The vector was digested according to standard restriction enzyme conditions in a final volume of 50 pl. The vector D N A was precipitated with 95% ethanol, washed with 70% ethanol, air-dried and resuspended to a final concentration of 100 ng/pl. 1 pl was run out on an agarose gel to ensure complete digestion. The vector was stored at -20° C.  2.15.2 Subcloning Cosmid EcoRI fragments were randomly cloned in order to isolate fragments for study.  1 pg  of D N A was double digested with EcoRI/Pvul using the double restriction enzyme digestion procedure (section 2.7.2). 5 pl of stop buffer was added and the sample was run into a 0.8% gel in I X T B E at 30 V , overnight. Pvul cleaves once within the vector and it was determined if there were any sites of cleavage in the cosmid insert. Once this had been determined, 1 pg of  Chapter 2: Materials and Methods  49  D N A was digested with EcoRl/Pvul, precipitated, and resuspended in 16 pi d H 0 . By digesting 2  with EcoRI in the presence of Pvul, the vector is cut into two pieces incompatible with ligation into EcoRI cleaved pBluescript. This selects against the ligation of vector D N A into pBluescript.  2.15.3 Ligation of Restriction Enzyme Digested Fragments into pBluescript Ligation reactions were carried out in a final volume of 20 pi containing approximately 100 ng QIAGEN gel purified D N A or 1 pg restriction enzyme digested cosmid D N A , I X N E B T4 ligase buffer, 20 ng pBluescript digested with the appropriate restriction enzyme, and 1U T4 D N A ligase (NEB). A positive control, containing only pBluescript with cohesive ends was included with each ligation performed. The ligation reaction was carried out overnight at 14° C.  2.15.4 Deletion Clones Plasmid deletion clones were constructed by digestion of 1 pg of the original clone with an enzyme present within the polylinker and the insert, followed by religation using the procedure outlined in section 2.15.3, with the exception that pBluescript was not included in the reaction.  2.16 Transformation of Ligated DNA into Competent E.coli Cells Competent DH5a E. coli cells (Inoue et a l , 1990) were removed from -70° C and thawed on ice. 50 pi was placed into a precooled 15 ml tube and 10 pi of the ligation reaction was added, mixed, and left on ice for 30 minutes. The cells were heat shocked for 45 seconds at 42° C and returned to ice for 2 minutes. 400 pi L broth was added and the cells were allowed to divide for 45 minutes at 37° C. 120 pi of transformed cells was pipeted onto agar plates, spread, and  50  Chapter 2: Materials and Methods  incubated overnight at 37° C. Agar plates containing ampicillin (50 pg/pl) were used for cosmid ends and X I A plates were used for D N A subcloned into pBluescript and for deletion clones. X I A plates: Agar plates (see section 2.4.1) plus 20 mg ampicillin, 25 mg X-gal, 50 mg IPTG NOTE: plates must be cooled and stored in the dark 2.17 D N A Sequencing  2.17.1 D N A Preparation for Sequencing  In order to obtain accurate sequencing results, the D N A to be sequenced must be especially pure. This is achieved through use of the Qiagen Plasmid Mini Kit (QIAGEN). A modified alkaline lysis protocol is used to isolate D N A : 1.5 ml of L broth, inoculated with a single bacterial colony containing the subcloned fragment of interest, grown overnight, was centrifuged and the pellet resuspended in 300 pl of cold Buffer P l . 300 pl of ice cold Buffer P2 was added, mixed gently by inversion and incubated on ice for 5 minutes. The sample was centrifuged for 10 minutes to pellet cellular debris, and the supernatant removed to a QIAGEN-tip preequilibrated with 1 ml of Buffer QBT. Buffer QBT and the supernatant create a low salt environment, allowing D N A to bind to the anion-exchange resin contained within the QIAGENtip. Once the supernatant had drained through the tip, 4 ml of Buffer QC, a medium concentration salt solution, was added to the tip to remove impurities, such as R N A and proteins. The purified D N A was eluted into a clean 1.5 ml eppendorf tube by addition of 800 pl of Buffer QF, a high salt buffer at pH7.0. The D N A was precipitated using 0.7X volume isopropanol, pelleted by centrifugation at room temperature, and washed with 70% ethanol to remove excess salt. The pellet was resuspended in 15 pl Milli-Q purified d H 0 . 1 pl of each sample was 2  51  Chapter 2: Materials and Methods  digested with EcoRI (section 2.7.1) and the fragments were separated on an agarose gel to determine D N A concentration.  2.17.2 Automated Sequencing Sequencing was carried out by the automated sequencer (Applied Biosystems Model 373 Stretch), using either the M13F ( 5 ' - C C C A G T C A C G A C G T T G T A A A A C G - 3 ' ) or the M13R (5'A G C G G A T A A C A A T T T C A C A C A G G - 3 ' ) primer, which flank the cloning sites of the pBluescript vector. Sequence was obtained from 500 ng of QIAGEN purified double stranded D N A template using ABI's AmpliTaq FS DyeDeoxy™ Terminator Cycle Sequencing. A l l four base reactions are carried out in one eppendorf tube on a thermocycler. The sequence obtained by this method can be 98.0% accurate to more than 650 base pairs. Sequencing was carried out by the U.B.C. sequencing laboratory. The procedure used is as follows: 8.0 pi of terminator premix, 500 ng of template, 3.2 pmol of primer (M13 or T3), and d H 0 to a final volume of 20.0 2  pi were mixed in a 0.6 ml eppendorf tube. The reaction mixture was overlaid with a drop of mineral oil and the tubes were placed in a preheated (96° C) thermocycler. Twenty-five cycles of: 96° C for 30 seconds, 50° C for 15 seconds, and 60° C for 4 minutes, were carried out followed by incubation at 4° C until reaction was removed from the thermocycler. Each reaction was transferred to a 1.5 ml eppendorf and precipitated with 1/10 volume 3 M sodium acetate, pH4.6, and 2 volumes 95% ethanol. This precipitation step removes excess dye terminators. The D N A was pelleted by centrifugation, dried, and stored at -20° C. Fragments were separated on sequencing gels using an automated sequencer (Applied Biosystems Model 373 Stretch).  Chapter 2: Materials and Methods  52  2.17.3 DNA Sequence Analysis  2.17.3.1 BLAST Sequences obtained from the subcloned fragments were analyzed using the basic local alignment search tool (BLAST). B L A S T uses a rapid database searching algorithm that optimizes local similarities between sequences and extends these alignments based on defined match and mismatch criteria (Altschul et al., 1990). The statistical significance of any similar segments between the query sequence and a given database sequence is evaluated and reported. Only P values of l.lxlO" or less were examined. The P value is the likelihood of the match 21  representing a random alignment. For the subclones analyzed, any match to the megasatellite sequence was at a P value less than 2xl0" . B L A S T queries Genbank, the National Institute of 33  Health database, maintained at the National Center for Technology (NCBI) (http://www.ncbi.nlm.nih.gov), which contains all reported nucleotide and protein sequences. BLASTnr accesses the non-redundant database, BLASTsts accesses the STS database and BLASTest accesses the EST database for matching sequences. A l l three types of B L A S T queries were carried out on the sequenced D N A .  2.17.3.2 Clustal W Sequences obtained from the subcloned fragments of 153G8 and 39A7 were compared to each other, and to the 4p megasatellite sequence (Genbank accession #: D38378) using the Clustal W program available at http://www2.ebi.ac.uk/clustalw (Thompson et al, 1994). Clustal W aligns sequences with respect to one another, allowing for insertions and deletions. Problems occur when two sequences are of very different lengths, or when there are long regions which can not  53  Chapter 2: Materials and Methods  be aligned. Therefore, when comparing subclone sequence with megasatellite sequence, B L A S T alignments were consulted for regions of homology, the regions were aligned using Clustal W, followed by alignment of sequences 'by eye'.  2.18 Cosmid Termini Isolation  2.18.1 Isolation of T3 End The following procedure is diagrammed in figure 6. Bglll is a 6 cutter that cleaves fairly frequently in human D N A . s-Cosl vector contains a single Bglll restriction enzyme site. Cosmids were digested with Bglll using standard conditions (section 2.7.1), the D N A was precipitated, and religated (section 2.15.3 with the exception of d H 0 used in place of plasmid 2  DNA). Religation results in a variety of products including deleted cosmids made up of the T3 end of the cosmid insert and 4300 bp of vector D N A containing the ampicillin resistance gene. Transformation (section 2.16) of these products into competent E. coli cells and growth on agar plates containing ampicillin results in a selection for the T3 end of the cosmid insert.  2.18.2 Isolation of T7 End 1 pg of cosmid was digested with EcoRI using standard conditions and run overnight at 30V into a 0.8% agarose gel in I X T B E (sections 2.7.1, 2.8.1). The gel was transferred to nylon membrane by Southern blotting, hybridized overnight at 42° C with T7 primer labeled by y-end labeling, washed, placed on film (sections 2.10 , 2.11.3, 2.12, 2.13), and the size of the EcoRI fragment containing the T7 end was determined. The cosmid was again digested with .EcoRI and run into 0.8% agarose in I X T B E at 30 V , overnight. The band of interest was extracted by low melting point agarose gel extraction (section 2.14.2).  Chapter 2: Materials and Methods  Bgm  ^  digestion with BgUl  Bgta  •  Religation, transformation, and growth on Amp+ agar  Figure 6: Construction of T3 end deletion clones from cosmid clones. Dark line represents vector D N A , light line represents human D N A , E=EcoKl. Selection for those colonies that grow on ampicillan results in isolation of T3 end of the human D N A insert.  Chapter 2: Materials and Methods  55  2.19 BAC Termini Isolation by Bubble PCR Bubble PCR was initially developed as a method for isolating Y A C termini (Riley et al., 1990). We have modified the technique to allow isolation of B A C termini. Briefly, ligation of a linker containing a region of mismatch, to restriction enzyme digested D N A , allows PCR to be carried out preferentially from only those fragments that contain vector plus insert D N A (see figure 7).  2.19.1 Digestion of DNA Approximately 500 ng of B A C D N A was digested with Ddel in a 30 pl reaction volume, final concentration of I X BSA, I X REACT®2, 4U enzyme. Reactions were carried out at 37° C for 2 hours. D N A was precipitated and resuspended in 12 pl of distilled water.  2.19.2 Preparation of Linker DNA The bubble linker was prepared by mixing equimolar amounts of bottom strand oligo 221 (5'C T C T C C C T T C T C G A A T C G T A A C C G T T C G T A C G A G A A T C G C T G T C C T C T C C T T G - 3 ' ) and top strand oligo 222 (5' - T N A C A A G G A G A G G A C G C T G T C T G T C G A A G G T A A G G A A C G G A C G A G A G A A G G G A G A G - 3 ' ) with 25 m M NaCl. The solution was boiled for 2 minutes followed by incubation for 5 minutes at 65° C. The solution was left at room temperature to cool.  Chapter 2: Materials and Methods  56  2.19.3 Ligation of Linker to B A C D N A The digested D N A was mixed with 15 u l of linker D N A , 3 u l of 5X N E B ligation buffer and 1 ul of N E B T4 D N A ligase enzyme. The ligation mixture was left at 14° C overnight and adjusted with I X TE to a final volume of 100 ul.  2.19.4 P C R Isolation of B A C Termini 10 pi of the ligation mixture was used in a standard PCR as in section 2.1 using 10 pmole each of oligo 224 and oligo T7 (T7 end) or oligo 224 and oligo SP6 (SP6 end) (see table 2 for primer sequences). The PCR annealing temperature was 40° C. Two samples of each B A C were run for T7/224 and 1 sample for SP6/224. 1 pi of each reaction was saved in case of insufficient amplification. One sample of T7/224 was digested with Pstl to remove vector sequences. This was done by drawing out the sample from below the oil and placing it on parafilm. The drop of sample was rolled across the parafilm to remove any residual mineral oil and then placed in a 500 pi eppendorf tube. A standard restriction digest by Pstl was carried out using the PCR sample as the D N A source (without any precipitation from the PCR mix). After 1 hour, stop buffer was added to the digested sample and the remaining samples, and they were loaded into a 2% agarose gel. Gels were run for approximately 2 hours at 100 V in I X TBE. Digested T7 ends and undigested SP6 ends were extracted from the gel either by the D E A E protocol or by L M P agarose excision (section 2.14).  Chapter 2: Materials and Methods  Insert  Ddel Ddel  Ddel  Ddel  B A C vector  Digestion with Ddel T7 SP6 Ligation of linker (=  1' T7 SP6 PCR using primer 224 and either primer T7 or SP6 T7 SP6 224 SP6 224  a. extension from SP6 or T7 b. primer 224 anneals to new strand  SP6  T7 224 T7 224  c. exponential amplification digestion with PstI to remove extra vector sequences  224  Figure 7: Bubble P C R protocol for B A C end isolation. Inset: The bubble linker. Oligo 224 is identical to one strand of the linker but not complementary to the other strand due to a region of mismatch. Therefore, amplification can not occur from oligo 224 unless extension has occurred from either the SP6 or T7 primer allowing the SP6 and T7 ends of a human D N A insert to be preferentially amplified.  Chapter 2: Materials and Methods  58  2.20 Preparation of D N A from Whole Blood  2.20.1 Isolation and Lysis of White Blood Cells NfLCkTris solution was warmed to 37° C. 5 volumes of this solution was added to whole blood in 50 ml polypropylene centrifuge tubes and incubated at 37° C for 5 minutes. Samples were centrifuged for 10 minutes at 2000 rpm. The supernatant was aspirated leaving 4-5 ml above the pellet. The pellet was resuspended in 10 ml saline, centrifuged at 2000 rpm for 10 minutes, the supernatant aspirated and the saline wash repeated. The cells were resuspended in 2 ml high T E buffer and immediately lysed by the injection of 2 ml of lysis mixture using a 5 ml syringe with a 16-18 gauge needle. The injection of the lysis mixture sufficiently mixes the suspension to cause complete, instantaneous lysis. Lysate was stored at 4° C. NH Cl:Tris solution 0.1395M NH C1 0.017M Tris, pH7.65 4  4  Saline 0.85% NaCl High T E Buffer lOOmM Tris, pH8.0 40mM E D T A , pH8.0 Lysis Mixture lOOmMTris, pH8.0 40mM E D T A , pH8.0 0.2% SDS  2.20.2 D N A Extraction A n equal volume of TE-equilibrated phenol was added to the lysate and gently mixed by inversion to a milky white emulsion. The sample was placed on a rotator and gently mixed for  Chapter 2: Materials and Methods  59  10 minutes, followed by centrifugation at 2000 rpm for 4 minutes. The upper aqueous phase was removed with a large bore pipet made by cutting the tip off of a Pasteur pipet and fire polishing it. The phenol extraction was repeated until no interphase remained. A n equal volume of 24:1 chloroform:isoamyl alcohol was added, the sample placed on a rotator for 10 minutes and centrifuged for 10 minutes at 2000 rpm. The upper aqueous layer was removed to a 15 ml tube.  2.20.3 DNA Precipitation The volume of the aqueous layer was determined and a 1/10 volume of 4 M ammonium acetate was added. A n equal volume of isopropanol was added, the tube swirled to entangle the D N A into a small ball and the D N A removed from the solution using a curved end Pasteur pipet. The D N A was washed using a stream of 70% ethanol, allowed to briefly air dry and dissolved in 1 ml low T E buffer. The D N A was allowed to dissolve overnight on the rotator at room temperature. The concentration of D N A was estimated by running 1 pl in 5 pl stop buffer into a 0.8% agarose gel. Low T E Buffer l O m M Tris, pH8.0 1 m M E D T A , pH8.0  2.21 Epstein Barr Transformed Cell Line Epstein Barr transformed lymphoblastoid cell lines, established from patients G.S. (MGV280), T.P. (GM13540, Coriell Institute), and the Dhooge et al. (1994) patient (MGV-292) were maintained in continuous suspension culture. G.S. carries a de novo inversion duplication (8)(pl2-»p23.1) (Dill et al. 1987), T.P carries a de novo inversion duplication (8)(pl 1.2->p23) (table 1), and the Dhooge et al. (1994) patient carries a maternally inherited distal 8p duplication  Chapter 2: Materials and Methods  60  of undetermined band origin and orientation (direct or inverted). Cells were grown in RPMI 1640 medium (StemCell Technologies) supplemented with 15% fetal bovine serum (Gibco/BRL). Cultures were split bi-weekly to maintain optimum growth.  2.22 Metaphase Chromosome Preparation  2.22.1 Pre-harvest Treatments Epstein Barr transformed lymphoblastoid cell line cultures were harvested 2 days after subculturing. Chromosomes were prepared according to standard protocols (Verma and Babu, 1989) with some modifications. 15-17 hours prior to harvest 1/100 volume methotrexate (1.25 pg/ml) was added to cell culture to synchronize cell growth by arresting cells in S phase. 4 A to l  5 hours prior to harvest, cells were rescued from arrest by removing as much medium as possible, replacing with fresh pre-warmed medium, and adding 1/100 volume bromodeoxyuridine (1 mg/ml). Thirty minutes prior to harvest, all but 10 ml of medium was removed and 0.5 ml colcemid (10 pg/ml) was added.  2.22.2 Chromosome Harvest Cell cultures were transferred to 15ml centrifuge tubes and centrifuged for 10 minutes at 200 g. The supernatant was removed and the pellet was thoroughly resuspended. 8 ml of prewarmed 0.75 M KC1 hypotonic solution was added, the cells were carefully resuspended by swirling, and placed at 37° C for 5-10 minutes. 2-3 ml 3:1 methanol:glacial acetic acid fixative was added to each tube, dropwise, with swirling. Cells were pelleted by centrifugation at 200 g for 10 minutes, the supernatant removed and the cell pellet resuspended. The cells were washed a  61  Chapter 2: Materials and Methods  further 3 times in 5 ml fixative. After the final wash the cells were resuspended in 10 ml fixative and stored at -20° C.  2.23 Slide Preparation and Storage Fixed cells were pelleted by centrifugation for 10 minutes at 200 g, the supernatant removed and the cells resuspended in enough fixative to result in a slightly cloudy mixture. Pre-washed slides stored at 4° C in 95% ethanol were rinsed multiple times in distilled water to remove all traces of ethanol and result in a smooth sheen to the slide when removed from the distilled water. A small volume of the cell suspension was taken up with a Pasteur pipet and a single drop was dropped onto one end of the slide held at a 45 ° angle. The solution was allowed to move down the length of the slide and then flicked to move the solution across the width of the slide. Slides were allowed to dry and were examined under phase contrast to assess quality. Slides were used immediately or were stored in 70% ethanol at -20° C (Jauch et al., 1990).  2.24 Fluorescence In Situ Hybridization (FISH) One use of fluorescence in situ hybridization is to allow direct visual localization of a human D N A clone on metaphase chromosomes. This is particularly useful in investigations of aberrant chromosomes in patient metaphases.  2.24.1 Nick Translation Labeling of Probes Hybridization probes were labeled with biotin by nick translation using the Gibco/BRL nick translation kit. To 1 pg of D N A was added 5 pl Solution A4, 1 pl 1 m M biotinylated dUTP (Boehringer Mannheim), and d H 0 to 45 pl. 5 pl of Solution C (DNA Polymerase/DNasel) was 2  added and labeling was allowed to proceed for 1 A hours at 14° C. 5 pl of Solution D (300mM l  62  Chapter 2: Materials and Methods  Na^EDTA, pH8.0) and 1.25 pl 5% (w/v) SDS was added to stop the reaction. Probes were mixed with 67.84 pg human placental D N A (Sigma) and d H 0 to 100 pl. 1/10 volume of 3 M 2  ammonium acetate was added. 2 volumes of 95% ethanol were added to precipitate the D N A and the solution was placed at -70° C for 30 minutes to overnight. D N A was pelleted by centrifugation, washed two times with 70% ethanol, dried and resuspended in 8 pl I X TE (final concentration approximately 100 ng/pl) Solution A4 0.2mM each dATP, dCTP, dGTP 500mM Tris-HCl, pH7.8 50mM M g C l lOOmM 2-mercaptoethanol 2  2.24.2 Pre-Hybridization Slide and Probe Preparation Slides were pre-treated with 2X SSC at 37° C for 30 minutes followed by dehydration through a room temperature ethanol series (70%, 85%, 95%, 2 minutes each). Denaturation was carried out in 70% formamide/2X SSC, pH 7.0 at 74° C for 2 minutes followed by dehydration through an ice cold ethanol series (70%, 85%, 95%, 2 minutes each). Hybridizations using cosmid 153G8 were carried out by Stanya Jurenka. 100 ng of labeled probe was mixed with 10 pl of Hybrisol VII (Oncor) and denatured at 74° C for 10 minutes. 1 pl of D8Z1 ct-satellite centromere probe (Oncor) was added to the hot solution. This allows some denaturation of the repetitive centromere D N A but allows clearer visualization of both probes than if the Oncor probe is denatured completely. For biotin labeled B A C 223B23, 100 ng of probe was mixed with 5 pl of Hybrisol VII (Oncor) and denatured at 74°C for 10 minutes. The denatured probe was placed at 37° C for 1 Vi hours to allow preannealing of highly repetitive sequences. Just prior to hybridization, 5 pl of Hybrisol VII (Oncor) was heated at 74° C for 5 minutes, 1 pl of  Chapter 2: Materials and Methods  63  D8Z1 a-satellite centromere probe (Oncor) was added to the hot solution, mixed, the solution was allowed to cool slightly and was mixed with the B A C 223B23 probe.  2.24.3 Hybridization The Hybrisol VII (Oncor) diluted probe was pipeted onto the slide and a small coverslip was placed over it. The coverslip was sealed with Elmers Rubber Cement and placed in a humid chamber at 37° C overnight.  2.24.4 Post-Hybridization Washes The coverslip was removed and the slide placed in 50% formamide/2X SSC, pH7.0 for 15 minutes at 43° C followed by 8 minutes in 2X SSC at 37° C. The slide was washed three times in I X PBD for two minutes. 1XPBD 32.57 g N a H P 0 - d H 0 17.3 g Na2HP0 (anhydrous) 9 g NaOH 2 ml Triton X d H 0 to 2L 2  4  2  4  2  2.24.5 Detection, Microscopy, and Photography Hybridized probes were visualized using FITC-avidin (FA) which has a strong affinity for biotin and amplified using cycles of biotinylated anti-avidin D (BAAD) followed by FA. 100 pi of antibody diluted in P M N (5 pg/ml) was pipeted onto wet slides, covered with a plastic coverslip and placed at 37°C for 5 minutes. Excess antibody was removed by three - 2 minute washes in I X PBD. The procedure was repeated, in the order: FA—»BAAD—»FA. The D N A was counterstained using propidium iodide (PI) mixed with antifade (0.6 pg/ml) and visualized  Chapter 2: Materials and Methods  64  using a Zeiss fluorescence microscope, equipped with a propidium iodide/FITC specific filter set (excitation BP450-490, dichroic FT510, barrier LP520). Photographs were taken with 1600 A S A colour film (Fuji) and processed.  1XPMN 50 ml I X PBD 2.5 g skim milk powder place at 37° C overnight, pellet sediment, filter sterilize supernatant plus 50 pi N a azide (0.2% w/v), store at 4° C  65  Chapter 3: Results  Chapter 3: Results  3.1 Polymorphic STS Genotyping STS genotyping at D8S503 (figure 3, table 2) was carried out on 9 inv dup(8p) patients previously reported by Floridia et al. (1996) to investigate whether this marker is located in the single copy region of the inv dup(8p) chromosome. Marker D8S503 is located on the sexaverage genetic map (Dib et ai, 1996, Gyapay et al., 1994) proximal to D8S349 (figure 3) which is deleted from the inv dup(8p) chromosome (Floridia et al., 1996). If this marker is included in the single copy region then D8S503 is the most informative distal marker in this region. Nine inv dup(8p) patients reported by Floridia et al. (1996) as, 10 (CM), 13 (TS), 9 (CI), 14 (MM), 2 (PJ), 5 (SA), 15 (TG), 1 (GL), and 7 (AM), and renamed patients 1 (CM), 2 (TS), 3 (CI), 4 (MM), 5 (PJ), 6 (SA), 7 (TG), 8 (GL), and 9 (AM) for this report, were genotyped at marker D8S503 (primer sequences and annealing temperature listed in table 2). A l l nine patients were previously shown to have inherited the aberrant chromosome from their mother (Floridia et al., 1996). Patients 2, 3, 6, 7, 8, and 9, were informative for the inheritance of a single D8S503 allele from each parent, confirming that D8S503 is present on the aberrant chromosome and therefore must be located in the single copy region. Patients 1, 4, and 5, were informative at D8S503 for the inheritance of a single paternal allele and at least one maternal allele, confirming that D8S503 is present on the aberrant chromosome. Raw genotyping data for D8S503 is presented in figure 8 and summarized in table 3. Genotyping results reported by Floridia et a/.(1996) for D8S201, D8S349, D8S252, D8S265, D8S552 and D8S135, for these 9 patients are presented in table 3 to allow comparisons to be made to patients 10-14 reported in section 3.12.  Chapter 3: Results  1 3 4 i-CM-, r- TS - , | - CI —| j- MM—, PJ - , 11 1 2 2 2 3 3 3 4 4 4 5 5 5 P F M P F M P F M P F M P F M r  ?  S A - , r-TG-n 0 6 6 6 7 7 7 2 P F M P F r  GL-, 8 8 8 M P F M n  r  A M 9 9 9 P F M n  Figure 8: Genotyping of 9 families with STS polymorphic marker D8S503. Numbers represent patient number, initials are those of the patient, P=proband, F=father, M=mofher. 134702 is a CEPH family member included as a positive control. The autoradiograph is a composite of two different exposure times.  fct  U CO  <  u ca u  <  Ov  CQ CJ  —1  a < u  u  2  <  CQ  o  < <  CQ  CQ U  < CQ UCQ  U CQ CQ <  Q CQ  a CQ <<  —1  a  CQ  CQ  o  CQ C J < CQ CQ O  CJ  a  <  CQ C J CQ CQ CQ CQ CQ CQ < < <  <  a  CJ CQ  < <  < CQ  UCQ <  u  CQ Q~ U CQ <  <  t/3  Q U  00  <  co  < CQ U  C J < CQ  < <  CQ  CQ  CJ  uo CQ CQ  CQ u < CJ< CQ < <  CQ <  OH  U CQ <  <C CQ CQ <  Q CQ  CQ  a <c_ Iu <  CJ  I  U CQ  <  >  < <  U ffl CQ < CQ <  CQ CQ C J < < CQ  CQ  CQ <  I  CJ  U <_  u  Q <  CQ  u  Q U  CJ < Q U  U CQ  OO  CJ  H  00  u 2  a  CQ  CQ <  CQ < U CQ  cn  CJ  < eS  CQ < CQ  -*-»  CS  <<  cd  cd  "ed  es  ^  C l D . C l f t C H C i D . C i D . o o o o o o o o o o o o o o o o o o  c  o  T3  < < CQ  CJ  o  CJ CQ  "a,  <  Q  M  C»OOOOOOCNCN<NCN CH  00  OH'  -H r-H CN CJ CJ CN CJ OH 0 0 0 . 0 OH OH CH OOOOOOOOOOOOOOOOOO  2  CO w  §  oo oo oo oo oo  H  O  2  -H  cN  ro  -sj-  Q  P.  T t t t 00 00—. 00 OO t CN CN CN ^ H tc tc rt: at CN  O  Q  L-  H-»  CJ cn co cn ro ro N CN CNCN CN oi CCL. Ci. Ci. O. cn ro co cn CH  CQ  Q  ed  H - t ^ H - » H - ' H H H H 4 _ » + J  uo vo io co Q  es  ed  p C C C c CJ CJ CJ  '3D  CO CN (N uo -H 00 00 00 GO 00  Q  ed  cd  rororororororororo CNCNCNCNCNCNCNCNCN  uu  0\ N m  Q  cd  H-»  a c CJ  ^  uo 00 00  Q  cd  +H  cd  C J C J C J C J C J C J C J C J C J  +-»  Q CQ  <  < CQ  cd  UH  o o o o o o o o o o o o o o o o o o  < Q CQ CQ  CQ  cd  E E E E E E S E E  5 CQ 9 < <«  < CQ  cd  H.  CQ CQ  « 3  CS  C C C C c u E CJ CJ u CJ CJ CJ  co  o  cc!  <  CQ < <  u s  CQ  CQ < CQ  CO  Q CQ  <  CQ CQ <  u CQ  <  < o  J  2  co H O <C  uo vb  oo  o  Chapter 3: Results  6 8  3.2 Refinement of the Map of the Region Predicted to Contain the Distal Element For inv dup(8p) patients with a cytogenetic center of symmetry in band 8p23.1, STS genotyping confirms that in 6/6 informative patients, D8S503 is located in the region of single copy at the cytogenetic center of symmetry (table 3). Genotyping also confirms the deletion of D8S349 from the inv dup(8p) chromosome in 11/11 informative patients (Floridia et al., 1996). The location of the distal repetitive element predicted by the hypothesis of a mechanism of formation mediated by inverted repeats, can then be refined to the region between marker D8S349 and D8S503. Currently this region is represented by a gap in the database of Y A C singly and doubly-linked STS contigs maintained by the MIT/Whitehead Institute for Genome Research (http://www.genome.wi.mit.edu). This gap is located between contigs WC8-0 (distal) and WC8-1 (proximal). In order to investigate this region at the molecular level, this contig gap must be spanned. Mapping efforts were focused on this goal. The MIT/Whitehead Institute data base was used to identify Y A C s for additional STS marker testing. The information in this database is collected by a 'pooled clone' PCR approach. Consequently, positive data must be confirmed and negative data is not reported. Thirteen Y A C s from contigs WC8-0 and WC8-1, comprising the proximal and distal ends respectively, were chosen for analysis (YACs 4-16, table 4). With the exception of 856d8, 843el, 889b7, and, 966b5, the Y A C s from this region were known to be chimeric, containing segments of D N A from more than one region of the genome. Chimeric Y A C s are useful for STS content contig assembly but are not useful for chromosome walking. STS content analysis, by PCR, of Y A C s 4-6 chosen from WC8-0, was carried out for markers D8S439, D8S277, D8S349, D8S349 (the same (CA) as D8S1819), and D8S1935. STS content analysis by PCR, of Y A C s 7-16 chosen n  Chapter 3: Results  69  from W C 8 1 , was carried out for markers D8S252, D8S503, D8S1825, D8S574, D8S516 and D8S542. A l l primer sequences and annealing temperatures are listed in table 2. A comparison of the MIT/Whitehead Institute STS content mapping, and results obtained in the chosen subset of Y A C s tested, is presented in table 5 and discussed below. STS content data collected for this thesis is summarized in figure 9. A n example of the raw PCR data is presented in figure 10.  3.2.1 Y A C STS Content Not Reported by MIT/Whitehead Institute For the region of WC8-0 studied all MIT/Whitehead positive results for the STS markers D8S277, D8S1819 and D8S1935 were confirmed. Y A C 764c7 was extended by the presence of markers D8S277 and D8S439. Y A C s 967cl 1 and 856d8 were found to contain marker D8S439. For the region of WC8-1 studied, almost all MIT/Whitehead positive results for the STS markers tested were confirmed (see section 3.2.2). Y A C s 966b5 and 889b7 were found to contain D8S1825. Y A C s 742dl2 and 843el were found to contain D8S542. Y A C s 845a9, 849h4, and 785d6 were found to contain D8S503. Y A C 845a9 was found to contain D8S516. Markers D8S252, D8S574, and D8S351, not analysed by the MIT/Whitehead Institute, were also tested for their presence in these Y A C s . Marker D8S252 is present in Y A C s 785d6, 742dl2, and 889b7. Marker D8S574 is present in Y A C s 849h4, 871a8, 742dl2, and 966b5. Marker D8S351 is present in Y A C s 785d6, 966b5, and 843el.  3.2.2 Negative YAC STS content Not Predicted by MIT/Whitehead Institute With the exception of 966b5, which did not contain D8S516 when tested in our lab, all results were in agreement with those predicted by MIT/Whitehead. This negative result could reflect a false positive in the MIT/Whitehead data as the frequency of false positives in these  Chapter 3: Results  Table 4: Y A C information. The subset of Y A C s investigated in this thesis have been placed on chromosome 8 by STS content (MIT/Whitehead Institute), n.k. = size is not known. "-"= Y A C is not known to be chimeric.  YAC 1. 810f8 2. 741h4 3. 787cll 4. 967c11 5. 856d8 6. 764c7 7. 918b5 8. 849h4 9. 871a8 10. 776f4 11. 785d6 12.742dl2 13.966b5 14.889b7 15. 845a9 16. 843e 1 17.915h4 18. 773g4 19. 799b 1  SIZE(kb) chimeric with chromosome: 1180 n.k. 1740 1110 550 1370 1320 1380 500 1740 1320 n.k. 1680 n.k. 600 580 740 1570 480  9 14 17 17 2 2 11 15 1,10 2 2 17 -  C/3  H  CD C/3  D8S516 D8S542  J3  o  U  CJ  Cl, CJ tH  o  CJ  o (=1 CD 00  cd  CJ  on cd  <D  fi.  .C  „ U £ < cd CD  oo CD  i  i  +  i  +  i  •  * +  *  i  *  •  H—»  00  D8S574  oo TJ  * +  D8S351  T3  *  IS  tH  I  i  #  I  •  CD  rH  CD  Cl  o  O  a <D o  CD  *^  CJ CJ  cd  t-H  _  * +  * +  S " CJ -2 + M a C CL,  *  I  * +  * +  #  D8S252  o a  +  *  *  * +  #  +  #  *  I  i  i  16. 843el  CU  +  15. 845a9  c . H o g co o  12. 742dl2  c/3  11. 785d6  H  D8S503 D8S1825  CD  =  cT C/3  H  Cl  13  o  «  .2 S to  «<  rrj  -s:  <D  s-  cd  00  ,  CD  fH  .5 ^3  *  •e  CD  o • CD  3  •S +3  CL, CD  tn  CD  1=1  O  cd  <  CD  cd  +  * +  +  * +  * +  *  #  *  #  c3 oo C/3 "cd rH  1  H  C/3  H  cd  YAC  s  —H  II  14. 889b7  CJ  O  *  o  13.966b5  * Ji  CD PH  •  i  10. 776f4  a CD  a CD oo  +  9. 871a8  o -i->  00 CD  8. 849h4  a o bO jjn  CD CJ  7. 918b5  CD  6. 764c7  a  5. 856d8  CD  4. 967cll  o  D8S1935  tH  D8S349  C/3  CJ  D8S439 D8S277  O  Chapter 3: Results  72  o  V  a1 IS  SH  -5 CO  ta g Is *r1  <-i  S3  CJ  S520  s3 T 3  g  O  o  00  O  co rtt  S542 S516 S351 S574 SI 825 S503 S252  O  CJ  Oi  H—» • - H  -3 -13  cd  03  CO  CO  Vi  £3  00 S3  o  CJ  •e5 g CJ3 UH  H—  m  CJ  a  CS  CJ  ~  -i? CJ  2  >H CJ  X)  t H  CJ  H »- CJ fcj c UH CS 2 ts s <4H "O O CO S cj H O +H CO  SI 935 S349  •a  o oo O  S277  • • «  S439  ©0 ©  o o  CS CJ CJ & C O CS C/5 O co o a H  > §  00 S3  C cS 'BH O  cS  *H->  CJ  "  J3 CJ  3 .3  & oo -t->  S518  "S £ SH  CO CJ  00  o o  CS CD CJ sX  CO H  CO  £ ts 12 ^  o  tr cs  Os  u  I  CJ  o  00 i n • H -  CJ  PH  Vi  S3 3^ CJ  Chapter 3: Results  73  Chapter 3: Results  74  data is known to be 6% (Hudson et al., 1995). However, given that large Y A C s are prone to deletion, the possibility also exists that a deletion has occurred in our clone that wasn't present in the C E P H mega-YAC pools. STS content analysis did not lead to closure of the gap between WC8.0 and WC8.1. A variety of approaches were undertaken to isolate clones spanning this region but, were unsuccessful (data not shown). Efforts were turned to analysis of a candidate repetitive element.  3.3 Isolation of an 8p Repetitive Element A novel repetitive element, termed a megasatellite, was isolated from 4pl5 and localized to 4pl5 and 8p23 by Southern blot analysis of single chromosome hybrids and by FISH analysis (Gondo et al., 1996, Kogi et al., 1997). The 4p element is a 4.7 kb EcoRI fragment, tandemly repeated 12-90 times in the Japanese population (Gondo et al., 1996, Kogi et al., 1997). The localization of this element to distal 8p, and the low copy number in the genome, made it a candidate for the proposed repetitive element involved in the formation of inv dup(8p) chromosomes. Chromosome 8 megasatellite clones were isolated and their organization on the chromosome investigated. The 4753 bp sequence of Gondo et al., 1996 (Genbank accession #: D38378) was used as a query sequence in a BLASTsts search of the Genbank STS database (http://www.ncbi.nlm.nih.gov). STS markers rsAVA13 (accession #U57857), STS4-310 (accession #L00791), SHGC4-672 (accession #G01777), SHGC4-1030 (accession #G01800), SHGC4-1135 (accession #G01806), SHGC4-1596 (accession #G01912), SHGC4-1739 (accession #G02001), SHGC4-1436 (accession #G01841), and SHGC4-1656 (accession #G05153) were recovered. Homology of these markers to the chromosome 4 megasatellite  Chapter 3: Results  75  sequence are shown in figure 11. rsAVA13 was derived from a chromosome 8 telomere half Y A C , the others, from chromosome 4 random clones (see table 2 for references).  3.3.1 Isolation of Megasatellite Containing Cosmids PCR products, amplified from the STSs spanning the megasatellite, were pooled and used as hybridization probes to the LA08NC01 cosmid library gridded on high density filters (See table 2 for STS markers, primer sequences and annealing temperatures. Primer pairs are available from Research Genetics, Inc.). Initially, only a single cosmid, 153G8, was isolated.  3.3.2 Characterization of Cosmid 153G8 Megasatellite STS PCR products were hybridized to EcoRI digested cosmid 153G8. Five EcoRI fragments; 4.7 kb, 3.35 kb, 2.4 kb, 2.3 kb, and 2.1 kb contain sequences which crosshybridize with the megasatellite sequence. Fragments are identified by cosmid name (in this case, 153G8), enzyme used to create the fragment (in this case, .EcoRI (E)), and the fragment size in kb. Therefore, these fragments would be identified as 153G8E4.7, 153G8E3.35, 153G8E2.4, and so on. STS P C R products were individually hybridized to £coRI digests of 153G8: 153G8E4.7 hybridized with all of the products, 153G8E3.35 hybridized only with the 3' products, 153G8E2.1 hybridized only with the 5' products, and 153G8E2.3/153G8E2.4 (not resolvable in this figure) hybridized with all of the products (figure 12). The 4p megasatellite sequence was analyzed for motifs, that with a single base pair change, would create an .EcoRI site. A mutation of G—»T at 2332bp would create an £coRI recognition site (GAATTC) resulting in two fragments, one of 2.3 kb and one of 2.4 kb. This would explain the apparent hybridization of all P C R products to 153G8E2.3/153G8E2.4.  Chapter 3: Results  76  The EcoRI blots were hybridized with y-end labeled T7 and T3 primers (see table 2 for sequences). The T3 end of 153G8 is contained within the 4.9 kb fragment and the T7 end within the 0.5 kb fragment. These fragments do not correspond to any of the fragments crosshybridizing with the megasatellite sequences. Therefore, the megasatellite sequences must lie internally in cosmid 153G8. As well, there are at most, 3 complete copies of the 4.7 kb EcoRI unit in this region of the genome. This is significantly different from chromosome 4, where the megasatellite is present in 12-90 copies. Cosmid 153G8 was digested with BamHI and EcoRI, blotted, and hybridized with the megasatellite STS PCR products. At the resolution obtained by digestion with BamHI, all EcoRI fragments containing the megasatellite, with the exception of 153G8E3.35, cross hybridized with the megasatellite PCR products. The 153G8E3.35 fragment is cleaved, by digestion with BamHI, into 153G8BE1.1 and 153G8BE2.2, and only 153G8BE2.2 hybridized with the megasatellite PCR products (data not shown). A restriction enzyme map of 153G8 was assembled using the enzymes, EcoRI, BamHI, Bglll, BssHll, and Kpnl. This map is shown in figure 13 and shows the location of the megasatellite sequences relative to the ends of 153G8. Throughout this thesis, references to megasatellite or flanking sequences with regard to 5' or 3' are made relative to this restriction map.  3.3.3 Characterization of Cosmid 39A7 A second screen of the cosmid library, using the pooled PCR products, isolated cosmid 39A7. Digestion of 39A7 and hybridization with the pooled megasatellite PCR products identified 5 .EcoRI fragments identical in size to the fragments for 153G8 with the exception that the 3.35 kb fragment of 153G8 is 3.8 kb in 39A7 (figure 14). Digestion of 39A7E3.8 with BamHI results  Chapter 3: Results  77  o o o  uo  to  o o in  G  "3-  U  o o o CD  O O  m o o o  m  o o  ID CN  o o o  OO  CN  o o  uo  o o o  o o in  o —  1  m vo cn ro ov un ON  £ cn  I  CN t~-  vo I  vo  vo  vo .—i  t  O  un cn cn I  I  O ,—c I  O  *-H  ^S* *^J" ^ rn  CJ CJ O _ _ O U U O l  oooo  ID ffi K ffi  oooS  a e 00 &0 00 t/> ffi ffi ffi H CZ)  (Z)  CZ) IZ)  Chapter 3: Results  Figure 12: Hybridization of megasatellite PCR products to cosmid 153G8. lane l ^ c o R I digested 153G8 banding pattern, lanes 2-10 hybridization of EcoRI digested 153G8 with P C R products: 2. rsAVA13, 3. SHGC4-1739, 4. SHGC4-1135, 5. SHGC4-1596, 6. SHGC4-672, 7. SHGC4-1656, 8. SHGC4-1436, 9. SHGC4-1030, 10. SHGC4-310.  79  Chapter 3: Results  60  CQ  60 CQ UJ oo CQ CQ  3 o II  bO CQ UJ  00  cu td on cd OX)  CQ  CD  UJ '  c/i  cd  CQ ' oo  C tH CU  00 . CQ  on  co o  oo  o  00 . CQ  IT) T3  on O  CO  O  UJ  00  o o UJ  on O , °  - H  S ll  00  a  CQ  ^  o H  ]  CQ  00 CQ CQ  i4 CQ UJ  r  J3  cu  iii  II  H  .S3P ftOX) fa  PQ  80  Chapter 3: Results U  u cd  •r, CD  C a o  •fa  13  vS  CQ  CD  & ts OO  cD  u .SP  < "3 ro  CQ  O  L  T  .2  H  ^  &:§  a * CD -73 <D g 43 CO  5-3 CD 4=  r-  CO  I—I  ID  taq  CQ ^ s ro  ON  <J r -  *°  c  ^  CO  co  T3 B3  00  O  ra  CD  00 co w>  o  'cd  3 < CD  a  cd  cd  o  c-3 CD  «j a  5  in 3 CCS CD 00  O >-  8 B  (D C cd CD c i &0 cj CD O *+H O  •x3 •^  j§ 0  cd  g  O  CD CD CD  "JB >  «8 § N 43  73  co  u >>iS  S , § 1  .SP  a a  g x  Chapter 3: Results  81  in cleavage into 39A7BE1.6 and 39A7BE2.2. The megasatellite PCR products hybridize to 39A7BE2.2 but not 39A7BE1.6. The T7 primer hybridizes to the 3.5 kb EcoRI fragment and the T3 primer to the 0.8 kb EcoRI fragment, indicating that the megasatellite must lie internally within this cosmid (data not shown). Therefore, this cosmid contains approximately three complete copies of the 4.7 kb unit. The cosmid 39A7 megasatellite can be differentiated from the cosmid 153G8 megasatellite by a difference in a single EcoRI fragment, 3.8 kb and 3.35 kb respectively.  3.3.4 Screening Cosmid Library with 4.7 kb Megasatellite Sequence A final screen of the cosmid library, using the 4.7 kb megasatellite fragment isolated from both 153G8 and 39A7 rather than the pooled PCR products, led to the isolation of 6 additional cosmids; 116B9, 113F7, 66A11, 108F1, 128H8, 172D7. In order to make comparisons to 153G8 and 39A7, these cosmids were digested with EcoRI and hybridized with the 4.7 kb megasatellite fragment. Cosmid 116B9 has a 3.35 kb fragment, 66A11 and 108F1 has a 3.8 kb fragment, 113F7 has too little of the megasatellite sequences to determine this fragment size, and 128H8 and 172D7 have a novel 1.8 kb fragment (figure 14). The LA08NC01 cosmid library was created from a single chromosome 8, therefore, differences in restriction fragment pattern must reflect the presence of at least 3 types of the megasatellite on chromosome 8, rather than polymorphism.  3.3.5 Restriction Fragments Characteristic of Type I, II, and III Megasatellite In order to characterize each megasatellite type, comparative restriction enzyme digestions were carried out on the cosmids. Restriction enzymes were chosen based on analysis of the  Chapter 3: Results  82  sequenced 4p megasatellite. Restriction enzymes were used in both single and double digestions, the digested D N A was transferred to nylon membrane, and hybridized with the 4.7 kb megasatellite fragment (results not shown). Figure 14 shows an example of these experiments. Ultimately, characteristic restriction fragment sizes were determined for each megasatellite type (table 6). Type I megasatellite is characterized by the presence of a 3.35 kb EcoRI fragment, presence of a 4.7 kb BssHll fragment, and of a single, 19 kb Ndel fragment. Type II megasatellite is characterized by the presence of a 3.8 kb EcoRI fragment, absence of a 4.7 kb ifosHII fragment, and presence of two Ndel fragments, one large and one 6.5 kb in size. Type III megasatellite was initially characterized only by the presence of a 1.8 kb EcoRI fragment, as both cosmids containing this type did not contain a complete 4.7 kb fragment, and therefore, appeared to terminate within the megasatellite. However, later experiments with Y A C clones indicate that both the 1.8 kb and 4.0 kb fragment are characteristic of the type III megasatellite. Cosmids 153G8 and 116B9 contain type I, 39A7, 108F1, 66A11 contain type II, 172D7 and 128H8 contain type III and, 113F7 contains type I or type II (figure 15).  3.4 Subcloning of EcoRI Fragments from Cosmids 153G8 and 39A7  The megasatellites found in cosmids 153G8 and 39A7 are internal. These cosmids were chosen for subcloning in order to isolate fragments flanking the megasatellites for analysis. A number of fragments from each cosmid were subcloned (table 7). The fragments were liberated from the cloning vector, and labeled for use as individual hybridization probes to .EcoRI digests of all of the cosmids. Results of these hybridizations are summarized in table 7. The sequence flanking all three types of megasatellite cross hybridized, including the 4.9 kb and 0.5 kb end  83  Chapter 3: Results  Table 6: Restriction enzyme fragment sizes characteristic of the 8p megasatellite sequences. The type I megasatellite sequences are distinguished by a 3.35 EcoRI fragment and a 4.7 kb iks-HII fragment. The type II megasatellite sequences are distinguished by a 3.8 kb EcoRI fragment and a 6.5 kb Ndel fragment. The type III megasatellite sequences are distinguished by a 4.0 kb and a 1.8 kb EcoRI fragment.  Megasatellite Type I Type II Type III  EcoRI fragments Nde I fragments 3.35 kb 3.8 kb 4.0 kb, 1.8 kb  1 large 1 large, 6.5 kb 1 large  BssHlI fragments 2 large, 4.7 kb 2 large 2 large  84  Chapter 3: Results  a  a.  oo  co  w 0 0  fN  fN ON  m  ON  NO  oo  3, 0 0  v©  C3 60 CD  E  oo Cv3  OX)  CJ  N  a  3  00  CJ  CJ  a 0(  85  Chapter 3: Results  Table 7: Hybridization of randomly cloned EcoRI fragments to megasatellite containing cosmids. Fragments were randomly cloned from cosmids 153G8 and 39A7. Megasatellite containing cosmids were digested with EcoRI, transferred to nylon membrane, and hybridized with: a. EcoRI fragments of cosmid 153G8. b. EcoRI fragments of cosmid 39A7. A l l fragment sizes are in kilobases. N H = no hybridization to this cosmid.  a. 153G8 JScoRI hybridization fragment 4.9  3.3  Hybridizes Hybridizes Hybridizes Hybridizes Hybridizes Hybridizes Hybridizes Hybridizes  3.25 2.7 1.6 0.9 2.7 1.6 N H N H 1.6 0.8 2.65 1.6 3.25 2.65 1.6 3.25 2.65 N H N H N H 2.9 N H N H 2.9  to 153G8 EcoRI fragment: to 116B9 EcoRI fragment: to 113F7 EcoRI fragment: to 39A7 EcoRI fragment: to 108F1 EcoRI fragment: to 66A11 EcoRI fragment: to 172D7 EcoRI fragment: to 128H8 EcoRI fragment:  4.9 0.9 NH NH 2.8 6.5 NH NH  2.7  1.6 2.6  2.6 3 2.6 3 2.6 3 2.6 3 2.6 3 NH NH 2.6 9.3 2.6 9.3  b. 39A7 .EcoRI hybridization fragment: 1.6  1.2  2.6  Hybridizes Hybridizes Hybridizes Hybridizes Hybridizes Hybridizes Hybridizes Hybridizes  1.2 1.2 1.2 1.2 1.2 NH 2.9 2.9  2.6 3 2.6 3 2.6 3 2.6 3 2.6 3 NH NH 2.6 9.3 2.6 9.3  to 153G8 EcoRI fragment: to 116B9 EcoRI fragment: to 113F7 EcoRI fragment: to 39A7 EcoRI fragment: to 108F1 EcoRI fragment: to 66A11 EcoRI fragment: to 172D7 EcoRI fragment: to 128H8 EcoRI fragment:  1.6 1.6 1.6 1.6 1.6  NH 2.9 2.9  3  3  2.8 NH 2.8 2.8 2.8 NH NH 2.8 2.8  0.5 0.5 4.9 4.9 4.9 NH NH 9.3 9.3  Chapter 3: Results  86  fragments of 153G8. A pictorial representation of this is shown in figure 15 where regions of cross hybridization are represented as overlapping although the cosmids are not believed to truly overlap in the genome. Cosmid inserts containing each type of megasatellite were labeled and each type was hybridized back to cosmids and a B A C containing a megasatellite (figure 16) (see below for data on the B A C ) . Some differences in restriction fragment size are apparent, however, cross hybridization was demonstrated across at least 40 kb. Therefore, it appears as though each class of megasatellite is embedded within a large reiterated sequence (LRS).  3.5 Sequencing of Clones 153G8E3.35 and 39A7E3.8  The megasatellites of 153G8 and 39A7 differ in their 5' EcoRI restriction fragments. These differences may offer clues to the evolution of the megasatellite sequences on chromosome 8. The 5' EcoRI restriction fragments, 153G8E3.35 and 39A7E3.8, are both cleaved by BamHI into two fragments. In each case only one of the fragments, 153G8BE2.2 and 39A7BE2.2 respectively, hybridize with the megasatellite sequences. The non-hybridizing fragment, 153G8BE1.1 and 39A7BE1.6, may contain novel sequences, useful as hybridization probes to study these specific megasatellite types in inv dup(8p) patient. In order to sequence in both directions from the BamHI site located within each EcoRI fragment, deletion clones 153G8BE1.1, 153G8BE2.2, 39A7BE1.6, and 39A7BE2.2 were constructed by the procedure outlined in figure 17. The approach chosen for construction depended on in which orientation(s) the fragment had been cloned into the vector. Clones are labeled as forward or reverse to allow clones of different orientation to be distinguished from one another. Sequencing from both ends of 153G8E3.35 and 39A7E3.8 and in both directions from the internal BamHI site in each fragment was carried out using an automated sequencer (ABI  Chapter 3: Results  88  Chapter 3: Results  89  Model 373). The regions sequenced are depicted in figure 18. Approximately 500 bp of sequence was obtained in the direction of each arrowhead. B L A S T searches and Clustal W alignment were used to compare the sequences. Alignments were made between corresponding subclone sequences and the 4p megasatellite (Appendix 1). In other words, sequence from the EcoRI site of 153G8BE2.2 was compared to sequence from the EcoRI site of 39A7BE2.2 and to the 4p megasatellite, and so on. The sequences read from the EcoRI sites of 153G8BE2.2 and 39A7BE2.2, and from the BamHI sites of these fragments, respectively, have a high degree of homology to each other and to the 4p megasatellite. 153G8BE1.1 and 39A7BE1.6 also have a high degree of homology with each other and some homology with the megasatellite, when read from both the BamHI and EcoRI site (figure 18). This is surprising because these fragments do not hybridize with the megasatellite sequence. It is unclear whether a single 500 bp insertion/deletion, or i f insertions/deletions of one or two nucleotides at multiple locations throughout the fragments, lead to the 500 bp difference in fragment size. From the number of single base pair deletions/insertions in the sequence alignments, the latter may be more likely.  3.6 Isolation of BACs Containing Megasatellite Sequences Since cosmids containing all three types of megasatellite cross-hybridized an attempt was made to isolate larger B A C clones that might contain unique flanking sequences for each type of megasatellite. The chromosome 8 enriched B A C library was screened with the 4.7 kb megasatellite fragment. B A C 87B23 was isolated, subjected to the restriction enzyme digestions used to characterize the megasatellite, and found to contain a type I repeat. B A C 87B23 has been sized using PFGE and the insert is approximately 165 kb (data not shown). B A C 87B23 contains the STS markers D8S1619, D8S1993, and D8S1640. These 3 markers have been placed  Chapter 3: Results  90  on the chromosome 8 map, by Y A C contig STS content mapping at the MIT/Whitehead Institute, to the region between D8S265 and D8S552 (figure 19). Attempts to isolate additional megasatellite containing clones from the chromosome 8 enriched B A C library were unsuccessful. A n attempt was made to isolate megasatellite containing B A C clones from the Research Genetics B A C library (Sambrook et al., 1989) from which the enriched chromosome 8 B A C library was made. Hybridization of the 4.7 kb megasatellite fragment to this library would result in isolation of a preponderance of clones containing chromosome 4pl5 sequences. Therefore, the Research Genetic, Inc. library was screened with the T7 and SP6 termini of B A C 87B23. A n extremely large number of positive signals, estimated to be in excess of 3,000, were obtained. Only strong signals were chosen, and still over 100 positive clones were obtained. To decrease the number of clones for analysis, only those B A C s contained in the 8p enriched B A C library were chosen (data listed in Appendix 2). None of these B A C s contained a megasatellite. In order to address the question of why the megasatellite may be underrepresented in the chromosome 8 B A C library, an estimate of the amount of highly repetitive D N A in the region surrounding the megasatellite was made. Total human D N A , labeled by random prime labeling, was hybridized to EcoRI digested cosmid 153G8 and B A C 87B23 (figure 20). One band in 153G8, containing the T7 end, hybridized to the probe. Therefore, the region including and immediately flanking the megasatellite contains none of the highly repetitive D N A , such as Alu and LINES, found in high copy number in the genome. A number of bands in 87B23 hybridized to the probe. The enrichment of the B A C library for chromosome 8 sequences was carried out by hybridization of Inter-Alu PCR products to the total B A C library, therefore a deficit of these sequences in the region immediately surrounding the megasatellite may explain the failure to  Chapter 3: Results  91  isolate B A C s containing the megasatellite. However, the region flanking the D N A contained within the cosmid may be better represented in the library, as evidenced by the large number of clones isolated when screening the Research Genetics B A C library. Isolation of B A C 87B23 allowed placement of the type I megasatellite, based on STS content, on the Y A C contig physical map. However, large clones containing the type II and type III megasatellites were not isolated. Therefore, Y A C s were investigated for their megasatellite content in order to place the remaining megasatellites on the 8p map.  3.7 Megasatellite Content in Chromosome 8 YACs A minimum tiling path in Y A C s , across the short arm of chromosome 8, was assembled by examination of the MIT/Whitehead Institute database. FISH analysis confines the chromosome 8 megasatellites to distal 8p. Therefore, Y A C s spanning from the telomere region to marker D8S552 (located in 8p22) (figure 19) were examined for their megasatellite content using the characteristic restriction enzyme analysis and hybridization with the 4.7 kb megasatellite fragment (figure 21). Three Y A C s were found to contain the megasatellite. Y A C 967c 11, contains a type II megasatellite, Y A C 764c7 (chimeric for chromosome 8) contains a type III megasatellite and Y A C 773g4 both a type II and type III megasatellite. No Y A C s containing the type I megasatellite were recovered. Isolation of Y A C s containing the entire type III megasatellite allowed the characteristic bands to be modified to include a 4.0 kb .EcoRI fragment. As well, Ndel digestion results in a single large band and BssHll digestion results in two large bands (not resolved in figure 21) and no 4.7 kb band (table 6). Four other Y A C s (737e5, 809h8, 871f3, 920dl2), known to contain the markers D8S1619, D8S1993 and D8S1640 found in B A C 87B23 (which contains a type I megasatellite) were analyzed for megasatellite sequences (data  92  Chapter 3: Results  1.1  153G8E3.35 I  1.6  2.2  2.2  39A7E3.8 l _  Figure 18: Schematic representation of sequenced fragments. 153G8E3.35 and 39A7E3.8 are the 5' flanking fragments of the type I and type II megasatellites respectively. Sequences are represented by arrowheads, pointing the direction of sequencing. Sequences are: a. 153G8BE1.1-M13R, b. 153G8BE1.1-M13F+153G8BE2.2-M13F, c. 153G8BE2.2-M13R, A . 39A7BE1.6-M13R, B . 39A7BE1.6-M13F+39A7BE2.2-M13F, C. 39A7BE2.2-M13R. A l l sequences share some homology with the 4p megasatellite sequence. Sequence a is highly homologous to A , sequence b is highly homologous to B , sequence c is highly homologous to C (see Appendix 1).  93  Chapter 3: Results  ,  ^ 3  j=  in  O  .£  co .2?  0 0  "S  „ §  2 o § q s £ ^  5  a.  3 55 ^  S552 S1619 SI 993 SI 640  t  SI 695 00  U  S265 S520 S1721  =3  ^  .a  S351  S503 S252  4)  SI 935 S349  S277 O ©  oo CO  o oo U  S439 tj  ^  S518 00  cd  *  S262 S201 S264  CD a  S7  4-S  <*>  M  «  ox)  g  — tS «  J  J  a B CD  H  t  9 x> 12 S3  cu a> ~\ p p *  ^ v • ® ® fa cd  Chapter 3: Results  r  EcoWX -  i 5 3 G 8  8 7 B 2 3  1 5 3 G 8  1  1  8 7 B 2 3  Figure 20: Hybridization of total human D N A to clones containing the megasatellite. Cosmid 153G8 and B A C 87B23 were digested with EcoRI. The EcoRI restriction fragment pattern is shown on the left, hybridization of total human D N A is shown on the right.  Chapter 3: Results  95  r—  ON  ON ~  m  as  so  NO  J=  JD  03  —i  ON J D  Tr  +H  CO CD  I  E  L0  O  in  «*3 <H_  CJ  BE)  OO  ON NO  A  or—  £ -2  t  CJ  to  i— o ~H ^H  r-~ ON ON  ON  TJ-  OS J 3  so  "  m  tr5 00 — <U HS  JD  _00  ^  00  TJ-  ON  ci  —  f£ oo  V~>  <  TJ-  J 3  —  c NO  r-- o CJ  [  CJ J 3  in  0 eg 00  Tf  ON J=  t  9  —  O  a a  o o CS N  i f . £ 4 3  IO W ^ M t-H  o '  a  CJ  in —  1— 00 CU c o E •- o moi ^ m ni o m  O  1  L  Cd  .  £  s  SS  O O H O H W  CD cd bH  CJ  co CJ  60 CJ  &o g; U oo  +H CS  IS  bH  =3  cs  CO CS  CJ  s  I g  r-  60 33 CJ _e C+H  2 -S O  S  q 43 . a co  o ~H ^H — js  1  CN  S-l C J CJ  -4-t CJ  so as m -o — — JD  Cd  60 CL)  ON so p- o -H -H  r~  +•»  >s cd < -S 43 § CO CO 2 cj cj q  JO J *  M  _N 'C cj  E  W  r>  OJj  cd +3 4= •  CS  O r-<  pj o o (A —.  i  .3  o;  C+H O  m  P3  « U  <  1  T  (NsO-st U1N ON NO  CJ  -T3 CC JJ ^<U .13 cs CJ cS co 43 co cS fcS 00 33 00 C J 0  ON so NO JD  CO CJ  ^  & o  ON ON JD '—i  Os — m  cd to  >s  43  s 3 ^5  cj o ON  3  CJ oo CJ  JD  7  T3  js Tf  ON SO NO  03 C3 BC CD  a 3  I  I  f-;  00  I  m  n  I  ^H CN  60  ^  60  "C co  o cs 43 a § 40  £  2  96  Chapter 3: Results  not shown), no type I megasatellite was recovered in these Y A C s . The MIT/Whitehead database maps Y A C 967cl 1 and Y A C 773g4 to two different locations on chromosome 8, based on STS content (figure 19), lending further evidence for the existence of more than one location of the megasatellite on 8p. The MIT/Whitehead Institute database reports a 6% false positive rate in their data (Hudson et al., 1995). Therefore, the STS content of these Y A C s must be confirmed.  3.8 STS Content Analysis of Megasatellite Containing Clones STS content analysis of clones containing megasatellite sequences is important for accurately placing the megasatellites on chromosome 8p. PCR analysis has been carried out in our laboratory on the Y A C clones for the STS markers D8S1640, D8S1993, D8S1619, D8S1819, and D8S1935 (with the exception of D8S1819 and D8S1935 data presented in section 3.2, data not shown). However, analysis of these clones, and the B A C 87B23, by hybridization of the PCR products of these markers more accurately confirms STS content than does PCR. A partial PCR product of D8S1819 (same (CA) as D8S349), constructed by digestion of the n  PCR product with Haelll to remove the ( C A ) , was used as a hybridization probe to clones n  containing the megasatellite; B A C 87B23 (type I), and the Y A C s 967c 11 (type II), 764c7 (type III), and 773g4 (types II and III). Its presence in Y A C s 967c 11 and 764c7, and its absence in B A C 87B23 and Y A C 773g4, was confirmed (figure 22a). The PCR product of D8S1935 was used as a hybridization probe to the same clones. Its presence in Y A C 967c 11, and absence in B A C 87B23 and Y A C s 764c7 and 773g4, was confirmed (figure 22b). B A C 87B23 contains the markers D8S1640, D8S1993, D8S1619, and a type I megasatellite. The content of these STS markers in Y A C s containing megasatellite sequences was investigated. The PCR product of each of the three markers (primer sequences and annealing temperatures  Chapter 3: Results  97  listed in table 2) was hybridized to EcoRI and Hindlll restriction enzyme digests of clones containing the megasatellite; Y A C s 764c7 (type III), 967cl 1 (type II), and 773g4 (type II and type III), and B A C 87B23 (type I) (figure 23). Y A C s 764c7 and 773g4 contain D8S1640, and 773g4 contains D8S1993. Y A C 967cl 1 does not contain the three markers. The fragment sizes seen in the Y A C s were consistent with those seen in 87B23. Unexpectedly, D8S1640 hybridizes to two Hindlll bands in Y A C 764c7. As well, D8S1993 hybridizes to two EcoRI bands in Y A C 773g4. However, the sequences reported for the PCR products D8S1640 and D8S1993 do not contain either an EcoRI or a Hindlll restriction enzyme cleavage site. These data suggest that either these sequences are present more than once in the clones or, a restriction enzyme recognition site has been introduced into the sequence. Regardless of the correct interpretation, these data suggest that large reiterated sequence may extend to include the markers D8S1640, and D8S1993. Investigations were undertaken to determine the size of the LRS in which each class of megasatellite is embedded.  3.9 Analysis of the LRS on Chromosome 8p Additional experiments were carried out on the B A C 87B23 and the Y A C s 967c 11, 764c7, 773g4, and 799b 1 in order to determine the size of the LRS. Y A C 799b 1, which does not contain a megasatellite, was included in the experiments because it contains D8S1640, D8S1993, D8S1619, and D8S552, placing it proximal to the megasatellite on the Y A C contig map (figure 19). B A C 87B23 contains the markers D8S1640 and D8S1993, which may be included in the LRS. In order to compare homology between Y A C s , hybridization probes were prepared from the insert of B A C 87B23, isolated from a pulsed field gel, and the SP6 and T7 ends of 87B23,  Chapter 3: Results  98  isolated by bubble PCR. B A C 87B23 and Y A C s 967cl 1, 764c7, 773g4, and 799M were digested with Sstl, separated on an agarose gel, transferred to nylon membrane and hybridized with the insert of B A C 87B23, the T7 end of 87B23, and the SP6 end of 87B23, respectively. The results of these experiments are presented in figures 24 and 25, and summarized here. Y A C 799b 1 does not hybridize with either end of 87B23 (data not shown) but cross hybridizes with the insert of 87B23. However, the majority of these bands are not the same size as those present in the other Y A C s (figure 24, bands are faint, all ~6 kb and larger). Y A C 764c7, a known chimera, contains the SP6 end of 87B23 in a 9.7 kb Sstl fragment, and contains a number of Sstl fragments that cross hybridize with the insert of 87B23 with similar in size to those in Y A C s 967cl 1 and 773g4. Most surprisingly, Y A C s 773g4 and 967cl 1 have almost identical hybridization patterns when hybridized with the insert of 87B23 (figure 24), and this pattern differs from the Sstl pattern of B A C 87B23. These Y A C s map to different regions of 8p using STS content (figure 19), therefore, the similarity in hybridization pattern must represent homology of these regions and not true overlap. The T7 end of 87B23 hybridizes to two Sstl fragments in 773g4, one 20 kb, the other just slightly larger than 20 kb (figure 25). Y A C 967c 11 does not hybridize with either the T7 or the SP6 end of 87B23. The similarity in the hybridization pattern of Y A C s 967c 11 and 773g4 using B A C 87B23 for hybridization could mean that the LRS is not entirely contained within B A C 87B23. Consequently, the ends of B A C 87B23 were used to isolate walk clones.  Chapter 3: Results  a. D8S1819 probe  r  Hindlll  Sstl  kb  20-  b. D8S1935 probe r  EcoRI —|  8 7  9 6 7  2 3  c  7 6 4  1 1  c  g  7  4  B kb  •  I— ss/i — |  7 7 3  kb  ' *  12  11-  Figure 22: D8S1819 and D8S1935 content analysis by hybridization. With the exception of B A C 87B23, all clones are YACs. D8S1819 is a polymorphic STS marker containing a (CA)n. The (CA)n has been removed by digestion and recovery of flanking sequence. D8S1935 is a non-polymorphic STS marker, a. Clones digested with Hindlll, or Sstl, and hybridized with the D8S1819 probe, b. Clones digested with EcoRI, or Sstl, and hybridized with the D8S1935 probe.  100  Chapter 3: Results  09 CJ  i—i H^H  i-H  vO  t-~ E> ps oj)Tf  53 MO Q o 0 (NI  jjj OS VO P - O  l-H —I  oo t^- CQ r- p»  CN  CO  m  u  oc^f  PS  r» vo  u  o p-  2 as vo r- o  ~-  oo r- CQ  CN  ^  m  o  -O  M  O  he xcept ion of BA 7B: iem t. Clo ne were digest ed \ vith Ec ofD 93, c. the P proidue  <+H  •a  IL  8S1  clo  a  j_  OS  (N  Qs  CJ  PS ^  co oo  J=  +H  Ip  or-  N  i OS C".  c/3 O .o  L  r-i m ^  p- P- ro  b/)Tf  p- vo  as  • I  >s  HQ  u  o p-  TP  < VOP~ O  oo c- CO  <  l-H l-H .  j CN P V  HO -4  r  3o0 SJ  (  I  p- vo  P S OOTf Tf  o ^ as vo p- cj vo L_ oo p~ CQ c/3 L 00 Q  ca  CN P I  o p»  m  >>  CN PS  Tf  CJ  —  I)  vS  TJ  eg  CJ  Hj"J  4)  0  o •— M  .CJi  < CN  -H  ©  _o  CO  t+H O  —  Ofj  Q  sis ck  oo p- pq  c o o  H C/J  o p—i  JO  <  I*^  p- p~ P S ofiTf  os vo r- o  "C  c/3  CJ  Tf  CJ N  u  O  S  p- vo  ft  ofY and sated t left of ea chh he PC pro duct of >8S1 een de fron auto ad  r  ^  » r- cq  o —  CJ tH  M CC  4—1  JO  d  CJ  . •—  ha  >3q as vo r- cj —• —  L  Sb  CJ E  hy bridi dDNA fra he PCR raph  4—1  M C i t  E  i•f-»: >  GO CJ  •aCJ g N .5  Cs. An and brid: eous  [  p- IS P S bOTf  d  .5? <  c^ H  X  W  Figure 24: Analysis of homology between YACs by hybridization of B A C 87B23. Clones are digested with Sstl. With the exception of 87B23, all clones are YACs. The autoradiograph on the left is a composite of two different exposure times.  Chapter 3: Results  a. 87B23 SP6 end  b. 87B23 T7 end  Sstl  8 7 B 2 3  9 6 7 c 1 1  Sstl  7 6 4 c 7  7 7 3 g 4  8 7 B 2 3  9 6 7 c 1 1  7 6 4 c 7  7 7 3 g 4  20  9.0-  Figure 25: Analysis of YACs by hybridization of B A C 87B23 T7 and SP6 ends. Clones were digested with Sstl and hybridized with a. 87B23 SP6 end, b. 87B23 T7 end. Extraneous lanes have been deleted from the autoradiographs.  Chapter 3: Results  103  3.9.2 Chromosome Walking from B A C 87B23  The cosmid and enriched chromosome 8 B A C library were screened with the T7 and SP6 ends of 87B23 to isolate clones for hybridization against the Y A C s . The T7 end of 87B23 hybridized to cosmids 113F7, 172D7 and 128H8 (figure 26c) that were previously isolated (figure 15, 16). Cosmids 172D7 and 128H8 contain a type III megasatellite, cosmid 113F7 contains either a type I or a type II megasatellite. A l l three cosmids contain 3' sequences flanking the megasatellite, but no 5' flanking sequences. As well, only ~9 kb of D N A is common to these 3 cosmids, but not common to the other cosmids containing megasatellite sequences (figure 15). Therefore, the T7 end of 87B23 is located within the LRS 3' region flanking the megasatellite, and does not represent the true end point of the LRS. The distance from the T7 end of 87B23 to the megasatellite is estimated from figure 15 to be approximately 20 kb. The SP6 end of 87B23 hybridized to cosmid 128H10. This cosmid was not investigated further because the T7 end was anchored within 20 kb of the megasatellite. The T7 end hybridized to two BACs, while the SP6 end hybridized to four BACs, isolated from the 8p enriched B A C library. Three of these BACs, 223B23, 67B6 and 286D2, were chosen for further analysis (figure 27). A l l three have Sstl patterns that differ from the Sstl pattern of 87B23 (figure 27a). BACs 67B6 and 223B23 hybridized to the T7 end of 87B23, B A C 286D2 to the SP6 end of 87B2. B A C 223B23, almost completely cross hybridizes with B A C 87B23. B A C 67B6 cross hybridizes to B A C 87B23, however, by summing the sizes of the restrictions fragments that do not cross hybridize, B A C 67B6 is estimated to extend 20 kb past the T7 end of B A C 87B23 (figure 27a, b). B A C 286D2 cross hybridizes to B A C 87B23,  Chapter 3: Results  104  however, by summing the sizes of the restriction fragments that do not cross hybridize, B A C 286D2 is estimated to extend 80 kb past the SP6 end of 87B23 (figure 27 a, b). The human D N A inserts of 223B23, 67B6, and a partial insert of 286D2 including the SP6 end of 87B23 and the ~80 kb region extends past B A C 87B23, were isolated by PFGE. Their respective sizes are 130 kb, 97 kb, and 80 kb (data not shown). Each insert was used as a hybridization probe to Sstl digests of B A C 87B23, and Y A C s 764c7, 967c 11, and 773 g4 (figure 28) and compared to the hybridization patterns of B A C 87B23 (figure 24). B A C 223B23 hybridizes to all but the 4.3 kb fragment hybridized by 87B23 and does not hybridize to any new fragments. B A C 67B6 does not hybridize to novel bands in the Y A C s . The partial insert of B A C 286D2 hybridizes to the fragment of 87B23 containing the SP6 end but does not hybridize to the Y A C s . This suggests that the LRS ends near the SP6 end of B A C 87B23 and may end just past the T7 end of 87B23. The insert of B A C 223B23 is almost completely homologous with B A C 87B23, including the T7 end of 87B23 which is estimated to be within -20 kb of the megasatellite, but the Sstl restriction fragment patterns differ (figure 27a). These data suggest that B A C 223B23 may contain a megasatellite differing from the type I megasatellite in B A C 87B23.  3.10 Lack of Megasatellite in BACs 223B23, 67B6, and 286D2  B A C 223B23 was analyzed for its megasatellite content by digesting with restriction enzymes EcoRI, Ndel, and ifosHII hybridizing with the 4.7 kb megasatellite fragment (data not shown). B A C 223B23 does not contain a megasatellite. As well, hybridization of the 4.7 kb megasatellite fragment to Sstl digests of BACs, including 223B23, 67B6 and 286D2, showed that none of the BACs homologous with 87B23 contain megasatellite sequences (figure 29). However, 223B23  Chapter 3: Results  105  has almost complete homology to 87B23 (only the 4.3 kb fragment, containing megasatellite sequences, does not hybridize with 223B23), is 130 kb in size, and overlaps with the T7 end of 87B23, which, based on cosmid data, is approximately 20 kb away from the megasatellite. The lack of a megasatellite in BAC223B23, and the isolation of more than 3,000 B A C clones in the library screen of the Research Genetics, Inc. B A C library with the T7 and SP6 ends of 87B23, raises the question of whether these sequences originate from 8p. This was addressed by FISH.  3.11 Mapping the LRS by FISH using BAC 223B23 B A C 223B23 is estimated to contain approximately 110 kb of the 5' flanking sequence and approximately 20 kb of the 3' flanking sequence of the megasatellite (figure 27), and is therefore a useful FISH probe for the LRS. FISH of the biotin-labeled insert of 223B23 to normal control chromosomes resulted in a strong hybridization signal at 8p23, analogous to the location of the megasatellite, and weaker signals on many other chromosomes (figure 30).  3.12 Characterization of Five Additional Patients A patient with what appeared to be a standard inv dup(8p) chromosome, three patients with unusual inv dup(8p) chromosomes, and the patient of Dhooge et al. (1994) were analyzed by STS genotyping and FISH analysis.  3.12.1 Polymorphic STS Genotyping in Five Patients STS genotyping at D8S201, D8S349, D8S503, D8S265, D8S552, and D8S135 (figure 3, table 2), was carried out on these five patients (patients 10-14) and their parents (or sibling, in the case of patient 14) to allow comparison between these patients, and previously reported inv dup(8p) patients. Markers D8S201 and D8S349 are located to the region of 8p deleted from the  Chapter 3: Results  a. insert 87B23  b. SP6 end 87B23 Sstl  c. T7 end 87B23  Sstl  Sstl  l  kb  20  kb  20  kb  2 8 H  1 2 H 1 0  1  1  1  7 2  7 2  7 6  D 7  G  C  3  2  20 14  9.4  4.3 3.8  2.8 2.3  Figure 26: Analysis of cosmid clones isolated by chromosome walking from B A C 87B23. The T7 and SP6 ends of B A C 87B23 were used to screen the chromosome 8 specific cosmid library. Cosmid clones were digested with Sstl and hybridized with: a. the insert of 87B23, b. the SP6 end of 87B23, c. the T7 end of 87B23.  107  Chapter 3: Results  m  m  CN  CQ Foo '-i-H  o CO  —  CN  <  o  O  — CO  NO  -5  a £ CD  C/3  4=  C+H  cs d  NO OH  ca  oo  c  <u  FX  g c  CS  4H  o  t-- oo p—i  - 1  Co  3 cs  CD  05  «  cs  VI  CD  43  .2 cS  £ 1.a Td f  O  CM  Vi  —  T3  rn -  in —I  CN  CD  •<  CQ PQ  £; ^  ^  S3  CD O CD U CD oo co " X> < cd CO. CD 0  r  z  oo r- CQ  ca £ =  CN  O  (NOOvOQfN  iH Cw  (N e> CQ  CN  m  .£  L  £  cS  ,  cf  Z  co  co  CN  CD  CQ  9  OS  co  <  CD  CS  CD J3  >  o  CD > cn  £ 2 ° X o  ON ^  t~ vd  co  C  E  o  r9  •  t > CQ 4D T3 CD ts (N  CN  o  1M CD CO  .a CD  •B  T J C4CD c  -S  £  m r) — CC o •r-oo  CS  u  CD co 00  b  N  E "C c o  o <  co O  £ 42 o —  CQ 4CD= 4=  §  4 ? CD  >—H e  C+H  O  CD  CO  CS ti  H °  CD '-+H  O  45  fi C o cs N  < s  'C  CN  X3  1 0  CO T 3  CD =3 3 CD  ^H  43  H-» 1 / 5  43  a  "«HJ  CD CS  00  53  r  2  Chapter 3: Results  108  -7-S c« .N  £  • "5 MX)  a b  o  5  D. 03 CJ «o 5  £  d  O  ° _ ,tu o^ & E >  8  ^  X3  O  °1£  r  Jo o o £ O, oo on C CJ  .-H  r- so  TT  co « ° cj S •u  o r-  CJ  U XL c  SO  CQ  H  03  o so t> o •—• •-<' oo r*^ CQ fN coj  CO  <  u  ffl  2  o CJ « 5  ooa .2  S^ I  X!  b c *o t; c« 03  X  (N  8S  x  o 2 o <  " 3  ^  £ CQ " oo CQ  ^ H "° u <d  > mS C/3  C  i>.2 § <  O  -  & 8 8 CQ  00  (~o  cj  ^  Chapter 3: Results  kb  6 7 B 6  2 2 3 B 2 3  •Sstl 7  6 D 2  B 2 3  f  20 -  9.4  4.3 3.8  2.8 2.3  Figure 29: Analysis of megasatellite content of BACs isolated by chromosome walking. Clones were digested with I and hybridized with 153G8E4.7 and 39A7E4.7 which contain the complete megasatellite sequence.  Chapter 3: Results  110  Figure 30: FISH analysis with B A C 223B23. The insert of B A C 223B23 and a chromosome 8 specific centromere probe were hybridized to normal metaphase chromosomes. Hybridization occurred to 8p23 (large arrowheads), at the location analogous to the hybridization location of the megasatellite, as well as to number of other locations throughout the genome.  Chapter 3: Results  ill  inv dup(8p) chromosome. D8S349 is deleted in 11/11 informative patients (Floridia et a l , 1996) and represents the most proximal marker in the region deleted from the inv dup(8p) chromosome. Markers D8S503 and D8S265 are localized to the single copy region at the center of symmetry of inv dup(8p) chromosomes (Floridia et al., 1996, and table 3). D8S552 is duplicated on the inv dup(8p) chromosome in 9/9 informative patients and delimits the distal extent of duplication in those patients who have a single copy region (Floridia et al., 1996). Marker D8S135 is located in band 8pl 1.22 and may be present in one or two copies on the inv dup(8p) chromosome, depending on the extent of chromosomal D N A included in the duplication. Examples of raw genotyping data are presented in figure 31. Genotyping results are summarized in table 8. Patient information, not reported elsewhere, and a discussion of genotyping results, is presented below. Interpretations of genotyping results were made in the following way: In all cases, at a given locus, a single allele was considered to be present on the normal chromosome of the patient. Then, based on how informative the genotyping was, the number of alleles on the aberrant chromosome could be interpreted as: zero (deleted), one (single copy), two (duplicated), zero/one, one/two or uninformative. Inheritance of a single allele at a given locus was considered evidence that the marker was deleted from the aberrant chromosome and that the chromosome was derived from the parent whose allele was not inherited. The inheritance of three alleles at a given locus was considered evidence that the marker was duplicated on the aberrant chromosome and that the chromosome was derived from the parent from whom two alleles had been inherited. The inheritance of two alleles from one parent was considered evidence that the aberrant chromosome was derived from an interchromosomal event.  Chapter 3: Results  a. D8S349  10(TP)  r  P  I4(MGP) M  F  11 (MD)  P  b. D8S503  10(TP)  14(MGP)  11 (MD)  M  12(DC) M  13(SW) M  P  M  c. D8S552  10(TP)  14(MGP) M  ll(MD)  12(DC) M  13(SW) M  1 3 4 7 0 2  l M  1 3 4 7 0 2  Figure 31: Examples of genotyping of patients 10-14. a. STS polymorphic marker D8S349, b. STS polymorphic marker D8S503, c. STS polymorphic marker D8S552. Patient numbers and initials are shown above each triad. P=patient, F=father, M=mother, S=sister, 134702 is a CEPH family member included as a positive control. Extraneous lanes have been deleted from the autoradiographs.  113  Chapter 3: Results  Table 8: Summary of polymorphic STS genotyping data for patients 10-14. Markers are listed in order from telomere to centromere on the normal chromosome. Within each triad, allele sizes are reported as A - D , A being smallest, D being largest. Genotypes are presented in the order patient, father, mother or, for patient 14, patient, father, sister. The dashed line indicates failure of amplification of alleles.  D8S201  IOTP  14MGP  B  AC AC AB AC C A B BC AC BC C AC BC AC BC BC BC AB AB AB BC C AB A AB ABC AB A A A A  B  D8S349 C D8S503 A B C D8S265 B C D D8S552 B C D8S135 A  BC BC AD AB AB  11 M D BC  12 DC A  A D A B CD  AC AB AB BD AC BC AB A B AC B AB B CD A B A C A B A A B AB  13 SW BC  AC ABC A ABC AB B AC AB BC A B  AC  B  AC BC AC AB B AB —  ...  A C BC  Chapter 3: Results  114  3.12.1 Patient 10 (TP) Patient 10 (TP) was born to karyotypically normal parents. As a child she presented with developmental delay, abnormal facies, and behavioural problems. At 29, she is severely mentally retarded, has a large mouth, large skull and upturned nose, and has a vocabulary of 10 words. Karyotype analysis revealed an inverted duplication of 8p: 46, X X , inv dup(8)(pl2->p23.1) de novo. Genotyping at distal markers D8S201 and D8S349 is consistent with lack of maternal alleles on the inv dup(8p) chromosome (table 8). Genotyping at markers D8S503 and D8S265 is consistent with inheritance of two different maternal alleles at these loci (table 8). Genotyping at D8S552 is consistent with inheritance of at least one maternal allele but is uninformative for copy number (table 8). It likely D8S552 is duplicated since markers distal to this locus are duplicated and the karyotype is consistent with duplication of the region to which D8S552 has been localized. A single maternal allele is inherited at D8S135 indicating that the duplication does not extend to include this locus. The absence of maternal alleles at D8S201 and D8S349, and the inheritance of two different maternal alleles at D8S503 and D8S265, is consistent with the inversion duplication being derived from a maternal interchromosomal event. With this interpretation, genotyping is consistent with paternal alleles being inherited in one copy at all informative markers.  3.12.2 Patient 11 (MD) Patient 11 (MD) presented with developmental delay, hypotonia, and an unusual face at approximately 1 year. She is 3 years old, at the time of this report. Cytogenetic analysis suggests that M D carries a classic inv dup(8p) chromosome with a center of symmetry at 8p23.1,  Chapter 3: Results  115  but the with the aberrant 8p telomere capped by satellites. Parental chromosomes are normal. A number of experiments were carried out in a diagnostic laboratory: FISH using a whole chromosome 8 paint showed that the entire aberrant chromosome, except the satellites, originated from chromosome 8; the distal p arm of the aberrant chromosome stains with silver staining specific to the nucleolar organizer region (NOR); FISH using the acrocentric a-satellite centromere probes (13, 14,15, 21, 22 (ONCOR)) showed no evidence of these sequences near the telomere of the aberrant chromosome. Genotyping of this patient at the above mentioned STS markers was consistent with a classic inv dup(8p) patient (table 8). Lack of inheritance of maternal alleles is seen at D8S349 and D8S201. At least one maternal allele is inherited at D8S503, and, at D8S265. Two different maternal alleles have been inherited at D8S552. D8S135 is uninformative. These data are consistent with maternal origin as a result of an interchromosomal event. The origin of the satellites is unknown but they are presumably derived from maternal chromosomes. With the interpretation that the aberrant chromosome is of maternal origin, a single paternal allele is inherited at all informative markers.  3.12.3 Patient 12 (DC) Patient 12 (DC), who is 6 years of age at the time of report, was investigated due to hypotonia as an infant. She has the mental capacity of a 3-4 year old, speaks 50 words, can walk, and is still in diapers. She has a large skull and ears, and congenital heart anomaly. Her karyotype is inv dup(8)(pl2->p23.?2). She has a large G band located at the center of symmetry. Parental chromosomes were normal. FISH analysis carried out in a diagnostic laboratory with a whole chromosome 8 paint indicates that the duplicated material is from chromosome 8. Genotyping at D8S201 and D8S349 is informative showing the inheritance of a single maternal  Chapter 3: Results  116  allele and a single paternal allele (table 8). These STS markers are normally deleted on the inv dup(8p) chromosome. However, the presence of these markers on the aberrant chromosome is consistent with the cytogenetic evidence for the presence of band p23.2. Markers D8S503, D8S265 and D8S552 are informative for the inheritance of at least one maternal and one paternal allele, but not for copy number. D8S135 is uninformative for copy number. From the genotyping results alone, the presence of a duplication could not be detected and the origin of the aberrant chromosome could not be determined. Whether this is a result of uninformative markers, or, whether the aberrant chromosome arose as a result of an intrachromosomal event is unknown. D C carries an inversion duplication of a different type to those previously reported.  3.12.4 Patient 13 (SW)  Patient 13 (SW) was ascertained with amenorrhea at 17 years of age. She has had some difficulties in school and is therefore thought to be mildly mentally retarded. She has a large skull but no striking dysmorphism. Analysis revealed a karyotype of inv dup(8)(p21.3->p23.?2). There is a G band at the center of symmetry of the aberrant chromosome. However, this band is not as large as that seen in patient DC. Parental karyotypes were normal. Genotyping is consistent with the inheritance of at least one maternal allele, and a single paternal allele, at D8S201 (table 8). Three alleles are inherited at D8S349 and at D8S503. Whether the additional alleles were inherited from the mother or the father could not be determined. Parental alleles at D8S552 failed to amplify. However, the patient is heterozygous. D8S135 is informative for the inheritance a single maternal and a single paternal allele. The origin of the aberrant chromosome could not be determined.  Chapter 3: Results  117  3.12.5 Patient 14 (MGP) Patient 14 (MGP) was previously reported by Dhooge et al. (1994). The aberrant chromosome was inherited from his mother, who carries the same chromosome. FISH analysis is consistent with the extra material being of chromosome 8 origin, however, orientation and whether it originated from bands p22—»p23.1 or from p21.3—>p22, is unknown (Dhooge et al., 1994). Maternal D N A was unavailable for genotyping, however, D N A from his sister who carries the same aberrant chromosome was genotyped. Genotyping of STS markers D8S201, D8S349, D8S503, D8S265, D8S552, and D8S135 was consistent with the inheritance of at least a single maternal allele and a single paternal allele (table 8). A third allele was not present at any of the loci.  3.13 Mapping the Megasatellite to Patient Chromosomes Using FISH The hypothesis for the formation of inv dup(8p) chromosomes predicts that the location of repetitive elements will not change when compared to the normal chromosome. A cosmid, 153G8, containing the megasatellite sequences (figure 13) was chosen for FISH analysis of patient metaphase chromosomes. Cosmid 153G8 was labeled with biotin by nick translation and used as a hybridization probe, along with a chromosome 8 specific centromere probe (Oncor), to patient metaphase chromosomes (FISH analysis kindly provided by S. Jurenka). Due to the lack of highly repetitive D N A in the cosmid (see section 3.6), preannealing was not required. Patients GS, a 'classic' inversion duplication (8p) patient (Dill et ai, 1987), TP, an inversion duplication (8p) patient lacking the single copy region, and M G P , either an inversion or direct duplication patient (Dhooge et al., 1994), provided target chromosomes. As can be seen in figure 32 the  UB  "3  o  0 00  •It  d  •a a  5  S 5 g  Chapter 3: Results  119  megasatellite is present on chromosome 4p, and on the normal chromosome 8, subtelomeric, in approximately band 8p23. Only a single signal, in addition to the centromere specific signal, can be seen on the normal chromosome. The significance of seeing a single signal with FISH analysis, when evidence suggests more than one discrete location of the megasatellite sequences, will be addressed in the discussion. GS appears to be a classic inv dup(8p) patient. However, presence of a single copy region has not been confirmed. When the megasatellite sequence is used as a FISH probe to the chromosomes of patient GS, there is a maintenance of location and signal intensity on the inversion duplication chromosome when compared to the normal chromosome (figure 32a). TP appears to be a classic inv dup(8p) patient with the exception that at D8S503 and D8S265, which are normally present in single copy on the aberrant chromosome, two maternal alleles were inherited. When the same FISH probe is used against the chromosomes of patient TP there is a more diffuse signal on the aberrant chromosome when compared to the normal chromosome (figure 32b). When the FISH probe is used against the chromosomes of patient M G P , who carries an 8p duplication of unknown orientation, there are two discrete signals on the aberrant chromosome, rather than the single signal seen on the normal chromosome (figure 32c).  Chapter 4:  120  Discussion  Chapter 4: Discussion  The aim of this thesis was to gather molecular evidence supporting the hypothesis of a mechanism of formation of inversion duplication 8p chromosomes mediated by inverted repetitive elements. This mechanism predicts the location of inverted repetitive elements at the cytogenetic center of symmetry of inv dup(8p), located within 8p23.1 in the majority of patients (table 1). Should these repetitive elements be sufficiently distant from each other on the chromosome, a single copy region at the cytogenetic center of symmetry is also predicted. This single copy region was shown to exist by genotyping of 5/5 informative inv dup(8p) patients (Floridia et al., 1996). These data predicted the localization of the proposed inverted repetitive elements at the boundaries of the single copy region, distally in the region between D8S265 and D8S349 and proximally in the region between D8S552 and D8S265 (see figure 3) (Floridia et al., 1996).  4.1 Refinement of the Location of the Distal Repetitive Element Predicted by the Hypothesis  STS genotyping was carried out at marker D8S503, located by Y A C content to the region distal to D8S265 but proximal to D8S349 (figure 3), for nine of the inv dup(8p) patients previously reported by Floridia et al. (1996). Six of these patients are informative for inheritance of a single D8S503 allele on the inversion duplication chromosome. Previously, a single copy region at the center of symmetry of the inversion duplication chromosome was shown in three of these patients (Floridia et al, 1996). Therefore, the single copy region has  Chapter 4: Discussion  121  been shown to exist in a total of 8/8 informative patients (Floridia et al., 1996, and table 3). Analysis with D8S503 allowed the location of the distal repetitive element to be refined to the region flanked by D8S503 and D8S349.  4.2 Refinement of the YAC Contig Map in the Region Predicted to Contain the Distal Repetitive Element A gap between Y A C contigs WC8.0 and WC8.1 (MIT/Whitehead Institute) occurs at the region predicted to contain the distal repetitive element. Thirteen Y A C s , flanking the gap between contigs WC8.0 and WC8.1, were analysed for STS content in order to confirm the MIT/Whitehead Institute data and to refine the map in this region. With the exception that Y A C 966b5 was not found to contain marker D8S516, all STS content reported by the MIT/Whitehead Institute was confirmed in our subset of Y A C s . Additional STS content mapping carried out with markers D8S252, D8S574, and D8S351, allowed restructuring of STS marker order in WC8.1, as well as extension of WC8.1 into the gap between WC8.1 and WC8.0 (figure 9). However, this gap has not yet been spanned. Refinement of the STS marker order in WC8.1 places D8S503 proximal, but close, to D8S252, on the edge of the gap between contigs (figure 9). D8S252 is not highly polymorphic (heterozygosity = 0.27). Thus, D8S503 is the most distal highly polymorphic marker (heterozygosity = 0.74) in the region spanned by WC8.1, which encompasses the single copy region found in inv dup(8p) patients. Therefore, these data suggest that, of the available polymorphic STS markers, D8S503 is the most useful for determining the distal extent of the single copy region in inv dup(8p) patients.  Chapter 4: Discussion  122  Although clones spanning the gap between contigs WC8.0 and WC8.1 were not isolated, the STS information collected will be valuable in closing this gap using other clone sources. Efforts were then focused on analysis of a candidate repetitive element.  4.3 A Candidate Novel Repetitive Element A novel repetitive element, a megasatellite (MS), was isolated from chromosome 4 clones, and mapped to 4pl5 and 8p23 (Gondo et al., 1996, Kogi et al., 1997). The low copy number and localization to 8p23 made it a likely candidate for the proposed repeat, predicted to lie in at least two locations on 8p23. Cosmid, B A C , and Y A C clones from chromosome 8, containing the megasatellite sequence, were isolated, mapped, and the organization examined.  4.3.1 Molecular Evidence for the Existence of at Least 4 Megasatellite Locations on 8p Three different types of megasatellite, contained in cosmid clones, were isolated from the LA08NC01 chromosome 8 cosmid library (figure 15) and analyzed. This library was created from a cell line containing a single chromosome 8 (Wood et al., 1992). The presence of three distinct types of megasatellite in clones isolated from this library can only be explained by the existence of at least three discrete locations of these megasatellites on chromosome 8. Isolation of B A C and Y A C clones containing the megasatellite sequences lent further evidence for at least 3 locations of the megasatellite on chromosome 8 and enabled the placement of the megasatellites on the Y A C contig map (figure 19). Y A C 967c 11 contains STS markers D8S439, D8S227, D8S349 (D8S1819), D8S1935 (figure 19) and a type II megasatellite (table 6, figure 23). Y A C 764c7 is a chromosome 8 chimera, containing STS markers D8S439, D8S277, D8S349 (D8S1819), D8S1640 (figure 19), and a type III megasatellite (table 6, figure 23). Y A C  123  Chapter 4: Discussion  773g4 contains STS markers D8S1695, D8S1640, D8S1993 (figure 19), and a type II and type III megasatellite (table 6, figure 23).  Y A C 799b 1 contains STS markers D8S1640, D8S1993,  D8S1619, D8S552 (figure 19) and does not contain a megasatellite (figure 23). B A C 87B23 contains STS markers D8S1640, D8S1993, D8S1640 (figure 19), and a type I megasatellite. These data, combined with the knowledge that 3 different megasatellite types exist in a cosmid library constructed from a single chromosome 8, place the megasatellite in at least four discrete locations on chromosome 8.  4.3.2 The Megasatellite is Embedded within a Large Reiterated Sequence  Cross-hybridization between the sequences flanking the different megasatellites was shown by isolation of flanking D N A from type I and type II cosmids followed by hybridization to all megasatellite containing cosmids (table 7, figure 15). Therefore, each class of megasatellite must be embedded in a large reiterated sequence (LRS) and as such, the proposed 'repeat' includes the megasatellite and flanking sequences. In order to determine the size of the LRS, the insert and ends of B A C 87B23 were hybridized to Y A C s 967cl 1, 764c7, 773g4, 799M. A l l four Y A C s share homology with the insert of 87B23, but only 764c7 contains the SP6 end and only 773g4 contains the T7 end (figures 24, 25). Y A C s 967c 11 and 773g4, which map to two discrete regions of the genome, have similar hybridization patterns when the insert of 87B23 is used as a hybridization probe. However, Y A C 967c 11 failed to hybridize with either end of 87B23. These data can be interpreted in at least two ways: 1. that the LRS is completely contained within B A C 87B23, or 2. it is not completely contained within B A C 87B23 due to a variety of possibilities.  Chapter 4: Discussion  124  To distinguish between the two possibilities, clones homologous to the ends of 87B23 were isolated (figures 27). Three cosmids, known to contain different types of megasatellites, hybridized with the T7 end of 87B23 (figures 15, 27). Therefore, the LRS is not entirely contained within B A C 87B23. Two BACs, 286D2 and 67B6, extend -80 kb and -20 kb, respectively, in either direction from 87B23. However, the partial insert of 286D2 did not hybridize to any of the Y A C s . And, upon comparison to the hybridization pattern obtained with the insert of 87B23, the insert of 67B6 did not hybridize to any novel Sstl fragments in the Y A C s . One interpretation of these data is, that one endpoint of the LRS is located within 87B23 very near the SP6 end, and, the other endpoint of the LRS is located just beyond the T7 end of 87B23. From these data, it can be concluded that the LRS is at least 165 kb. A third B A C , 223B23, isolated by library screening, is apparently homologous to 87B23, although the megasatellite is absent (figure 27), and therefore, the three markers D8S1640, D8S1993, and D8S1619, are included in the LRS. The inclusion of these markers in LRS offers an explanation for the hybridization of the PCR product of D8S1993 to two EcoRI fragments of Y A C 773g4, and for the hybridization of D8S1640 to two Hindlll fragments of Y A C 764c7 (figure 23). These Y A C clones may contain two copies of these sequences. Alternatively, these clones may contain copies of these markers that differ from the sequenced PCR products and contain restriction enzyme sites such that digestion separates the sequence onto two fragments.  4.3.4 Orientation of the Megasatellite in B A C 87B23  The orientation of the megasatellite within B A C 87B23 was determined by localization of the T7 end of B A C 87B23, in cosmids that contain the megasatellite and 3' flanking sequences. B A C 87B23 contains the STS markers D8S1640, D8S1993, and D8S1619, which are ordered  Chapter 4: Discussion  125  relative to the megasatellite based on Y A C content data. Cosmids containing the T7 end of B A C 87B23 do not contain the 5' flanking region of the megasatellite, or the three STS markers. Therefore the orientation of the megasatellite within B A C 87B23, relative to the three markers, can be determined. The orientation of the type I megasatellite on chromosome 8 relative to figure 19 is: telomere - T7 end - 3'<-5'MS - D8S1640 - D8S1993 - D8S1619 - SP6 end - centromere  4.4 The Organization of the Megasatellite on 8p  There are at least 4 locations of the megasatellite on chromosome 8p. However, the organization of the megasatellites on the chromosome is unknown. It is likely that all 8p megasatellites (MS) are embedded within a large reiterated sequence that includes D8S1640, D8S1993, and D8S1619 (which will be referred to as the marker cluster when all three are present in a clone). It is also likely that the Y A C s 967c 11, 764c7 and 773g4 are not chimeric. These Y A C s contain marker NIB 1550, which is located on chromosome 17. Y A C s 967cl 1 and 764c7 contain the chromosome 17 marker WI-3139. If indeed all of these Y A C s are chimeric, it is remarkable and improbable that they would all be chimeric with the same region of chromosome 17. Thus, the inferred chimerism is likely to be an artefact due to amplification within the chromosome 8 LRS corresponding to a paralogous region on chromosome 17. Further, Y A C s may be unstable, depending on the D N A insert. The inability to isolate clones spanning the gap between the MIT/Whitehead Institute Y A C contigs WC8.0 and WC8.1 suggests that this region is unstable in Y A C s . Therefore, it seems likely that Y A C 764c7 is not chimeric for chromosome 8 sequences, but rather that a large deletion occurring in this clone has deleted D8S1935 and the type II megasatellite (figure 33), and that the type III megasatellite in  Chapter 4: Discussion  126  this Y A C clone is actually located in the Y A C contig gap between WC8.0 and WC8.1. Therefore, the type III megasatellite would be located in the distal region as well as the proximal region as shown by the presence of type II and III megasatellites on Y A C 773 g4. B A C 87B23 contains the marker cluster and a type I megasatellite. The marker cluster is placed on the Y A C contig map (figure 33) distal to D8S552. A type I megasatellite was not isolated in Y A C s .  Y A C s containing megasatellites may be unstable and undergo  rearrangements that delete the type I megasatellite, and therefore, the type I megasatellite may be located in the proximal region. Alternatively, the type I megasatellite may be located within the gap between the Y A C contigs in the distal region. Both of these explanations are supported by evidence that when the insert of this B A C is used as a hybridization probe, the Sstl hybridization pattern differs from the Sstl hybridization patterns of Y A C s 967cl 1 and 773g4 (figure 24). This suggests that B A C 87B23 may be from a different region of the chromosome than contained within these Y A C s . Analysis of Y A C content of the megasatellite and the marker cluster orients the three markers relative to the megasatellite: M S - 1640 - 1993 - 1619 (figure 19).  4.4.1 The Distal 8p Map Location of the Megasatellite  In the distal region between D.8S349 and D8S503 the hypothesis predicts the location of a repetitive element involved in the formation of inv dup(8p) chromosomes. Y A C 967c 11 contains D8S439, D8S277, D8S349, D8S1935, and a type II MS. It does not contain any of the STS markers from the marker cluster. The presence of D8S349 places this Y A C on the genetic map. Y A C 764c7 also contains D8S439, D8S277, and D8S349, but does not contain D8S1935. It does, however, contain a type III MS and the STS marker D8S1640. The content of D8S349 places this Y A C on the genetic map. Assuming, as discussed above, that this Y A C has  127  Chapter 4: Discussion ON  cu rx -a  2  e =tt .5 00  o  E o c 8  cn C ON co O N  CD  S552  <4  4 oo U  S1619 SI 993 S1640  CN-  ON ON  J  oo  m S1695  Pl cn  =3  3  S265  P. -5=1  «/•>  S520 S1721  Cvt  >  S351 vo  S503 S252  <l  _^  1  OO  is — Q w j2 o B  Tf  00  cS .Si  S1935 S349 CD  S277  o  NO ON  p oo U  S439 S518 S262  82 o  NO  S201  *i5  S264 S7  i ' i  |3  -  ^  H  3 co  -5 cu 6D  CD  CJ  >5  9i  a, a. sP S s P CD  CD  t-H  B  <|  CJ  ^  <  <H  CO  Chapter 4: Discussion  128  undergone deletion, then the type III megasatellite may be located proximal to the type II megasatellite in this region. Therefore, the organization in the distal region (figure 33), where the orientation of the bracketed region is unknown, may be: tel - 349 - 1935 - M S H - 1640 - 1993 - 1619 - (MSIII - 1640 - 1993 - 1619) - 503 - cen placing the repeat, comprised of the megasatellite and the LRS, in the distal region predicted to contain repetitive elements involved in the formation of inv dup(8p) chromosomes.  4.4.2 The Proximal 8p Map Location of the Megasatellite In the region between D8S265 and D8S552 the hypothesis predicts the location of a repetitive element involved in the formation of inv dup(8p) chromosomes. Y A C 799bl contains, the marker cluster, D8S552, but no MS. The presence of D8S552 places this Y A C on the genetic map. Y A C 773g4 contains D8S1695, D8S1640, D8S1993, a type II and a type III MS. The presence of D8S1695 places this Y A C on the physical map overlapping Y A C 915h4 which contains D8S265 and D8S1695 (figure 33). From these data , the organization in the proximal region (figure 33), where orientation of the bracketed regions is unknown, may be: tel - D8S265 - D8S1695 - (1619-1993-1640-MSII)/(MSIII)-1640-1993-1619 - D8S552 - cen placing the repeat, comprised of the megasatellite and the LRS, in the proximal region predicted to contain repetitive elements involved in the formation of inv dup(8p) chromosomes.  4.4.3 Sequences Flanking the Megasatellite are Not Chromosome 8 Specific A n attempt to isolate B A C clones containing the megasatellite, by screening of the Research Genetics B A C library with the T7 and SP6 ends of B A C 87B23, which contains an internal megasatellite, was unsuccessful due to the overwhelming number of positive clones obtained  Chapter 4:  Discussion  129  (>3,000). Attempts to chromosome walk from the ends of B A C 87B23 led to isolation of clones, such as B A C 223B23, that with the exception of the megasatellite, are apparently homologous to the LRS. These data suggested that these clones may be from elsewhere in the genome. When used as a hybridization probe for FISH analysis, B A C 223B23 hybridizes to multiple locations throughout the genome, including 8p23. The signal on 8p23 is stronger than those seen elsewhere in the genome, likely due to the localization of at least 4 copies of this sequence to this region. Therefore, the sequences flanking the megasatellite are not unique to chromosome 8, and chromosome walking, within this region, will be difficult. The size of LRS is estimated to be at least 165 kb on chromosome 8. However, it will be difficult to estimate the size as one can not be sure that clones are truly overlapping with the chromosome 8 sequences without sequencing. The hybridization of B A C 223B23 to multiple locations throughout the genome supports the assumption that the inferred chimerism of Y A C s 967c 11, 764c7, and 773g4 with chromosome 17 could be due to cross hybridization of the LRS with a paralogous region of chromosome 17. 4.5 The Evolution of 8p Megasatellite The novel 4p repetitive element, or megasatellite, reported by Kogi et al. (1997) is defined by a 4.7 kb EcoRI fragment tandemly repeated 12-90 times on a single chromosome 4. This repetitive element has been localized to chromosome 8p23, which suggests that at some time in the past, and perhaps still, the 4p megasatellite was mobile. The 8p megasatellite is comprised of approximately 3 complete copies of the 4p repetitive element, that are contained within a number of different sizes of EcoRI fragments. The 4.7 kb fragment is maintained on chromosome 4, but not on chromosome 8, suggesting that the origin of the sequence is chromosome 4p.  Chapter 4: Discussion  130  Each class of chromosome 8p megasatellite is embedded within a region of sequence that is itself reiterated throughout the genome. This suggests that the LRS was mobile before the megasatellite intruded, and may still be mobile. The LRS, including the embedded megasatellite, is present in at least four locations on chromosome 8. This suggests that the original intrusion of the megasatellite onto chromosome 8 occurred within the reiterated region, followed by dispersion of the L R S and the embedded megasatellite to an additional three locations on chromosome 8p. This dispersion may have occurred by chromosomal interactions, or by transposition of the mobile region. The conclusion that this type of dispersion occurred is supported by sequence evidence that the chromosome 8 type I and type II megasatellites are more closely related to each other than to the 4p megasatellite (Appendix 1).  4.6 Analysis of Patients by STS Genotyping and FISH of the 8 p Megasatellite Previous studies of inv dup(8p) have shown that in all investigated informative cases, there is a deletion of distal 8p (Dill et al, 1987, Barber et al, 1994, Mitchell et al, 1994, de DieSmulders et al, 1994, Guo et al, 1994, Floridia et al, 1996), and there is a region of single copy located at the cytogenetic center of symmetry (Floridia et al, 1996). As well, all inv dup(8p) chromosomes arise by a maternal interchromosomal event (Barber et al, 1994, Floridia et al, 1996). Five patients (10-14), suspected of carrying an inv dup(8p) chromosome, were genotyped for STS markers mapping to the regions deleted, present in single copy, and duplicated, on the inv dup(8p) chromosome. FISH analysis with cosmid 153G8, which contains megasatellite sequences, was carried out on patients 10 (TP), 14 (MGP) (Dhooge et al, 1994) and a previously reported inv dup(8p) patient (GS) (Dill et al, 1987, Henderson et al., 1992). If the proposed repetitive element  Chapter 4: Discussion  131  mediating the formation of inv dup(8p) chromosomes is the repeat containing the megasatellite, then, one would expect a signal on the aberrant chromosome of equal intensity, size, and location from the centromere, as on the normal chromosome (see schematic, figure 5).  4.6.1 Patient GS  Patient GS was included in the analysis as she is believed to be a classic inversion duplication patient, although parental D N A was unavailable for genotyping. Hybridization of cosmid 153G8 containing the megasatellite sequences, to metaphase chromosomes from patient GS, occurred to chromosomes 4 and to the normal and aberrant chromosomes 8. The signals on the chromosomes 8 appeared to be of equal intensity and size, and were located at what appears to be the same distance from the centromere. This analysis supports the hypothesis that the LRS containing the 8p megasatellite, is the repetitive element involved in the formation of inversion duplication(8p) chromosomes.  4.6.2 Patient 10 (TP)  The inv dup(8p) chromosome of patient 10(TP) is derived from a maternal interchromosomal event which led to deletion of distal sequences but did not result in a region of single copy, at the center of symmetry, detectable with the available STS markers (table 8). This is the first report of an inv dup(8p) patient where a single copy region was not detected at the center of symmetry of this chromosome with the currently available markers. A modification of the proposed hypothesis of a mechanism mediated by inverted repeats would explain how the aberrant chromosome of patient 10(TP) could arise (figure 34). If there are at least two repetitive elements located at the distal breakpoint that are inverted with respect  132  Chapter 4: Discussion  03  o  a  C -° o 13 O T3  SP C  H  £  03  O  co  a o  -Mr  I I  - -«  o a  J3  a.  <4  -  - -4  'co  .2 " 3  H M -  >  s  03  .TS  Chapter 4: Discussion  133  to one another, then misalignment of homologues at these sequences, followed by recombination, would result in an inv dup(8p) chromosome with a very small single copy region at the center of symmetry. This single copy region would not be detectable with the currently available STS markers. Further, the repetitive sequences involved in the misalignment would be retained at the same location relative to the centromere, as those on the normal chromosome. However, the repetitive elements in the proximal region would be duplicated, resulting in three regions containing the repetitive sequences on the inv dup(8p) chromosome. Thus, when the repetitive element is used as a FISH probe to patient metaphase chromosomes, one would expect to see a more diffuse signal on the aberrant chromosome, when compared to the normal chromosome. FISH hybridization of cosmid 153G8 to metaphase chromosomes from patient 10(TP) results in a more diffuse signal on the aberrant chromosome, compared to the normal chromosome. Therefore, FISH hybridization of the cosmid (153G8), containing the megasatellite, supports both the hypothesis that this sequence is involved in the formation of the inv dup(8p) chromosome, and the hypothesis for a modification of the proposed mechanism explained above, (figure 32a). However, it is also possible that other mechanisms, or other repetitive elements, were involved in the formation of this chromosome.  4.6.5 Patient 14 (MGP) FISH analysis confirms that the extra material on the aberrant chromosome of patient 14 (MGP) is derived from chromosome 8 (Dhooge et al, 1994). In no case were there 3 alleles present at a given marker (table 8). Therefore, there is no genotype evidence for duplication. It is possible that this chromosome originally arose as a result of an intrachromosomal event. To  Chapter 4: Discussion  134  date, inv dup(8p) chromosomes have only been shown to arise by an interchromosomal event. Karyotype analysis predicts that the duplication encompasses either bands p21.3—»p22 or p22—»p23.1 (Dhooge et al, 1994). In either case, if the duplication was inverted, FISH of the megasatellite (cosmid 153G8), would result in a single signal on the aberrant chromosome, and if direct, two signals. Two discrete signals are seen on the aberrant chromosome when the megasatellite FISH probe is hybridized to the chromosomes of M G P (figure 32c). Therefore, FISH analysis supports the conclusion that this material is directly duplicated. Classic inv dup(8p) patients are severely mentally retarded and do not reproduce. Patient M G P is not severely mentally retarded, and inherited his aberrant chromosome from his mother, who carried the same chromosome. Cytogenetic, molecular, and phenotypic differences of this case compared to inv dup(8p) cases supports the conclusion that this patient has a direct, not inverted, duplication of 8p.  4.6.3 Patient 11 (MD) The inv dup(8p) chromosome of patient 11 (MD) is inherited through a maternal interchromosomal event that led to a deletion of distal sequences. The aberrant chromosome is uninformative for the markers located to the single copy region, and has a duplication of D8S552 (table 8). Genotyping, at the markers tested, suggests that she has a classic inv dup(8p) chromosome. However, the aberrant chromosome is capped by satellites. The mechanism by which this chromosome arose is unknown. Due to the similarity to other inv dup(8p) patients we propose that the underlying mechanism is identical, with the exception that the satellites of one of the acrocentric chromosomes were used to stabilize the aberrant chromosome.  Chapter 4: Discussion  135  4.6.4 Patients 12 (DC) and 13 (SW) The aberrant chromosomes of patients 12 (DC) and 13 (SW) are unusual in that the center of symmetry is more distal than that seen in the majority of inv dup(8p) patients (table 1), and a deletion of distal sequences was not detected with the STS markers tested (table 8). The origin of the aberrant chromosome, of both patients, could not be determined from the STS markers tested. FISH analysis carried out by S. Jurenka, confirms that the extra material on both DC and SW's chromosomes is of chromosome 8 derivation, however, the mechanism by which these inv dup(8p) patients arose is unknown. FISH analysis of normal chromosome 8 with the megasatellite probe (figure 32), and the flanking reiterated sequence probe (figure 30), rules out a location of these sequences distal to 8p23.1. Therefore, the region of duplication containing the megasatellite sequences is unlikely to be involved in the formation of these unusual chromosomes. However, this does not exclude the existence of other repetitive sequences, located in 8p23.2/8p23.3, which could mediate the formation of these chromosomes. Conversely, these chromosomes may arise by a mechanism different than that leading to inv dup(8p) chromosomes with a center of symmetry in 8p23.1.  4.7 FISH as a Tool for Determining the Presence of a Single Copy Region In 1996, Floridia et al. reported evidence for the existence of a single copy region, flanked by the duplication, in inv dup(8p) patients. In 5/16 patients, genotyping was informative for the presence of a single copy region. In an additional 9/16 patients, genotyping was uninformative for copy number. Therefore, in these nine patients, the presence of a single copy region was  Chapter 4: Discussion  136  demonstrated by FISH of STS PCR products to patient metaphase chromosomes, followed by quantitative analysis of the number of signals present on the aberrant chromosome. At least four chromosome 8p megasatellites have been located to the distal and proximal regions predicted by the analysis of 8p inversion duplications. When the megasatellite is used as a FISH probe to patient chromosomes, only a single signal is visible on both the normal and aberrant chromosome suggesting that these sequences are within approximately 2-3 Mb of each other and therefore not resolvable as discrete signals. From this, it is reasonable to assume that, depending on the organization at the center of symmetry, duplicated sequences from this region may not appear as discrete signals when used as FISH probes. Therefore, FISH against metaphase chromosomes, without quantification of signal intensity, is not an accurate tool for assessing copy number of STS markers localized to the region between the distal and proximal megasatellites.  4.8 A Single Copy Region May Not be Found in Rare Cases of Inv Dup(8p)  In 1996, Floridia et al., demonstrated the existence of a single copy region located at the center of symmetry of inversion duplication (8p) chromosomes. Currently, this single copy region has been reported in 9/9 informative patients (Floridia et al., 1996, Barber et al, 1998, and table 3). However, Patient 10 (TP), reported in this thesis, has an inversion duplication of bands (pl 1.2-»p23.1) but does not have a region of single copy detectable with the currently available STS markers. D8S265 and D8S503, which lie within the single copy region, show 3 alleles in this patient indicating duplication of the region on the aberrant chromosome. The proposed mechanism could still account for this patient if there are two of the proposed repetitive elements located in close proximity and in opposite orientation near the distal boundary of the  Chapter 4: Discussion  137  single copy region (see section 4.7.2). In other words, this mechanism would apply if there are multiple copies of the proposed repeat clustering in the regions predicted to contain these repeats. A similar situation may exist on chromosome 15q in the PWS/AS region (Robinson WP, personal communication, Robinson et al, 1997). A gap in the Y A C contigs of 8p exists at the region where the additional repeat would lie, therefore, it has not been possible to determine the number of megasatellites in this region, although a re-interpretation of the organization of Y A C 764c7 leads to the conclusion that the distal region contains two megasatellites. Interestingly, cytogenetic analysis suggests that the aberrant chromosome found in patient TP does not differ from other inversion duplication chromosomes with a center of symmetry at 8p23.1. Therefore, the size of this region is less than the size detectable by analysis of metaphase chromosomes, approximately 2-3 Mb.  4.9 Genomic Organization of 8p23.1  FISH analysis of the megasatellite by hybridization of a cosmid containing megasatellite to metaphase chromosomes, resulted in a single signal on the normal chromosome 8 (figure 32). Therefore, the distance between these sequences is below the level resolvable with FISH, estimated to be approximately 2-3 megabases (Mb) (Trask, 1991). Y A C analysis places the megasatellite in at least 2 discrete regions of chromosome 8p, located at minimum 8 c M (D8S503-D8S265) and at maximum 18 c M (D8S349-D8S552) apart, based on the sex-averaged Genethon genetic map (figures 3, 19) (Dib et al, 1996, Gyapay et al, 1994). On average, 1 c M is considered to be equal to 1 Mb, however, increased recombination can inflate the genetic distance relative to the physical distance. If 1 c M equals 1 Mb in this region, one would expect these sequences to be located between 8-18 Mb apart. However, evidence from FISH analysis  Chapter 4: Discussion  138  does not support a region of this size. Therefore, the genetic distance must be inflated in this region. The most obvious explanation for this inflation is that this region is prone to recombination. Examination of the male and female genetic maps suggests that recombination in females, and not males, is increased in this region. The estimate of the size of this region from the female genetic map is at minimum 9.9 c M and at maximum 21.7 c M . Whereas the estimate from the male genetic map is at minimum 2.3 and at maximum 8.7 c M , the lower end of which is consistent with FISH analysis (figure 3). A sex-specific difference in recombination rate in this region may explain why inv dup(8p) chromosomes are of maternal origin. Perhaps some difference in female meiosis allows recombination to occur in this region more frequently than males, and therefore, should these sequences be misaligned during meiosis, recombination would be more likely to occur at these misaligned sequences in females. A modification of the proposed mechanism mediated by inverted repeats may offer a more robust explanation of how inv dup(8p) chromosomes arise. Intrachromosomal misalignment and recombination at the proposed repetitive sequences, would lead to inversion of the region between the repeats (figure 1). As discussed in the introduction, there is evidence that this type of event occurs in the factor VIII gene and leads to severe haemophilia A (Lakich et al., 1993) and also occurs in the IDS gene and leads to Hunter syndrome (Bondeson et al., 1995). Since this region is located within a single R band, a paracentric inversion of this region would not be detectable at the cytogenetic level. Should a parent carry this inversion, then, at meiosis, the normal process of events to resolve pairing problems inherent to inversions would occur. Namely, a very small inversion loop would form. In the majority of cases recombination would occur outside the loop. However, in rare instances, recombination may occur within the loop, leading to a dicentric and an acentric product, each containing a region of single copy. Breakage  Chapter 4: Discussion  139  of the dicentric at anaphase, followed by telomere repair, could lead to the formation of an inversion duplication chromosome. The center of symmetry, including the single copy region, would be consistent in unrelated patients, and the extent of duplication would vary. This mechanism is attractive in that it requires only a single aberrant event followed by normal resolution of this event and does not exclude a mechanism based on interchromosomal misalignment of these repetitive sequences. Analysis of recurrent rearrangements often begins by analysis of novel junction fragments located at the rearrangement breakpoints. Chromosomes arising by the above mechanism would not have novel junction fragments. Therefore, should this mechanism apply, novel junction fragments will not be detected in the regions flanking the single copy region. 4.10 Conclusions  4.10.1 Investigation of Megasatellite as a Candidate for the Repetitive DNA Involved in the Formation of inv dup(8p) A hypothesis for the mechanism of formation of inv dup(8p) chromosomes predicts that inverted repeats mediate the formation of these chromosomes and, that these repeats are localized to the boundaries of the single copy region found at the center of symmetry of these chromosomes. The proximal boundary of the single copy region was previously localized to the region flanked by D8S265 and D8S552 (Floridia et al, 1996). In this thesis, genotyping of nine patients with STS polymorphic marker D8S503 refines the distal boundary to the region flanked by D8S503 and D8S349. A repeat sequence composed of a megasatellite embedded within large reiterated sequence, termed an LRS, has been identified and localized on the short arm of  Chapter 4:  Discussion  140  chromosome 8 to two regions, one flanked by D8S349 and D8S503, the other flanked by D8S265 and D8S552. The organization of this repetitive D N A on chromosome 8 is unknown. However, individual cosmids derived from a single chromosome 8, and containing one of three types of megasatellites embedded within the LRS, have been found. As well, Y A C content analysis predicts 2 discrete locations of the type II megasatellite. Therefore, there at least 4 discrete locations of this repeat sequence (LRS with an embedded megasatellite) on chromosome 8. The minimum size of the repeat has been estimated at 165 kb. When used as a FISH probe, the megasatellite sequences are seen as a single signal on distal 8p, indicating a maximum separation of approximately 2-3 Mb. The genetic distance between the proximal and distal sequences is estimated at 8-18 c M , generally interpreted as 8-18 Mb. This discrepancy in distance between megasatellites would be explained by an increased amount of recombination between these sequences, and appears to be due to increase in recombination in females. Two hypotheses have been proposed for the mechanism of the formation of inv dup(8p). These mechanisms predict the presence of D N A duplicated on the normal chromosome 8p and located to the region flanked by D8S349 and D8S503, and the region flanked by D8S265 and D8S552. We propose that the repetitive D N A localized to these regions, in this thesis, is the repetitive D N A involved in the formation of inv dup(8p).  4.11 Future Research Until now, the existence of duplicated D N A in the proposed regions for the formation of the inv dup(8p) chromosomes has been only a hypothesis. The mapping of the megasatellite and flanking duplicated D N A within the regions proposed to contain important sequences for the formation of inv dup(8p) chromosomes allows the investigation of this region from a different  Chapter 4: Discussion  141  approach than previously thought possible. We now have a candidate sequence that can be studied in a number of ways to determine if it is the sequence involved. Some experiments that will need to be carried out are: 1. Determination of orientation of repeat relative to placed STS markers; 2. Closure of the gap between WC8-0 and WC8-1 and a search for a megasatellite located within this gap; 3. Determination of exact number of repeats and their organization relative to each other; 4. Examination of inv dup(8p) patients for novel junction fragments at the repeat boundaries; 5. Examination of maternal D N A for the presence of novel junction fragments that may indicate the presence of an inversion of the region between D8S349 and D8S552 or/ FISH against stretched maternal D N A to determine the orientation of sequences within this region.  Bibliography  142  Bibliography  Abeliovich D, Dagan J, Werner M , Lerer I, Shapira Y , Meiner V : Simultaneous formation of inv dup(15) and dup(15q) in a girl with developmental delay: Origin of the abnormal chromosomes. Eur. J. Hum. Genet. 3: 49-55, 1995 Abruzzo M A , Hassold TJ: Etiology of nondisjunction in humans. Environ. Mol. Mutagen 25 Suppl. 26: 38-47, 1995 Altschul SF, Gish W, Miller W, Myers EW, and Lipman DJ: Basic Local Alignment Search Tool. J. Mol. Biol. 215: 403-410, 1990 Amos-Landgraf J, Gottlieb W, Rogan PK, Nicholls RD: Chromosome breakage in Prader-Willi and Angelman Syndrome deletions may involve recombination between a repeat at the proximal and distal breakpoints. Am. J. Hum. Genet. Suppl. 55: A38, 1994 Barber JCK, James RS, Patch C, Temple IK: Protelomeric sequences are deleted in cases of short arm inverted duplication of chromosome 8. Am. J. Med. Genet. 50: 296-299, 1994 Barber JCK, Joyce C A , Collinson M N , Nicholson JC, Willatt LR, Dyson H M , Bateman MS, Green AJ, Yates JRW, Dennis NR: Duplication of 8p23.1: a cytogenetic anomaly with no established clinical significance. J. Med. Genet. 35: 491-196, 1998 Bellanne-Chantelot C, Lacroix B, Ougen P, Billault A , Beaufils S, Bertrand S, Georges I, Gilbert F, Gros I, Lucotte G, Susini L, Codani JJ, Gesnouin P, Pook S, Vaysseix G, Lu-Kuo J, Ried T, Ward D, Chumakov I, Le Paslier D, Barillot E, Cohen D: Mapping the whole human genome by fingerprinting yeast artificial chromosomes. Cell 70: 1059-1068, 1992 Birnboim HC, Doly J: A rapid alkaline extraction procedure for screening recombinant plasmid DNA. Nucl. Acids Res. 7: 1513-1523, 1979 Bodrug SE, Ray PN, Gonzalez I:. Schmickel RD, Sylvester JE, Worton RG: Molecular analysis of a constitutional X-autosome translocation in a female with muscular dystrophy. Science 237: 1620-1624, 1987 Bodrug SE, Holden JJA, Ray PN, Worton RG: Molecular analysis of X-autosome translocations in females with Duchenne muscular dystrophy. E M B O J. 10: 3931-3939, 1991 Bondeson M L , Dahl N , Malmgren H, Kleijer WJ, Tonnesen T, Carlberg B M , Pettersson U : Inversion of the IDS gene resulting from recombination with IDS-related sequences is a common cause of the Hunter syndrome. Hum. Mol. Genet. 4(4): 615-621, 1995  Bibliography  143  Boue A , Groopp A , Boue J: Cytogenetics of pregnancy wastage. Adv. Hum. Genet. 14: 1-7, 1985 Brownstein B H , Silverman GA, Little D A , Burke DT, Korsmeyer SJ, Schlessinger D, Plson M V : Isolation of single-copy human genes from a library of yeast artificial chromosome clones. Science 244: 1348-1351, 1989 Buiting K , Greger V , Brownstein B H , Mohr R M , Voiculescu I, Winterpacht A , Zabel B, Horsthemke B: A putative gene family in 15ql 1-13 and 16pl 1.2: Possible implications for Prader-Willi and Angelman syndromes. Proc. Nat. Acad. Sci. U S A : 89: 5457-5461, 1992 Chance PF, Abbas N , Lensch N W , Pentao L, Roa B B , Patel PL, Lupski JR: Two autosomal dominant neuropathies result from reciprocal D N A duplication/deletion of a region on chromosome 17. Hum. Mol. Genet. 3: 223-228, 1994 Cheng SD, Spinner N B , Zackai E H , Knoll J H M : Cytogenetic and molecular characterization of inverted duplicated chromosomes 15 from 11 patients. Am. J. Hum. Genet. 55: 753-759, 1994 Cheung SW, Sun L, Featherstone T: Molecular cytogenetic evidence to characterize breakpoint regions in Robertsonian translocations. Cytogenet. Cell Genet. 54: 97-102, 1990 Choo K H , Vissel B , Brown R, Filby RG, Earle E: Homologous alpha satellite sequences on human acrocentric chromosomes with selectivity for chromosomes 13, 14, and 21: implications for recombination between nonhomologues and Robertsonian translocations. Nucleic Acids Res. 16: 1273-1284, 1988 Choo, K H , Earle E, McQuillan C: A homologous subfamily of satellite III D N A on human chromosomes 14 and 22. Nucleic Acids Res. 18: 5641-5648, 1990 Chumakov I M , Le Gall I, Billault A , Ougen P, Soularue P, Guillou S, Rigault P, Bui H, De Tand M F , Bariloot E, Abderrahim H, Cherif D, Berger R, Le Paslier D, Cohen D: Isolation of Chromosome 21-specific yeast artificial chromosomes from a total human genome library. Nat. Genet. 1: 222-225, 1992 Clayton-Smith J, Webb T, Cheng X J , Pembrey M E , Malcolm SE: Duplication of chromosome 15 in the region 15ql 1-13 in a patient with developmental delay and ataxia with similarities to Angelman syndrome. J. Med. Genet. 30: 529-531, 1993 Cole C G , Goodfellow PN, Bobrow M , Bentley DR: Generation of novel sequence tag sites (STSs) from discrete chromosomal regions using Alu-PCR: Genomics 10: 816-826, 1991 Dausset J, Cann H , Cohen D, Lathrop M , L J M , White R: Centre d-Etude du Polymorphisme Humain (CEPH): collaborative genetic mapping of the human genome. Genomics 6: 575-577, 1990  Bibliography  144  de Die-Smulders C E M , Engelen JJM, Schrander-Stumpel C T R M , Govaerts LCP, de Vries B , Vies JSH, Wagemans A , Schijns-Fleuren S, Gillessen-Kaesbach G, Fryns JP: Inversion duplication of the short arm of chromosome 8: Clinical data on seven patients and review of the literature. Am. J. Med. Genet. 59: 369-374, 1995 Deaven L L , Van Dilla M A , Bartholdi M F , Carrano A V , Cram LS, Fuscoe JC, Gray JW, Hildebrand CE, Moyzis R K , Perlman J: Construction of human chromosome-specific D N A libraries from flow-sorted chromosomes. Cold Spring Harb. Symp. Quant. Biol. 51: 159-167, 1986 Debella L, Schertzer M , and Wood S: Identification of a Novel Human Gene (GOR) localized to 8ql3-8q22. unpublished, 1998 Dib C, Faure S, Fizames C, Samson D, Drouot n, vignal A , millasseau P, Marc s, Hazan j , Seboun E, Lathrop m, Gyapay G, Morisette J, Weissenbach J: A comprehensive genetic map of the human genome based on 5,264 microsatellites. Nature 380: 152-154, 1996 Dill FJ, Schertzer M , Sandercock J, Tischler B, Wood S: Inverted tandem duplication generates a duplication deficiency of chromosome 8p. Clin. Genet. 32: 109-113, 1987 Dhooge C, Van Roy N , Crean M , Speleman F: Direct transmission of a tandem duplication in the short arm of chromosome 8. Clin. Genet. 45: 36-39, 1994 Earle E, Shaffer L G , Kalitsis P, McQuillan C, Dale S, Choo K H A : Identification of D N A sequences flanking the breakpoint of human t(14q21q) Robertsonian translocations. Am. J. Hum. Genet. 50: 717-724, 1992 Engelen JJM, de Die-Smulders C E M , Fryns JP, Hoovers J M N , Albrechts J C M , Loots WJG, Jacobs M E , Hamers A J H : Partial trisomy and monosomy 8p due to inversion duplication. Clin. Genet. 45: 203-207, 1994 Evans GA, Lewis K A , Rothenberg B E : High efficiency vectors for cosmid microcloning and genomic analysis. Gene 79: 9-20, 1989 Feldman GL, Weiss L , Phelan M C , Schroer RJ, Van Dyke DL: Inverted duplication of 8p: ten new patients and review of the literature. Am. J. Med. Genet. 47: 482-486, 1993 Feinberg A P , Vogelstein B: A technique for radiolabeling D N A restriction endonuclease fragments to high specific activity. Anal. Biochem. 132: 6-13, 1984a Feinberg A P , Vogelstien B: Addendum: A technique for radiolabeling D N A restriction endonuclease fragments to high specific activity. Anal. Biochem. 137: 266-267, 1984b Ferguson-Smith M A : Chromosome satellite association. Lancet i: 1156-1157, 1967  Bibliography  145  Flint J, Craddock CF, Villegas A , Bentley DP. Williams HJ, Galanello R, Cao A , Wood WG, Ayyub H , Higgs DR: Healing of broken human chromosomes by the addition of telomeric repeats. Am. J. Hum. Genet. 55: 505-512, 1994 Floridia, G, Piantanida, M , Minelli A , Dellavecchia C, Bonaglia C, Rossi E, Gimelli G, Croci G, Franchi F, Gilgenkrantz S, Grammatico P, Dalpra L, Wood S, Danesino C, Zuffardi O: The same molecular mechanism at the maternal meiosis I produces mono- and di-centric 8p duplications. Am. J. Hum. Genet. 58: 785-796, 1996 Fryns JP, Kleczkowska A , Dereymaker A M , Hoefnagels M , Geremans G, Marien J, Van den Berghe H : Partial 8p trisomy due to interstitial duplication: karyotype: 46,XX, inv dup(8) (p21.1-»p22). Clin. Genet. 28: 546-549, 1985 Fryns JP, Kleczkowska A , Vogels A , van den Berghe H : Normal phenotype and slight mental retardation in de novo distal 8p deletion (8pter-8p23.1). Ann. Genet. (Paris) 32: 171-173, 1989 Fuscoe JC, Clark L M , Van Dilla M A : Construction of fifteen human chromosome-specific D N A libraries from flow-purified chromosomes. Cytogenet. Cell Genet. 43: 79-86, 1986 Giacalone JP, FranckeU: Common sequence motifs at the rearrangement sites of a constitutional X/autosome translocation and associated deletion. Am. J. Hum. Genet. 50(4): 725-741, 1992 Gobin SJ, Alcaide-Loridan C, Bono M R , Ottone C, Chumakov I, Rothstein R, Fellows M : Transfer of yeast artificial chromosome into mammalian cells and comparative study of their integrity. Gene 163(1): 27-33, 1995 Gondo Y , Hando Y , Saitoh T, Okada M , Kogi S, Fukushiga S, Ikeda JE: Megasatellite D N A : A novel 4.7 kb tandem repetitive sequence found in human 4pl5. Am. J. Hum. Genet. Suppl. 59(4): A303, 1996 Gorinati M , Caufin D, Minelli A , Memo L, Gaspardo G, Dodero A : Inv dup (8)(p21.1->22.1): further case report and a new hypothesis on the origin of the chromosome abnormality. Clin. Genet. 39: 55-59, 1991 Grunstein M , and Hogness DS: Colony hybridization: a method for the isolation of cloned DNAs that contain a specific gene. Proc. Nat. Acad. Sci. 72:3961, 1975 Guichaoua M R , Devictor M , Hartung M , Luciani JM, Stahl A: Random acrocentric bivalent associations in human pachytene spermatocytes: molecular implications in the occurrence of Robertsonian translocation. Cytogent. Cell. Genet. 42: 191-197, 1986 Guo WJ, Callif-Daley F, Zapata M C , Miller M E : Clinical and cytogenetic finding in seven cases of inverted duplication of 8p with evidence of a telomeric deletion using fluorescence in situ hybridization. Am. J. Med. Genet. 58: 230-236,1995  Bibliography  146  Gyapay G, Morissette J, Vignal A , Dib C, Fizames C, Millasseau P, Marc S, Bernardi G, Lathrop M , Weissenbach J: The 1993-94 Genethon human genetic linkage map. Nat. Genet. 7(2): 246339,1994 Han JY, Choo K H A , Shaffer L G : Molecular cytogenetic characterization of 17 rob(13ql4q) Robertsonian translocations by FISH, narrowing the region containing the breakpoints. Am. J. Hum. Genet. 55: 960-967 Henderson K G , Dill FJ, Wood S: Characterization of an inversion duplication of the short arm of chromosome 8 by fluorescent in situ hybridization. Am. J. Med. Genet. 44: 615-618, 1992 Hongell K, Knuutila S, Westermarck T: Two cases of an abnormal short arm of chromosome 8 (8p+) associated with mental retardation. Clin. Genet. 13: 237-240, 1978 Hoo JJ, Chao M , Szego K. Rauer M , Echiverri SC, Harris C: Four new cases of inverted terminal duplication: A modified hypothesis of mechanism of origin. Am. J . Med. Genet. 58: 299-304, 1995 Hudson TJ, Stein L D , Gerety SS, M a J, Castle A B , Silva J, Slonim D K , Baptista R, Kruglyak L , X u S-H, Hu X , Colbert A M E , Rosenberg c, Reeve-Daly MP, Rozen S, Hui L, Wu X , Vestergaard C, Wilson km, Bae JS, Maitra S, Ganiatsas S, Evans C A , DeAngelis M M , Ingalls K A , Nahf RW, Horton Jr. LT, Anderson M O , Collymore AJ, Ye W, Kouyoumijian V, Zemsteva IS, Tam J, Devine R, Courtney DF, Renaud MT, Nguyen H, O'Connor TJ, Fizames C, Faure S, Gyapay G, Dib C, Morissette J, Orlin JB, Birren B W , Goodman N , Weissenbach J, Hawkins TL, Foote S, Page DC, Lander ES: A n STS-based map of the human genome. Science 270: 19451954, 1995 Inoue H, Nojima H , Okayama H : High efficiency transformation of Escherichia coli with plasmids. Gene 96: 23-28, 1990 Ish-Horowicz D and Burke JF: Rapid and efficient cosmid cloning. Nuc. Acid Res. 9: 2989, 1981 Jacobs, PA: The role of chromosome abnormalities in reproductive failure. Reprod. Nutr. Dev. Suppl. L63S-74S, 1990 Jauch A , Daumer C, Lichter P, Murken J, Schroeder-Kurth T, Cremer T: Chromosomal in situ suppression hybridization of human gonosomes and autosomes and its use in clinical cytogenetics. Hum. Genet. 85: 145-150, 1990 Jensen P K A , Junien C, Despoisse S, Bernsen A , Thelle T, Friedrich U , de la Chapelle A : Inverted tandem duplication of the short arm of chromosome 8: A non-random de novo structural aberration in man. Localization of the gene for glutathione reductase in subband 8p21.1. Ann. Genet. 25(4): 207-211, 1982  Bibliography  147  Kalitsis P, Earle E, Vissel B, Shaffer L G , Choo K H A : A chromosome 13-specific human satellite I D N A subfamily with minor presence on chromosome 21: further studies of Robertsonian translocations. Genomics 16: 104-112, 1993 Kleczkowska A , Fryns JP, D'Hondt F, Jaeken J, van den Berghe H : Partial duplication 8p due to interstitial duplication: inv dup (8)(p21.1->p22). Ann. Genet. 30(1): 47-51, 1987 Kogi M , Fukushigi S, Lefevre C, Hadano S, Ikeda JE: A novel tandem repeat sequences located on human chromosome 4p: isolation and characterization. Genomics 42(2): 278-283, 1997 Lakich D, Kazazian H H Jr., Stylianos E A , Gitschier J: Inversion disrupting the factor VIII gene are a common cause of severe haemophilia A. Nat. Genet. 5: 236-241, 1993 Leana-Cox J, Jenkins L , Palmer C, Plattner R, Sheppard L, Flejter WL, Zackowski J, Tsein F, Schwartz S: Molecular cytogenetic analysis of inv dup(15) chromosomes; using probes specific for the Prader-Willi/Angelman syndrome region: clinical implications. Am. J. Hum. Genet. 54: 748-756, 1994 Lopes J, Ravise N , Vandeberghe A , Palau F, Ionasescu V , Mayer M , Levy N , Wood N , Tachi N , Bouche P, Latour P, Ruberg M , Brice A , LeGuern E: Fine mapping of de novo C M T 1 A and HNPP rearrangements within CMT1 A-REPs evidences two distinct sex-dependent mechanisms and candidate sequences involved in recombination. Hum. Mol. Genet. 7(1): 141-148, 1998 Lupski JR, de Oca-Luna R M , Slaugenhaupt S, Pentao L , Guzzetta V , Trask B, Sucedo-Cardenas O, Barker DF, Killian J M , Garcia C A , Chakravarti A , Patel PI: D N A duplication associated with Charcot-Marie-Tooth disease type 1A. Cell 66: 219-232, 1991 Malmgren H , Carlberg B M , Pettersson U , Bondeson M L : Identification of an alternative transcript from the human iduronate-2-sulfatase (IDS) gene. Genomics 29(1): 291-293, 1995 Mascari MJ, Gottlieb W, Rogan PK, Butler M G , Waller DA< Armour JAL, Jeffrey AJ, Ladda R L , Nicholls RD. The frequency of uniparental disomy in Prader-Willi syndrome. New Engl. J. Med. 326: 1599-1607, 1992 Mattei JF, Mattei M G , Ardissone JP, Coignet J, Giraud F: Clinical, enzyme, and cytogenetic investigations in three new cases of trisomy 8p. Hum. Genet. 53:315-321, 1980 Mears AJ, Duncan A M V , Budarf M L , Emanuel BS, Sellinger B, Siegel-Bartelt J, Greenberg CR, McDermid HE: Molecular characterization of the marker chromosome associated with Cat Eye syndrome. Am. J. Hum. Genet. 55: 134-142, 1994  Bibliography  148  Minelli A , Floridia G, Rossi E, Clementi M , Tenconi R, Camurri L, Bernardi F, Hoeller H, Previde Re C, Maraschio P, Wood S, Zuffardi O, Danesino C: D8S7 is consistently deleted in inverted duplications of the short arm of chromosome 8 (inv dup 8p). Hum. Genet. 92: 391-396 1993 Mitchell JJ, Vekemans M , Luscombe S, Hayden M , Weber B, Richter A , Sparkes R, Kojis T, Watters G, Der Kaloustian V M : U-type exchange in a paracentric inversion as a possible mechanism of origin of an inverted tandem duplication of chromosome 8. Am. J. Med. Genet. 49: 384-387, 1994 Morin GB: Recognition of a chromosome truncation site associated with a-thalassaemia by human telomerase. Nature 353: 454-456. 1991 Mutirangura A , Kuwano A , Robinson WP, Greenberg F, Malcolm S, Ledbetter D H . Duplication of chromosome 15q Prader-Willi and Angelman syndromes: a gene dosage paradox. Am. J. Hum. Genet. Suppl. 53: A584, 1993 Nelson D L , Ledbetter SA, Corbo L, Victoria M F , Ramirez-Solis R, Webster TD, Ledbetter DH, Caskey CT: Alu polymerase chain reaction: a method for rapid isolation of human-specific sequences from complex D N A sources. Proc. Natl. Acad. Sci. U S A 86: 6686-6690, 1989 Nevin N C , Morrison PJ, Jones J, MacC Reid M : Inverted tandem duplication of 8pl2->p23.1 in a child with increased activity of glutathione reductase. J. Med. Genet. 27: 135-136, 1990 Nicholls RD, Fischel-Ghodsian N , Higgs DR: Recombination at the human a-globin gene cluster: sequence features and topological constraints. Cell 49: 369-378, 1987 North K N , Wu B L , Cao B N , Whiteman D A H , Korf BR: C H A R G E association in a child with de novo inverted duplication (14)(q22->q24.3). Am. J. Med. Genet. 57: 610-614, 1995 Page SL, Shin JC, Han JY, Choo K H , Shaffer L G : Breakpoint diversity illustrates distinct mechanisms for Robertsonian translocation formation. Hum. Mol. Genet. 5(9): 1279-1288, 1996 Poloni L, Maraschio P, Danesino, Zuffardi O: Two cases with a de novo inverted duplication of a segment of chromosome 8. Clin. Genet. 19: 530, 1981 Raeymaekers P, Timmerman V , Nelis E, Van Hul W, De Jonghe P, Hoogendijk JE, Baas F, Bakker DF, Martin JJ, De Visser M , Bolhuis PA, Van Broeckhoven C, and the H M S N Collaborative Research Group: Duplication in chromosome 17pl 1.2 in Charcot-Marie-Tooth neuropathy type 1A (CMT1A). Neuromusc. Disord. 1: 93-97, 1991 Redha A A , Murthy DS, Kandil H , Farag TI, Usha R, al-Awadi SA, Jeryan L A , al-Nagdy K, elGhanem M : Partial trisomy of short arm of chromosome 8 (46,XY,inv dup(8)(p21^pter) in a  149  Bibliography  Bedouin child with multiple congenital anomalies and mental retardation. Indian J. Pediatr. 61(3): 301-306, 1994 Rethore M , Aurias A , Couturier J, Dutrillaux B, Prieur M , Lejeune J: Chromosome 8: trisomie complete et trisomies segmentaires. Ann. de Genet. 20: 5-11, 1977 Riley J, Butler R, Ogilvie G, Finniear R, Jenner D, Powell S, Anand R, Smith JC, Markham AF: A novel, rapid method for the isolation of terminal sequences from yeast artificial chromosome (YAC) clones. Nuc. Acids Res. 18: 2887-2890, 1990 Robinson WP, Bottani A , Yagang X , Balakrishman J, Binkert F, Machler M , Prader A , Schinzel A : Molecular, cytogenetic, and clinical investigations of Prader-Willi syndrome patients. Am. J. Hum. Genet. 49: 1219-1234, 1991 Robinson WP, Binker F, Gin R, Vazquez C, Miiller W, Rosenkranz W, Schinzel A: Clinical and molecular analysis of five inv dup(15) patients. Eur. J. Hum. Genet. 1: 37-50, 1993 Robinson WP, Horsthemke B , Leonard S, Malcolm S, Morton C, Nichols RD, Ritchie RJ, et al., Report of the third international workshop on human chromosome 15 mapping 1996. Cytogenet. Cell Genet. 76: 1-13, 1997 Robinson WP, Dutly F, Nicholls RD, Bernasconi F, Penaherrera M , Michaelis RC, Abeliovich D, Schinzel A A : The mechanisms involved in formation of deletions and duplications of 15ql 1ql3. J. Med. Genet. 35(2): 130-136, 1998 Rose M D , Winston F, Hieter P: Methods in yeast genetics - A laboratory course manual. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1990 Saitoh S, HaradaN, Jinno Y , Hashimoto K, Imaizumi K, Kuroki Y , Fukushima Y , Sugimoto T, Renedo M , Wagstaff J, Lalande M , Mutirangura A , Kuwano A , Ledbetter D, Niikawa N . : Molecular and clinical study of 61 Angelman syndrome patients. Am. J. Med. Genet. 52: 158163,1994 Sambrook J, Fritsch FF, Maniatis R: Molecular cloning: a laboratory manual, 2 ed. Cold Spring Harbor, New York: Cold Spring Harbor Laboratory Press, 1989 nd  Schmickel R D and Knoller M : Characterization and localization of the human genes for ribosomal ribonucleic acid. Pediatr. Res. 11: 929-935, 1977 Schreck RR, Breg WR, Erlanger BF, Miller OJ: Preferential derivation of abnormal human Ggroup-like chromosomes from chromosome 15. Hum. Genet. 36: 1-12, 1977 Sonnhammer E.L.L. and Durbin R: A dot-matrix program with dynamic threshold control suited for genomic D N A and protein sequence analysis. Gene 167:GC1-10, 1995  150  Bibliography  Southern E M : Detection of specific sequences among D N A fragments separated by gel electrophoresis. J. Mol. Biol. 98: 503-517, 1975 Stetten G, Charity L L , Kasch L M , Scott A F , Berman C L , Pressman E, Blakemore K J : A paternally derived inverted duplication of 7q with evidence of a telomeric deletion. Am. J. Med. Genet. 68(1): 76-81, 1997 Sullivan B A , Jenkins LS, Karson E M , Leana-Cox J, Schwartz S: Evidence for structural heterogeneity from molecular cytogenetic analysis of dicentric Robertsonian translocations. Am. J. Hum. Genet. 59: 167-175, 1996 Taylor K M , Francke U , Brown M G , George D L , Kaufhold M L : Inverted tandem ("mirror") duplications in human chromosomes: inv dup 8p, 4q, 22q. Am. J. Med. Genet. 1: 3-19, 1977 Telvi L , Ion A , Carel JC, Desguerre I, Piraud M , Boutin A M , Feingold J, Ponsot G, Fellous M , McElreavey K : A duplication of distal X p associated with hypogonadotrophic hypogonadism, hypoplastic external genitalia, mental retardation, and multiple congenital abnormalities. J. Med. Genet. 33(9): 767-771, 1996 Therman E: Human chromosomes: structure, behavior, effects, 1 ed. Springer, New York, Heidelberg, and Berlin, 1980 st  Therman E, Susman B, Denniston C: The nonrandom participation of human acrocentric chromosomes in Robertsonian translocations. Ann. Hum. Genet. 53:49-65, 1989 Thompson J, Gibson T, Thompson J., Higgins DG, Gibson TJ: C L U S T A L W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, positionspecific gap penalties and weight matrix choice. Nucleic Acids Res. 22: 4673-4680, 1994 Tomfohrde J, Wood S, Schertzer M , Wagner MJ , Wells DE, Parrish J, Sadler L A , Blantor SH, Daiger SP, Wang Z: Human chromosome 8 linkage map based on short tandem repeat polymorphisms: effect of genotyping errors. Genomics 14(1): 144-152,1992 Trask BJ: Fluorescent in situ hybridization applications to cytogenetics and gene mapping. T.I.G.S. 7(5): 149-154, 1991 Trowell HE, Nagy A , Vissel B, Choo K H A : Long-range analysis of the centromeric regions of human chromosomes 13, 14, and 21: identification of a narrow domain containing two key centromeric D N A elements. Hum. Mol. Genet. 2: 1639-1649, 1993 Tuck-Muller C M , Martinez JE, Batista D A , Kearns WG, Wertelecki W: Duplication of the short arm of the X chromosome in mother and daughter. Hum. Genet. 91(4): 395-400, 1993  Bibliography  151  Ullmann A , Jacob F, Monod J: Characterization by in vitro complementation of a peptide corresponding to an operator-proximal segment of the p-galactosidase structural gene of Escherichia coli. J. M o l . Biol. 24: 339, 1967  Urban Z, Helms C, Fekete G, Csiszar K , Bonnet D, Munnich A , Donis-Keller H , Boyd CD: 7ql 1.23 deletions in Williams syndrome arise as a consequence of unequal meiotic crossover. Am. J. Hum. Genet. 59: 958-962, 1996 Van Dyke D L : Isochromosomes and interstitial tandem direct and inverted duplications. In The Cytogenetics of Mammalian Autosomal Rearrangements. Alan R. Liss, Inc. 635-665, 1988 Verma and Babu: Human chromosomes: manual of basic techniques. Elmsford, N Y : Pergamon Press, 1989 Vocero-Akbani A , Helms C, Wang JC, Sanjurjo J, Korte-Sarfaty J, Veile R A , Liu L, Jauch A , Burgess A K , Hing A V , Holt MS, Ramachandra S, Whelan AJ, Anker R, Ahrent L L , Chen M , Gavin M R , Iannantuoni K, Morton S M , Pandit SD, Read C M , Steinbrueck T, Warlick C, Smoller D A , Donis-Keller H : Mapping human telomere regions with Y A C and P l clones: Chromosome-specific markers for 27 telomeres including 149 STSs and 24 Polymorphisms for 14 proterminal regions. Genomics 36: 492-506, 1996 Wagner M J , Ge Y , Siciliano M , Wells DE: A hybrid cell mapping panel for regional localization of probes to human chromosome 8. Genomics 10: 114-125, 1991 Wandstrat A E , Leana-Cox J, Jenkins L, Schwartz S: Molecular cytogenetic evidence for a common breakpoint in the largest inverted duplications of chromosome 15. Am. J. Hum. Genet. 62: 925-936, 1998 Warburton D: Outcome of cases of de novo structural rearrangements diagnosed at amniocentesis. Prenatal Diagn. 4: 69-80, 1987 Weleber R G , Verma RS, Kimberling WJ, Fieger H G Jr, Lubs H A : Duplication-deficiency of the short arm of chromosome 8 following artificial insemination. Ann. Genet. 19(4): 241-247, 1976 Wilson OJ, Meaney C A , Hopwood JJ, Morris CP: Sequence of the human iduronate 2-sulfatase (IDS) gene. Genomics, 17: 773-778, 1993 Wolff DJ, Schwartz S: Characterization of Robertsonian translocations by using fluorescence in situ hybridization. Am. J. Hum. Genet. 50: 174-181, 1992 Wood S, Poon R, Ridell DC, Royle NJ, Hamerton JL: A D N A marker for human chromosome 8 that detects alleles of differing sizes. Cytogenet. Cell Genet. 42: 113-118, 1986  Bibliography  152  Wood S, Schertzer M , Drabkin H, Patterson D, Longmire JL, Deaven L L : Characterization of a human chromosome 8 cosmid library constructed from flow-sorted chromosomes. Cytogenet. Cell Genet. 59: 243-237, 1992 Wood S, Schertzer M , Yaremko M L : Sequence identity located CEBPD and FGFR1 to mapped human loci within proximal 8p. Cytogenet. Cell Genet. 70: 188-191, 1995 Yen PH, L i X M , Tsai SP, Johnson C, Mohanda T, Shapiro LJ: Frequent deletions of the human X chromosome distal short arm result from recombination between low copy repetitive elements. Cell 61: 603-610, 1990 Zackowski JL, Nicholls RD, Gray B A , Bent-Williams A , Gottlieb W, Harris PJ, Waters M F , Driscoll DJ, Zori RT, Williams C A . Cytogenetic and molecular analysis in Angelman syndrome. Am. J. Med. Genet. 46: 7-11, 1993  1  Appendix 1  Appendix 1: Sequence Comparisons  E Bam L J  E  E  _L  L  b Type I  3  .  3  5  4  .  2 . 3  7  2  Bam  E  E  _ L _  _L  _L  .  4  2.1  B  Type II  3.8  4  .  7  2 . 3  2  .  4  2 . 1  Location o f sequence within the type I and type II 8p megasatellites. E = EcoRI, B a m = BamHI, EcoRI fragment sizes are in kb. Alignments were made using clustalw, with reference to B L A S T alignment searches.  a. Comparison o f sequences A, a, and the 4 p m e g a s a t e l l i t e . type I I type I 4p  TNTTATTAAAGA7AAA7AAATACGCTGTGCT7A7^ATACCAT7\ANTTCATTGACT7AATCTCAGG 3 32 TATTATTAAGGAAGAAAATACGCTGTGCTAAATACTATACTTCCATTGACTATTCTCAGG 324 TCTCACTAGGGAAGAAAATACGCTGTGCTAAATACTATACTT-CATTGACTATTCTCAGG 442 3  type I I type I 4p  TCAGAAAGCACACTTCCGATTTCTTGTCCTTCT-GTCGCTGAGAGGATGATGATAGCTGC 2 73 TCAGAAAGCGCACTTCAGACTTCTTGTCCTTCCCGTTGATGAGAGGATGACGGTAGCTGC 264 TCAGAAAGCGCACTTTCGACTTCTTGTCCTTCC - GTCGCTGAGAGGATGATGGCAGCTGC 4482 ********* ***** ** * * * * * * * * * * * * ** * * * * * * * * * * * * * ******  type I I type I 4p  CAAAAGTACATA-CTTGGAAGTTCATCCCAGCACGAGCACACACACACATAAACACACAC 214 CAAAAGTACATAACTTGGAAGTTCATCCCAGCACAAGCACACACACACA C 214 CAAAAGTACCTA-CTTGGAGGTTCATCCCAGCACAAACACACACACACA CAC 4533  type I I type I 4p  ACACACACACACACACACACACACACACACACACACAGACACACACAGGGTTTCATAGGT 154 ACACAAACACACACACACACACACACACACAGAGAGAGATACACACACGGTTTCATAGGT 154 GCCCCCCCACACACACACACAAACACACTCACACACACACACGCACACGGTTTCCTAGGT 4 5 93 * * ************** ****** ** * * * * ** **** ****** *****  type I I type I 4p  AAAGATTTCTTCCCTGACATTCTTTTACCTAAAATAAGGCAACTGTGCGGCCACTGCCCA 94 AAAGATTTCTTCCCTGACATTCTTTTACCTAAAATAAGGCAACTGTGTGGCCACTGTCCC 94 AAAGATTTCTTCCCTGCCATTGCTTTACCTAAAATAAGGCAACTGTGAGGCCGCTGTCCC 46 53  * * * **  *** * * * * * * * * * * * * * * * * * * * * * ***  * *********  *******  154  Appendix 1  type I I type I 4p  AACCCGGTTACACTCATATTATATGTGCCTATCACCCTGAGGAGTAATTTGATTCAGGTG 34 AACCCGGTTACACTCATATTATATGTGCCTATCACCCTGAGGAGTAATTTGATTCAGGTG 34 AACCCGGTTACACTCCTATTATATGTGCCTATCATCCTGAGGAGTAATTTGATTCAGGTG 4713  type I I type I 4p  TTCTAGAAGTCATGATGTGGGCTGTGTCTGTTG TTCTAGAAGTCATGATGTGGGCTGTGTCTGTTG TTCTGGAAGTCATGCTGTGGGCTGTGTCTGTTG  *************** ******************  **** *********  ************************* 1 1 4746  ******************  b. Comparison o f sequences B, b, and the 4p m e g a s a t e l l i t e .  The BamHI s i t e i s  shown i n b o l d . Alignment b l : type I I type I 4p  -AATTNTCTGAGGAATGCAAGAGGATACAACCTAAGACAAAAAACTTAATTGAATCCTGA 406 -AATTNTCTGAGGAATGCAAGAGGATACAACCTAAGACAAAAAACTTAATTGAATCCTGA 4 05 -AATTCCCAGCG--ATGCCAGGGGACACACCCTGTGACTCCTTCCTGAATTGAGTGCTGA 57 **** * * * * * * * ** * * * * * * * * * * * * ** * * * * * * * * * * *  type I I type I 4p  TATTTCATTAGTAAATAGGGTAATTGATGGATAAATGTAATGGTCTCGGTGGGTGGACAG TATATTATTAGTAAATAGGGTAATTGATGGATAAATGTAATGGTCTCGGTGGGTGGACAG TATTTGATTGGCTTATCGCGCACCTGATGAGTGGGTGGGGTGTTCGCGGTTGGTGGGGGT  type I I type I 4p  TAGTTAAATAAGGGCTGATGCAGCAAGATAATTATTTAAAGGCGTTTGAAAGAAA TT 521 TAGTTATATAAGGGCTGATGCAGCAAGATAATTATTTAAAAGAGTTTGAAAGAAA TG 522 GACTTACAGAAGGGCTGATGCGGCCAGAGAGCTCGTCA TTTGAA- GACTCTCTC 17 0 * *** * * * * * * * * * * * * * ** *** * * * * * * * * * * ** *  type I I type I 4p  GAAACAGGAGAGTGGATGTATTCAGCTAAAATAAAATCCGGAAGCCCTGAA-ATAAATNT GAAACAGAAGAGTGGATGTATTCAGCTAAAATAAAATCCGGAAGCCCTGAA-ATAAATNT GGAAG-GGATAGCGTCT TTCTGC A--ACCTGCGG--TCCCAGCAGACAAACCT  type I I type I 4p  CATTTTGCGGGTAAAAAAATGGCATTGGAGGA-GATTCTGGGTCAATCATG-AAGCTGTG 640 CATTTTGGGTGTAAAAAAATGGCATTAGAGGA-GATTCTGGGTCAATCATCCAAGCTGTG 64 0 TGTGATCCTCGTTCCA--GTCGACATGGAGGACGACTCAC--TCTA-CTTGGGAGGTG- - 2 71  type I I type I 4p  AAAGTTGCATCTTGGAAGCAGGATCCCTGTAATGAAACGAGACTTGTTTATCAGAGGTGG 700 AAAGCTGCATCTTGGAAGCAGGATCCCTGTAATGCAACGAGACTTGTATATCAGAGGTGG 700 - -AGTGGCAGTT- -CAACCACTTTCC AA AAC TCACATCTTC 308  type II type I 4p  TCTTTCAGANGAAAANATTTNGAAGAATGGCCCCTTCCTTTTGTGTATTTGACGATTAGA 760 TCTTTCAGTGGAAAAGATTTTGAAGAATGGACCCTTCCTTTTGTGTATTTGACGATTAGA 760 TCGGCCCGATGCA--GCTTTTGCTGAAATCCAGCGGACTTCTCT CCCTGAGA 358  *** * *** *  * **  *  **  ** * * *  * * ** *  *  ***  **  *  *  *  *****  *** **  * *  ** **  *  *  **  *  * ***  * ***** ** **  ***  **  ** ** * * * *  ***  466 46 5 117  *****  *** * * * ***  ** * * *  *  582 581 218  ** **  *** *  *  155  Appendix 1  type I I type I 4p  TNTCATGCCAAATTTCGGGTTTTAAACTCTATTTAANCNTTANAA-NAATTANCTANAAT 819 CTTCATGCCAAATCTCGGGTTTAAAACTCTATTTAAACATTAACA-GAATTAATTAAAAT 819 AGTCA CCACTCTCATGT- -GAGACCCGTGTCGACCTCTGTGACGATTTGGCTCCTGT 413  type I I type I 4p  GGCCAAAAAACAAGAAATTTTTTGATT--AGGAATCGTCAAATATTCATTTCTTGTTAGA 877 GGCAAAAAATCAAGAAATTTTTTGATT--AGGAATCATCACATATTCATTTCTTGTTAGG 877 GGCAAGACAGC TTGCTCCCAGGGAGAAGC TTCCTCTGAG- 4 52  type I I type I 4p  TACAGTTACCANACACCACCTACCGGAGAGAAACAATTGTGGAGAATGGCC-CCTTATTT TACAGTTATCAAAGATGACCTACCAGAGAGAAACGATTGTGGAGAATGGCC-CCTTACTT TAGC AG - GA- GACCTGC TGCG GTGGGGGCTGGGCTCCAGA ** * * * **** * * * * * * * * * * * * ** *  type I I type I 4p  TTGTTTATTTGCTGATNAGATTTCNTA-GTCCATTTC-TCATT--AGGTNCAGAGAT-CN 991 TTGTNTATTTGCTGATAAGATTTCNTACGTCCATTTTATCATT--AGGTNCNAAGAT-CA 992 ATATG- -GGAAATACCTGCTACGTGAACGCT-TCCTTGGAGTGGCTGACATACA 541  type I I type I 4p  AAGTTGACCTACCCAAGGAGTTGA-GATATCCAGGNACNNAAA-CTCAGGGCACGGTAGA 1049 AAGTTGACCTACCAAAG-AGTTGAAGATGTCCAGGA-CANNAA-CTCAGGGCTCAGTAGA 104 9 CACCGCCCCTTGCCAACTACATGC TGTCCCGGG--AGCACTCTCAA ACGTGTCA 593 • * * * * ** * ** * *** ** * **** * *  type I I type I 4p  ACCACAGAA-TCTTGGGTGAAATATTGCTCAAGAACAATAATGTNCTTATTCACC-GTGT ACCNCNGAAATCTTGGGTGAAATATTGCTCAAGAACAAAAATGTNCTTATTCCCAANTGT TCGTCACAA GGGCTGCA TGCTCT - GTACTA TGCAAGCTCACA  type I I type I 4p  TTCTGTGTGACATGTGTGGA-AACCAAAGTGCTTTGGAGCCTGACCCGAAGACTGAATTT 1166 TTGTGTGTGACATGTGTGTTGAATTAAAGTGCAATG-AGCATGACATGCAGGCANGACAT 1168 TCACACGGGCCCT CCACAATCC- -TG-GGCACGTCATCCAGCCCTCACAG 681 * *** * * * * * * ** ** * * ** * *  type I I type I 4p  CNTATTCGGCTC-CCTCCNNATCNGTTGTGTAGCCATTANAGAAAACCCAATTCCTAAGG 122 5 CN-ATTCGGCTCNCCTCAAAAGCAGTTNTGAA-CCTTAAAGGACAAC AATCCTA-GG 1222 -GCATT-GGCT GCTGGCTT CCATAGAGGCAAGC 712  type I I type I 4p  TCCCGCCTTAATGAANATTAGANCCATCCTTCCNCCTGTTGTNTAAANCCNCCGCATCTN TCCCGC-TTAAAGAAA--TAAGACCATCCC-CCACCC--TGTGTTGAACCCCCGCNTCTG AGGAAG ATGCCCAT GAA  type I I type I 4p  GGGCTTGCTCATGNATTCCTGGGGGATCATTCCCCAAAAAAATGGGTGGGCTCCCTTCCT 1345 GA--TTGCTCCTGT-TTCT--GGGGAACATCCTCCTGAAAA-TGGCGG TCCTTTCC- 1326 TTTCTCATGT- -TCACTGTGGA TGCCATGAAAAAGGCATG CCTTCC- 773  type I I type I 4p  CCCGG-- 1350 CCCN 1330 -CGGGCA 78 0  ***  * * * **  **  * ** *  *** * * * *  *** *  •k -k -k -k k  *  *  **  k k  ** ** *  *** ****  * *  *** *  *  *  *  kk kk  * ** *  ** *  * ***  *  *  * * * * * **  * *  *  -k -k  k  ** *  *  *  ****  936 936 490  k kk k  1107 110 9 634  * *  * * * * *  **  * **  *  k k k -k -k  *****  * *  ****  ** * * * **  k  *  *  *  * ****  128 5 1276 72 9  156  Appendix 1  Alignment b2: type I I type I 4p  AACACACACAAACTGTTTAATATGAACACAATTGTTAAATAGCATTGTTATAACATGAAA 226 CAAACACACAAACTGTTTAATATGAACACAATTGTTAAATAGCATTGTTATAT-ATGAAA 22 5 CACACACGCACACGGTTTCCTAGGTAAAGATTTCTTCCCTGCCATTGCTTTAC-CTAAAA 462 7  type I I type I 4p  TAAGGCAAATGTTTAGTTANTATCCTAACCCGGTTCCCATTCCTAACATATTTGCT-CAT 2 85" TAAGGCAAATGTTTAGCTATTATCNTAGCCCG-TTCCCACACCTAACATATTTGCTTCAT 284 TAAGGCAACTGTGAGGCCGCTGTCCCAACCCGGTTAC-ACTCCTATTATATGTGC 4682  ******** ***  *  * **  * * * * * ** * *  ****  **** ***  Alignment b3: type I I type I 4p  TTTAAAGGCGTTTGAAAGAAATTGAAACAGGAGAGTGGATGTATTCAGCTAAAATAAAAT TTTAAAAGAGTTTGAAAGAAATGGAAACAGAAGAGTGGATGTATTCAGCTAAAATAAAAT CTGTATTGATAAT-AAAGGAAAGCAAACACAGGAGTGTGTGTATTCAACTGAAATAAATT  559 558 26 09  type I I type I 4p  CCGGAAGCCCTGAAATAAATNTCATTTTGCGGGTAAAAAAATGGCATTGGAGGAGATTCT CCGGAAGCCCTGAAATAAATNTCATTTTGGGTGTAAAAAAATGGCATTAGAGGAGATTCT CAGAAAGCCCTGAAATCAATCTCACTGGGTGTGTTTAAAAATGGCATTTGGGGAATTTCT  619 618 266 9  type I I type I 4p  GGGTCAATCATG-AAGCTGTGAAAGTTGCATCTTGGAAGCAGGATCCCTGTAATGAAACG 6 78 GGGTCAATCATCCAAGCTGTGAAAGCTGCATCTTGGAAGCAGGATCCCTGTAATGCAACG 6 78 GGGTCATTTGTC-CAGCTGCGAAAGCTGCATCTCTGAAGCACAGTCCCTGTCCCGCAGTG 2 728 ****** * * ***** ***** ******* ****** ******* * * *  type I I type I 4p  AGACTTGTTTATCAGAGGTGGTCTTTCAGANGAAAANATTTNGAAGAATGGCCCCTTCCT AGACTTGTATATCAGAGGTGGTCTTTCAGTGGAAAAGATTTTGAAGAATGGACCCTTCCT AGACTTATTTATCCGACGTGGTGTTTCCGTGGAAATGATTGTGGGAAATGGCCCCTTCCT  type I I type I 4p  TTTGTGTATTTGACGATTAGATNTCATG 766 TTTGTGTATTTGACGATTAGACTTCATG 766 TTTCTCTATTTGCTGATTAGACTTCATGGTCCCTTTCTCGTCAGGTACAGTGATCAAAGT 2 84 8  *  *  *  * * * * * **  *****  * * ************ *** *** *  * * **  * * * * * * * * * * * ** * * * * * * * * * *  *** * ******  *******  *****  * * * * * * * * ** * * * * * * * *  ************ * ***  ****  ***  *  *****  ****  ********  73 8 738 2 788  *****  Alignment b4: type I I type I 4p  ATNTCATGCCA- -AATTTC GGGTTTTAAACTCTATTTAANCNTTANAANAATT 809 ACTTCATGCCA- -AATCTC GGGTTTAAAACTCTATTTAAACATTAACAGAATT 809 AAGCCCTGAAATCAATCTCACTGGGTGTGTTTAAAAATG GCATTTGGGGAATT 2666  type I I type I 4p  ANC TANAATGGCCAA AAAACAAGAAATTTTTTGATTAGGAATCGTCAAATAT 861 AAT TAAAATGGCAAA AAATCAAGAAATTTTTTGATTAGGAATCATCACATAT 861 TCTGGGTCATTTGTCCAGCTGCGAAAGCTGCA TCTCTGA AG CACAG-T 3867  *  * **  *  ** * *  * * * **  *** *  * **** *** *  *  * * ***  * **  *  ** *  ****  *  157  Appendix 1  type I I type I 4p  TCATTTCTTG TTAGATACAGTTACCANACACCACCTACCGGAGAGAAACAATTGTGG 918 TCATTTCTTG TTAGGTACAGTTATCAAAGATGACCTACCAGAGAGAAACGATTGTGG 918 CCCTGTCCCGCAGTGAGACTTATTTATCCGACGTGGTGTTTCCGTG-GAAATGATTGTGG 3 92 7  type I I type I 4p  AGAATGGCCCCTTATTTTTGTTTATTTGCTGATNAGATTTCNTA-GTCCATTTC-TCATT 976 AGAATGGCCCCTTACTTTTGTNTATTTGCTGATAAGATTTCNTACGTCCATTTTATCATT 978 GAAATGGCCCCTTCCTTTTCTCTATTTGCTGATTAGACTTCATG-GTCCCTTTC-TCGTC 3 985 *********** **** * * * * * * * * * * * * *** *** * **** *** ** *  type I I type I 4p  AGGTNCAGAGATCNAAGTTGACCTACCCAAGGAGTTGA--GATATCCAGGNACNNAAACT 1034 AGGTNCNAAGATCAAAGTTGACCTACCAAAG-AGTTGAA-GATGTCCAGGA-CANNAACT 103 5 AGGTACAGTGATCAAAGTTGACCAACCCCAG-AG- -GAAAGCTGCCCAGGG--CACAACT 404 0 **** * * * * * * * * * * * * * * * * * ** ** * * * * * * * * * ****  type I I type I 4p  CAGGGCACGGTAGAACCACAGAA-TCTTGGGTGAAATATTGCTCAAGAACAATAATGTNC 1093 CAGGGCTCAGTAGAACCNCNGAAATCTTGGGTGAAATATTGCTCAAGAACAAAAATGTNC 1095 CAGGGCTCCGTAGAACCACAGAA-TCTTGGGCGCAACCCTGCTCAAGCACCCAAATGTGC 3 999 * * * * * * * * * * * * * * * * * * * * * * * * * * * ** * * * * * * * * ** ***** *  type I I type I 4p  TTATTCACC-GTGTTTCTGTGTGACATGTGTGGA-AACCAAAGTGCTTTGGAGCCTGACC 1151 TTATTCCCAANTGTTTGTGTGTGACATGTGTGTTGAATTAAAGTGCAATG-AGCATGACA 1154 ATACGAACA-GGGTCTCCGTGTGACGTGTGTGAA-AACTACAGTGTGATG-AGCATGACT 4156  type I I type I 4p  CGAAGACTGAATTTCNTATTCGGCTC-CCTCCNNATCNGTT 1191 TGCAGGCANGACATCN-ATTCGGCTCNCCTCAAAAGCAGTT 1194 GGCAGACAGCTTATCG-ATTGGGCTCCCCTCAAAATCGGTTATGAGCATTCAAGCACACC 4215 * ** * ** * * * * * * * * * * * * * * ***  * * **  **  *  *  * **  ** *  * *** *  *  *  ******* ******  **  * * * ****  * ****  *******  ** * * * * * * *  Alignment b5: type I I type I 4p  CCGGAGAGAAACAATTGTGGAGAATGGCCCCTTATTTTTGTTTATTTGCTGATNAGATTT 959 CCAGAGAGAAACGATTGTGGAGAATGGCCCCTTACTTTTGTNTATTTGCTGATAAGATTT 959 TCCGTG-GAAATGATTGTGGGAAATGGCCCCTTCCTTTTCTCTATTTGCTGATTAGACTT 2 812  type I I type I 4p  CNTA-GTCCATTTC-TCATTAGGTNCAGAGATCNAAGTTGACCTACCCAAGGAGTTGA- - 1015 CNTACGTCCATTTTATCATTAGGTNCNAAGATCAAAGTTGACCTACCAAAG-AGTTGAA- 1017 CATG-GTCCCTTTC-TCGTCAGGTACAGTGATCAAAGTTGACCAACCCCAG-AG--GAAA 2867  type I I type I 4p  GATATCCAGGNACNNAAACTCAGGGCACGGTAGAACCACAGAA-TCTTGGGTGAAATATT GATGTCCAGGA-CANNAACTCAGGGCTCAGTAGAACCNCNGAAATCTTGGGTGAAATATT GCTGCCCAGGG--CACAACTCAGGGCTCCGTAGAACCACAGAA-TCTTGGGCGCAACCCT  type I I type I 4p  GCTCAAGAACAATAATGTNCTTATTCACC-GTGTTTCTGTGTGACATGTGTGGA-AACCA 1132 GCTCAAGAACAAAAATGTNCTTATTCCCAANTGTTTGTGTGTGACATGTGTGTTGAATTA 1136 GCTCAAGCACCCAAATGTGCATACGAACA-GGGTCTCCGTGTGACGTGTGTGAA-AACTA 2 982  * * * ****  * *  * *  **** ***  *****  * * * * * * * **  *******  ***********  ** * * * * * *  * * * * * * * * * * * * * * * * * * * **  **** ********* ***  ** **  **  * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * **  * * * * * * **  *  ** *  ******* ******  **  *  *  1074 1076 2 924  158  Appendix 1  type I I type I 4p  AAGTGCTTTGGAGCCTGACCCGAAGACTGAATTTCNTATTCGGCTC-CCTCCNNATCNGT AAGTGCAATG-AGCATGACATGCAGGCANGACATCN-ATTCGGCTCNCCTCAAAAGCAGT CAGTGTGATG-AGCATGACTGGCAGACAGCTTATCG-ATTGGGCTCCCCTCAAAATCGGT  ****  ** * * * * * * *  * ** *  **  *** ***** ****  1191 1194 3 04 0  * * **  type I I type I 4p  TGTGTAGCCATTANAGAAAACCCAATTCCTAAGGTCCCGCCTTAATGAANATTAGANCCA TNTGAA--CCTTAAAGGACAAC-AAT-CCTA-GGTCCCGCTTAAAGAAATA--AGA-CCA TATGAG--CATTCAAGCACACC-GATGCCCA-GGTCCCGGCTGCAGGAATA--AGA-CCC  type I I type I 4p  TCCTTCCNCCTGTTGTNTAAANCCNCCGCATCTNGGGCTTGCTCATGNATTCCTGGGGGA 1311 TCCC-CCACCC--TGTGTTGAACCCCCGCNTCTGGA--TTGCTCCTGTTT--CTGGGG-A 12 98 TCCA-GGGTCT--TGTGTGAAGCCTCGGCATCTGCA--TTGCTCATGCTT--CTGGGG-A 314 5  type I I type I 4p  TCATTCCCCAAAAAAATGGGTGGGCTCCCTTCCTCCCGG 13 50 ACATCCTCCTGAAAA- TGGC - - GG - TCCTTTC - CCCCN 13 31 TCATTCTCCTGAAAA-TGGT--GGCTCCTTTC-TCCCTGTGGAGCATCTTTCTAAGCAGT 32 01  c.  * **  * **  ***  *  * * * * **  ** * * *  *** *  ** ** * * * * * * * *  * ** * *  **** ***  ***  ** * * * * * *  *  *  * * * * * * **  *  12 51 1246 3 093  * * * **  ****** *  *** *  Comparison o f sequences C, c, and the 4 p m e g a s a t e l l i t e .  type I I type I 4p  AATTCTGTCCCTCTGAG AATGCCGGA AATTCTGTCCCTCTGAG AATGCTGGA GTTGGGAAAATCCCTAGAGCCAGGATCTTCATTCCCGCTAAGCCAGACAGCCGGAAGACA  type I I type I 4p  CAGGTTTACCTTCATCACCATAA-AATT 53 CAGGTTTACCTTCATCACCATAA-AATT 53 CACCCAAATTCTGTCCCTCTTACTTCAGGGAACATGTCCACTTTCGGCAGCATTACAATT 355 9  type I I type I 4p  TTGGAAC-AAATGTGGTAACTGCAGGTTCTCCCCACAATGAGTAACTGAAAATTGAGGCA TTGGAAC-AAATGTGGTAACTGCAGGTTCTCCCCACAATGAGTAACTGAAAATTGAGGCA TTGGCACCAAATGTGCTAACTGCAATTCCACCATACAATGCGTAACTGGAAATGGAGGCA  type I I type I 4p  GTATTTCAGATCCTAAAAAACTGATGAAGTAATTCGCCACACATTTGGGTTGTTTTTGCC 172 GTATTGCAGATCCTAAAAAACTGATGAAGTAATTCACCACACATTTGGGTTGTTTTTGCC 172 ACATCTCCGATCCTGAACGATCGATGCGAGAATCCAGGATATGCACGGCTTATTTTGGCC 3679  type I I type I 4p  TTTTCCTACTATGAAGATTGCTAATAGGAAGAATAGTGTACAAGTATCTGTTTGATTCCC 232 TTTTCCTACTATGAAGATTGCTAGTAGGAAGAATGGTGTACAAGTATCTGTTTGATTCCC 232 TTTTCCCACTGAAACAAGGGCCAGTATTAAAAATGGCACGCTATCCTCTGTTTCACTCCC 3 73 9  type I I type I 4p  TGCTTTTAGAATCCTTTGCTTGTTTGTGTTGTTTGTCTGTTCCTTCTTGAGACAGGATGT 2 92 TGCTTTTAGAATCCTTTGCTTGTTTGTGT-GTTTGTCTGTTCCTTCTTGAGACAGGATTT 2 91 TGCTTTTA-AAC GTCTCCGA TGTTTCTCCCTGAGACAGCGCCT 3 781  ***  * * * * ** **  ** **  * * * * ** * * * * * * * * * * * * * * *  **  * * * * * * * **  ****** ***  * * * * * * * * **  *  *  *  * * **  ****  ** *  ** * * * *  *  ** * * *  ** * * * * * * * *  ****** ******* **** ******  *** *  ** * **  ** * * *  *  26 26 3 52 9  * *  * *  112 112 3 619  ** ** * * * * * * *  ******* * ****  * * * * ** * * * * * * * * *  *  Appendix 1  159  type I I type I 4p  CACTCCCAGTCCAGCCCAGGATTTTCCCAGTT-TTGTTAATTTTTGTTGTTTCCCTTTTG 3 51 CACTCCCAGTCCAGCCCAGGCTTTTCC-AGTTGTTGTTAATTTTTGTTGTTTCCCTTTTG 3 50 CACTTCC-GTC-AGCCG-GGCTTTTCTACGGT ATAATTTTCCTTGTTTGC-TTTTG 3 83 3  type I I type I 4p  TTCAATTTTTAGAAATTTGTTATTTTATTTCCTATTGAATTTTAAGGCATTTTTAGATAT 411 TTCAATTTTTAGAAATTTGTTATTTTACTTCCTATTGAATTTTAAGGCATTTTTAAATAT 410 TCCAAAT- -TAGAACTTT-TTATTTCATCTC-TAGGAAACGTTGATCCATTATCACATAC 3889  type I I type I 4p  GTTATTAAAACATTATCCACACAGGCCGTGTTGTTTACATTGCAATTATTTCCACCA-TC 4 70 TTT-TTAAAAANTTTTCCCCA-NNCCCGTNTT--TTAANTNGNAAT-ATCTCCCCCG-TC 464 GTA-TGGAAATATTATC-ACACATGCTGTGAG--ATACGTTGTTTTTATTTTCATCAATT 3 94 5  type I I type I 4p  CCCTTTAAGAAACCAAAGAAT TACCTTA CC - TTNAAAAAACAAAGGTTT NACCTTA CC--TTAATAAACAAAAGGTTATAGCTGGGATACCTTCTGAGTTCTCAAGTTTTTTGTTT  * * * * ** * * * * * * *  * *** *  *  **  *  ** * * * * *  ***** *** ****** *  ***  ** **  * ** * * * * ** *  **  *  * **  * *  *******  ** **  **  **  *****  ** *  * *  ****** * *****  **** * * ***  * ** * *  *  *  4 98 4 91 4 0 03  160  Appendix 2  Appendix 2: BACs Isolated from the Research Genetics BAC library  Probes used for isolation: 87B23 T7 and SP6 ends. B A C clones isolated were in excess of 3,000. These B A C s represent the strongest positive signals for clones present in the enriched chromosome 8 B A C library. 255E8 285P3 300F13 351120 262F24 273J3 282B18 257E19 223B23 308D20 286D2 328H13 87G21 67B6 164L15 110A12  Note: 223B23, 67B6, 286D2 shown in a separate experiment to overlap with 87B23 but not contain the megasatellite  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0088696/manifest

Comment

Related Items