UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Genetic mechanisms of nondisjunction in humans Gair, Jane Louise 2005

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata

Download

Media
831-ubc_2005-104945.pdf [ 16.19MB ]
Metadata
JSON: 831-1.0092280.json
JSON-LD: 831-1.0092280-ld.json
RDF/XML (Pretty): 831-1.0092280-rdf.xml
RDF/JSON: 831-1.0092280-rdf.json
Turtle: 831-1.0092280-turtle.txt
N-Triples: 831-1.0092280-rdf-ntriples.txt
Original Record: 831-1.0092280-source.json
Full Text
831-1.0092280-fulltext.txt
Citation
831-1.0092280.ris

Full Text

GENETIC MECHANISMS OF NONDISJUNCTION IN HUMANS by JANE LOUISE GAIR B.Sc, The University of British Columbia, 1998  A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY In THE FACULTY OF GRADUATE STUDIES (Medical Genetics)  THE UNIVERSITY OF BRITISH COLUMBIA June 2005 © Jane Louise Gair, 2005  Abstract  Missegregation of chromosomes in meiosis, or nondisjunction, occurs relatively frequently in humans, and results in pregnancy loss. There is a correlation with advancing maternal age, but the cause of the dramatic increase of aneuploidy, and specifically trisomy (the presence of three copies of a chromosome rather then two), seen with age remains unknown. There is evidence to suggest that chronological age is less important than biological age for trisomy risk, and that regardless of their chronological age, some women are at a greater risk of having a trisomic pregnancy after having already experienced one. Several features of chromosomes are associated with aging, such as a decrease in telomere length, an increase in replication asynchrony at loci including centromeres, and an increase in somatic cell aneuploidy with increasing age. For some chromosomes (15 and 21) an association has also been observed between maternal age, reduced recombination along the chromosome, and risk for nondisjunction. In this project, I have investigated whether or not some women are predisposed to having a trisomic pregnancy. That is, can we predict who will have recurrent trisomy? After analyzing telomere length for an association, no significant decrease in length was seen for women experiencing recurrent trisomy when compared to control women. There was a trend, however, towards longer telomeres in women with a "good" reproductive history (children after 37 years of age) compared to women with a "poor" reproductive history (trisomy and/or recurrent trisomy). As well, although there was no significant increase in replication asynchrony in mothers of trisomy as a group, the younger mothers (<35 years old) of trisomies had increased replication asynchrony when compared to controls of the same age. The relationship between recombination and nondisjunction has been well established and my studies of chromosome 15 confirmed that decreased recombination is associated with meiosis I errors and increased recombination is associated with meiosis II errors. A family with an apparent inherited predisposition to missegregation of chromosome 21 seemed the ideal family in which to study possible genetic mechanisms of nondisjunction. Although nothing conclusive could be determined to be causing their segregation problems, a cryptic rearrangement involving the centromere of chromosome 21 is the most likely explanation. Finally, telomere length was not found to be shortened in children conceived through intracytoplasmic sperm injection (ICSI).  iii Table of Contents Abstract  ii  Table of Contents  iii  List of Tables  vi  List of Figures  viii  List of abbreviations  xii  Acknowledgements  xiv  Dedication Chapter 1:  xv Introduction  1  1.1  Opening remarks  1  1.2  An Overview  2  1.3  1.2.1  Meiosis 1.2.1.1 Timing 1.2.1.2 Recombination 1.2.1.3 Chromosome cohesion and segregation 1.2.2 Mitosis 1.2.2.1 Timing 1.2.2.2 Recombination 1.2.2.3 Chromosome cohesion and segregation  2 2 7 10 16 16 17 17  Aneuploidy in Humans (overview)  19  1.3.1  Trisomy 1.3.1.1 Trisomy mosaicism 1.3.1.2 Origin of mosaicism 1.3.1.3 Uniparental disomy 1.3.2 Monosomy 1.3.3 Segmental aneuploidy 1.3.3.1 Deletions 1.3.3.2 Duplications 1.4  20 21 22 25 -26 27 28 28  Factors that may affect risk of trisomy  29  1.4.1  29 31  Review of Epidemiology 1.4.1.1 Aging  iv  1.4.2 1.4.3 1.4.4 1.4.5 1.4.6 Chapter 2: 2.1  1.4.1.2 Recombination Ovarian aging Chromosome structure 1.4.3.1 Centromeres 1.4.3.2 Telomeres Replication timing Gene mutations Environmental  32 35 37 37 40 41 42 44  Materials and Methods  .47  Subjects and sample collection  47  2.1.1  47  Trisomic pregnancy mothers and controls  2.2  Tissue Culture and harvest  48  2.3  Molecular methods  49  2.3.1 2.3.2  49 50 50 51. 51  2.3.3 2.4  2.5 Chapter 3:  DNA extractionfromblood Microsatellite marker typing 2.3.2.1 Origin of trisomy 2.3.2.2 Recombination mapping (genetic mapping) Telomere length assay  Cytogenetic methods  53  2.4.1  Q-FISH  53  2.4.2  Interphase FISH  57  Statistical analysis  59  Telomere length in women who have experienced a trisomic pregnancy  3.1  Introduction  63  3.2  Results  66  3.3 Chapter 4:  Discussion 78 Replication timing and somatic aneuploidy in women who have experienced a trisomic pregnancy  4.1  Introduction  4.2  Results  83 8 8  V  4.3 Chapter 5:  Discussion  96  Recurrent trisomy 21 in a Northern BC family  5.1  Introduction  100  5.2  Results  105  5.3  Discussion  116  Chapter 6:  Telomere length in IVF/ICSI cases  6.1  Introduction  126  6.2  Results  128  6.3  Discussion  131  Chapter 7:  Recombination at common deletion breakpoints and nondisjunction of chromosome 15  7.1  Introduction  133  7.2  Results  141  7.3  Discussion  150  Discussion and Conclusion  156  Chapter 8: References  162  Appendix 1  183  vi List of Tables  Table 1-1: The main differences between males and females in meiosis  14  Table 1-2: The main differences between meiosis and mitosis  18  Table 3-1: List of patient samples and pregnancy history (N-37 for Southern and N=l 3 for Q-FISH). SA = spontaneous abortion, T = term birth, E = ectopic pregnancy, SB = stillbirth and TA = termination 70 Table 3-2: List of control samples (n=39 for Southern and n=l 1 for Q-FISH) Table 4-1: The amount of replication asynchrony observed at 15Z4 in cases (N=7) and controls (N=19)  72 89  Table 4-2: Results for replication timing asynchrony, % aneuploidy (single or triple signals) and HPRT as an internal control to ensure nuclei replication. * Replication asynchrony is significantly different at 21q22 between cases and controls. No other significant differences were observed 90 Table 4-3: When replication asynchrony is analyzed at each locus based on age at first trisomy, there is a significant difference seen when comparing the younger members of each group (<35 years of age) at 21q22 (p=0.001; t-test). No other significant differences are seen 92 Table 4-4: The amount of replication asynchrony observed at 21q22 in cases (N=l 1) and controls (N=17) ....92 Table 4-5: Somatic aneuploidy data for cases (N=2) and controls (N=4) at all 3 loci. FISH efficiency, as determined by the number of 0 signals is included 95 Table 4-6: The amount of replication asynchrony observed at 15Z4 and 21q22 in cases (N=2) and controls (N=5 for 21q22 and N=4 for 15Z4) using HPRT as an internal control; i.e. HPRT was always observed as an SD pattern before the nucleus was scored for the other probe 96 Table 5-1: Microsatellite markers spanning chromosome 21 were analyzed in all members of the pedigree and the genotype data is found below 104 Table 5-2: Somatic aneuploidy in controls (N=4) compared to 2 of the mothers from the pedigree. % aneuploidy was analyzed at 15Z4, 21q22 and HPRT on the X chromosome.. 109 Table 5-3: Microsatellite marker results for chromosome 13 markers. D13S633 was informative and was found to cosegregate with the duplicated alleles on chromosome 21 in the same individuals (1-2 and II-3) Ill  Vll  Table 5-4: Genotyping data for both polymorphisms at MTHFR for all family members (N=9) and controls (N=8) 113 Table 5-5: Frequency of maternal MTHFR C677T genotypes in case and control groups. 114 Table 5-6: Frequency of maternal MTHFR A1298C polymorphisms in case and control groups 114 Table 6-1: Telomere lengths measured by Southern blot. The average telomere length for the ICSI group is not significantly different than the control group (p=n.s.; t-test) 130 Table 7-1: The most recent physical and genetic map data for 15ql l-ql3. The CEPH (control) male and female genetic maps of this region (738CA to D15S144) are estimated to be 27 cM and 32 cM respectively 143 Table 7-2: Genetic maps of CEPH male and female controls compared to the MI and Mil error groups of mothers. The total genetic map length for MI error mothers was estimated to be 13.8 cM while the genetic map for Mil error mothers was 36 cM 149  Vlll  List of Figures  Figure 1-1: Diagram of meiosis. A single chromosome pair with a single crossover is shown. This leads to the formation of 4 distinct gametes  6  Figure 1-2: Recombination heteroduplex, formed when a single strand (represented by the dotted lines) from a broken duplex "invades" an intact double- stranded DNA duplex 8 Figure 1-3: The kinetochores of sister chromatids are side-by-side at MI, and face the same pole. This results in both sister chromatids moving to the same pole in anaphase I. Adapted from Paliulis and Nicklas, 2000 11 Figure 1-4: The kinetochores of the sister chromatids are back-to-back at M i l , and face opposite poles. This results in the sister chromatids moving to opposite poles in anaphase II. Adapted from Paliulis and Nicklas, 2000 12 Figure l-5a: Mosaicism can result from anaphase lag in meiosis. This diagram shows loss of one of the three copies of the chromosome at the 2 meiotic division resulting in two cell lines; one with 46 chromosomes and one with 47. This is also called "trisomic rescue". The extra chromosome is isolated in a separate cell and is eventually lost 23 nd  Figure l-5b: Mosaicism can also arise during a nondisjunction event in mitosis. A cell with two copies of a chromosome givesriseto one cell with three copies and another with only one copy of that chromosome. Here the cell with the three copies of the same chromosome will continue to grow, but the cell with only one copy of the chromosome will not 24 Figure 1-6: Unequal recombination between misaligned sister chromatids or homologous chromosomes containing highly homologous sequences leads to both deletion and duplication products. This is seen here, where there is either one copy or three copies of the sequence, represented by the arrows, rather than two .27 Figure 2-1: Computer generated image of a metaphase preparation after hybridization with PNA probes for the telomeres and counterstained with DAPI. An algorithm segments the telomeres and allows the TFL-TELO program to determine the amount of fluorescence at each telomere segment. The numbers assigned to the chromosomes are arbitrary 56 Figure 2-2: Diagram of the dot hybridization assay used to assess replication timing in interphase nuclei. Two single dots or two double dots for a single locus indicate a synchronous pattern of replication, while a single dot and a double dot found within the same nucleus indicates replication asynchrony 59 Figure 3-1: A Southern blot showing typical smears resultingfromhybridization of a telomere repeat (TTAGGG)-specific probe to digested genomic DNA. Each lane represents one sample. The darkest point in the smear indicates the average TRF (terminal restriction fragment) length in each sample and its size is estimated using the molecular weight ladders shown in lanes 1 and 20 67  ix Figure 3-2: TRF length by age after measurement by Southern blot in all cases (N=37 in red) and controls (N=46 in blue). The correlation coefficients are not significantly different (p=n.s.; Fisher r-to-z transformation) 68 Figure 3-3: A metaphase preparation after Q-FISH with PNA probes for the telomeres. A computer program TFL-TELO measures the amount of fluorescence emitted from each telomere and calculates a value that can be converted to a kilobase length. All of the telomere lengths are then averaged to arrive at an average value for telomere length for that sample 69 Figure 3-4: Telomere length by age after measurement using Q-FISH in cases (N=T3 in red) and controls (N=l 1 in blue) 70 Figure 3-5: Correlation between the Southern method of analysis and Q-FISH for telomere length measurement in samples measured with both methods (N=25) 74 Figure 3-6: Telomere length measurements for all samples measured by Southern analysis (N=95 in red) and Q-FISH (N=28 in blue). TRF lengths are significantly longer (p<0.0001; t-test) but correlation coefficients are not significantly different between the two methods (p=n.s.; Fisher r-to-z transformation) 74 Figure 3-7: TRF length by age in all cases (N=37 in red) and "reproductively healthy" (RH - children over age 37 years) controls (N=17 in blue). The correlation coefficients are not significantly different (p= n.s.; Fisher r-to-z transformation) 76 Figure 3-8: TRF length by age in high-risk cases (women experiencing >1 trisomy) (N=22 in red) and RH controls (N=17 in blue). The correlation coefficients are not significantly different (p = n.s.; Fisher r-to-z transformation) 76 Figure 3-9: TRF length by age in women experiencing their first trisomy at <35 years old (N=10 in red) and women experiencing their first trisomy at or after 35 years of age (N=27 in purple). The correlation coefficients are not significantly different (p= n.s.; Fisher r-to-z transformation) 77 Figure 3-10: Diagram of the structure of the telomere. Southern analysis measures the length of the telomeric repeats (TTAGGG) as well as a small portion of the telomere associated repeats (TAR). This structure is referred to as the terminal restriction fragment (TRF) 78 Figure 4-1: Replication asynchrony with age at 15Z4 in cases (N=7 in red) and controls (N=19in blue)  90  Figure 4-2: Replication asynchrony with age at 21q22 in cases (N=l 1 in red) and controls (N=17in blue) 95 Figure 5-1: Pedigree of recurrent trisomy 21  101  X  Figure 5-2: FISH with probes for the 13/21 centromere and 21q22 was performed on samples from individuals II-3 and III-13. Two signals on each of the chromosomes 21 and one signal on each of the chromosomes 13 was observed. No translocation to another chromosome could be detected 106 Figure 5-3: An amplified signal at the 13/21 centromere in II-3 as well as a diminished signal at 21q22 in this same sample was observed, prompting investigations into a possible inversion 107 Figure 5-4: When each FISH probe (21q22 on the left metaphase and 13/21 centromere on the right metaphase) was hybridized to samples from II-3, no split signals were observed.. 108 Figure 5-5: Haplotype reconstruction based on the microsatellite marker typings. Two individuals were found to have 3 alleles at 3 loci (as shown by the red shaded in box). Each shift in colour along either the maternal chromosomes (orange and yellow) or the paternal chromosomes (purple and blue) indicates a crossover. The red box enclosing two alleles is the only region shared by all women in the pedigree 112 Figure 5-6: Telomere length as measured by Southern analysis in 4 mothersfromthis family (N=4 in red) compared to controls (N=46 in blue)  115  Figure 6-1: Distribution of telomere lengths, measured by Southern blot analysis, in the ICSI newborns (N=24) compared to newborn controls (N=16) 129 Figure 7-1: 15ql l-ql4. All six breakpoints are shown. Locations of both LCR15s and 7/£7?C2-containing dupliconsfromPujana et al. 2002 are shown. Paternally expressed genes are boxed in blue and the maternally expressed gene is boxed in pink. IC=imprinting center 136 Figure 7-2: 15ql l-ql3 divided into 11 intervals in order to analyze recombination patterns  140  Figure 7-3a: IndividualsfromCEPH families were typed for microsatellite markers and recombinants were identified. Observed crossover events are indicated, where each coloured line represents the region within which the crossover was localized in an individual meiosis. Maternal crossovers (of N=246 total meioses) are indicated in pink and male crossovers (of N=247 total meioses) are indicated in blue 144 Figure 7-3b: IndividualsfromUPD15 and T15 families as a result of an MI (of N=132 total meioses) or an M i l (of N=25 total meioses) error in meiosis were typed for microsatellite markers and recombinants were identified. Observed crossover events are indicated, where each coloured line represents the region within which the crossover was localized in an individual meiosis 145  xi Figure 7-4: The estimated c M values for C E P H males and females (controls) compared to M I and M i l error chromosomes. A significant difference is seen in interval 1 with females having more recombination than males (p=0.02; Fisher's exact). No other significant differences among groups were seen 146 Figure 7-5: Chromosome 15ql l-ql3 divided into 11 intervals from centromere on the left to microsatellite marker D15S144 on the right. C E P H control male and female cM/Mb data is shown. Note the sex-specific recombination hotspots, in particular, the two female hotspots of recombination in intervals 1 and 8 and the male recombination hotspot in interval 9. Intervals 3 and 5 appear to regions of high homologous recombination in both sexes 147 Figure 7-6: Chromosome 15ql l-ql3 divided into 11 intervals from centromere on the left to microsatellite marker D15S144 on the right. MI error mothers and M i l error mothers cM/Mb data is shown compared to C E P H control female data. There is not a significant decrease in recombination in general in this centromeric region for the MI error mothers (p =n.s.; x ), while there is a significant increase in recombination for the M i l error mothers (p<0.001; x) 150 2  2  List of abbreviations Abbreviation  Full Name  ABI AD AS BAC bp BP BP A BrdU BWS C cM CDP CENPs CEPH CPM CVS DAPI DD dF DG/VCFS DIG DNA DS DSBs FISH HPRT HR IC ICSI IFI IVF LCR MI Mil mat Mb MLH1 MR mtDNA MTHFR MTRR ND  Applied Biosystems Alzheimer's disease Angelman syndrome Bacterial artificial chromosome Base pairs Breakpoint Bisphenol A Bromo-deoxy-uridine Beckwith-Weidemann syndrome Control Centimorgan Chemiluminescence detection phosphatase Centromere bonding proteins Centre d'Etude du Polymorphisme Humain Confined placental mosaicism Chorionic villus sampling 4,6-diamidino-2-phenylindole Double-double nucleus signal by FISH Deionized formamide DiGeorge/velocardiofacial syndromes Digoxigenin Deoxyribonucleic acid Down syndrome Double strand breaks Fluorescence in situ hybridization Hypoxanthine guanine phosphoribosyl transferase Homologous recombination Imprinting centre Intracytoplasmic sperm injection Integrated fluorescence intensity In vitro fertilization Low copy repeat Meiosis I Meiosis II Maternal Megabase DNA mismatch repair protein Mental retardation Mitochondrial DNA Methylenetetrahydrofolate reductase Methionine synthase Nondisjunction  NF1 pat PBS PCR PNA PWS Q-FISH RH RSA SA SC SCP3 SD SMR SMS ss SS STS T21 T TAR TRF UPD WBS  Neurofibromatosis 1 Paternal Phosphate buffer solution Polymerase chain reaction Peptide nucleic acid Prader-Willi syndrome Quantitative FISH Reproductively healthy Recurrent spontaneous abortion Spontaneous abortion Synaptonemal complex Synaptonemal complex protein 3 Single-double nucleus signal by FISH Standardized morbidity ratio Smith-Magenis syndrome Single-stranded Single-single nucleus signal by FISH Sequence tagged site Trisomy 21 Thymidine Telomere-associated repeats Terminal restriction fragment Uniparental disomy Williams-Beuren syndrome  xiv Acknowledgements I would like to thank Wendy Robinson for being a very supportive and helpful supervisor over the years. I would also like to thank everyone in the Robinson lab for their help and support. In particular the technical support of Ruby Jiang has been greatly appreciated. There are many projects that I could not have done without the help of Maria Penaherrera, Chiho Hatakeyama and other students in the lab. Thanks to all members of my committee, Dr. Carolyn Brown, Dr. Ann Rose, Dr. Fred Dill, Dr. Valia Lestou and Dr. Peter Lansdorp. I would like to thank Peter Lansdorp and Elizabeth Chavez for their technical and collaborative support with the telomere chapter. Thanks for your patience and discussions. I am also indebted to Carolyn Brown and all members of the Brown lab for support, help and collaboration. I could not have proceeded with any of my projects without the patient and control samples and I thank the genetic counselors and geneticists at the Medical Genetics Clinic at Children's and Women's Hospital for their constant help and recruitment. The Women's Health Clinic and the recruitment by Mary Stephenson was also invaluable. Thank you to Helene Bruyere and the Cytogeneticists at Vancouver General Hospital for their expertise and help. Also the Cytogenetics lab at Children's and Women's Hospital for constant discussion, technical support and friendship. Finally I would like to thank Fred Dill who inspired me in my undergraduate Human Cytogenetics course to pursue genetics research. Thank you for staying with me and offering support and thesis edits even into your retirement.  XV  Dedication The completion of this thesis would not have been possible without the support of all of my family and friends who were always there for me. Through all of my ups and downs you were always there to support me and I am grateful for that. You know who you are.  1 Genetic Mechanisms of Nondisjunction in Humans Chapter 1: Introduction 1.1  OPENING REMARKS: In humans, missegregation of chromosomes at meiosis occurs at a high frequency -  the highest among all mammals yet observed. It is estimated that 15-20% of all pregnancies result in spontaneous abortion. At least 50% of these losses are chromosomally abnormal and more than half of these are trisomic, i.e. have an extra copy of one chromosome, making trisomy the most common chromosome abnormality in humans. Trisomy was first described in man in 1959 (Lejeune et al, 1959) and has been recognized as playing a major role in human genetic disease. Trisomy is caused by an error in meiosis such that homologous chromosomes do not disjoin (separate) properlyfromeach other. When both chromosomes of a pair or both sister chromatids of a chromosome move to the same pole together the resulting gamete will have two copies of that chromosome rather than one. These chromosomes have been said to have "nondisjoined". Despite the immense importance of meiotic nondisjunction (ND) to human reproduction and genetic disorders of newborns, little is known of its etiology. The only factor clearly associated with risk of trisomy is advanced maternal age (Hassold and Chiu, 1985). The risk of trisomy for women under 25 years of age is 2% and this dramatically increases with age to approach 35% in women over 40 years of age (Hassold and Hunt, 2001). The association between increasing maternal age and trisomy is arguably the most important etiological factor in human genetic disease (Hassold and Sherman, 2000). The relationship between recombination and chromosomal nondisjunction has been studied extensively along chromosome 21. There appears to be a relationship between altered recombination, both amount and placement, along the chromosome and nondisjunction for  2 21. Altered recombination has been linked to trisomy for other chromosomes and is clearly a major factor affecting N D . The goal of this project was to identify other factors that may affect, and be associated with, nondisjunction in humans. Specifically telomere length, replication timing and the levels of somatic aneuploidy (the gain or loss of whole chromosomes) were explored in women who had experienced a trisomy pregnancy, and compared to women of the same age who had not. I hypothesized that aneuploidy may be an effect of aging and that women experiencing trisomy would have shorter telomeres (due to increased rate of aging), would experience more replication asynchrony and increased levels of somatic aneuploidy when compared to control women aging at a normal rate.  1.2  AN OVERVIEW:  1.2.1 Meiosis: Meiosis is the process that takes place in the specialized tissues of the germline in both males and females. In humans, errors are found both in products of meiosis and products of conception. Errors that occur during the process of meiosis in the female ovary are responsible for the majority of the aneuploidy found in human pregnancies. 1.2.1.1 Timing: Meiosis is the process of producing cells with half the number of chromosomes. This process takes place in the germline during gametogenesis to produce eggs in females and sperm in males. In females, oogenesis, the process of forming eggs, takes place in the ovaries beginning in early fetal life. The ovary contains many follicles each of which contains a single egg (oocyte). At birth each female carries a lifetime supply of developing oocytes, and beginning at puberty, one oocyte is normally released once a month until menopause, which occurs when the threshold minimum of ~1000 oocytes is reached (Faddy  3  and Gosden, 1996). In males, spermatogenesis is the process of forming sperm in specialized gonads called testes. Human males produce 100-200 million sperm per ejaculate on a continuous basis throughout their lives beginning at puberty (Turnpenny and Ellard, 2005). Meiosis can be separated into two major cell divisions: meiosis I (MI) which is a reductional division, such that the chromosome number is reduced from the diploid to haploid number, and meiosis II (Mil) which is an equational division similar to mitosis. Each of these divisions can be further subdivided into four main stages: prophase, metaphase, anaphase and telophase. As shown in Figure 1-1, there is replication or doubling of the genetic material in each chromosome during the preceding S phase such that each homologue consists of two identical sister chromatids held together by proteins called cohesins. Prophase of MI (prophase I) can be even further subdivided into five main steps, each of which is outlined below. M i l follows the same main stages without the reduction, as the chromosome number is already haploid (Strachan and Read, 1999). Leptotene:  The replicated chromosomes begin to condense and become visible. The two  sister chromatids comprising each chromosome cannot be distinguished at this step. Zygotene:  After finding one another, the chromosomes pair with the help of an axial  element that contains cohesins. Each pair of homologous chromosomes aligns lengthwise with each other forming a bivalent, sometimes referred to as a tetrad because it contains four sister chromatids. The cohesins help to keep sister chromatids together so that they do not prematurely separate, i.e. before anaphase II, and similarly help keep homologous chromosomes together so that recombination can occur. During zygotene, synapsis, defined as the close pairing of homologues, occurs and the synaptonemal complex (SC) is formed. The SC consists of lateral elements located between the sister chromatids and a central  4  element connecting these lateral elements. Homologous exchange or recombination takes place across this structure and begins at this point in meiosis I, and also helps to keep the homologues together. Pachytene:  Synapsis is complete by pachytene and recombination nodules appear and are  thought to represent regions where recombination is occurring. After completion of recombination the SC begins to break down and the chromatids begin to pull apart. Diplotene:  Chiasmata are now visible. The two components of each bivalent begin to  repel each other, each homologue keeping the sister chromatids intact at their centromeres. This is the point at which meiosis is arrested (called dictyotene) in human females until puberty at which point a single egg will complete MI at each ovulatory cycle. In human males there is no arrest in meiosis. Diakinesis:  The chromosomes reach maximal condensation at this point, and are now  clearly visible. In order for the homologues to pair and align themselves in zygotene of prophase, they must first find each other. It is unknown exactly how this happens but two chromosomal regions have long been thought to play important roles in mediating early pairing: telomeres and centric heterochromatin (Walker and Hawley, 2000). Alignments of these two types of chromosomal domains may represent early steps in the process of pairing, and disruption of telomere-telomere attachments may impair eventual synapsis and recombination. The observation in the nucleus of the "telomere bouquet", which is the formation of a structure that looks like a bouquet of flowers formed by telomeres that appear to be interacting with each other, in the nucleus of cells demonstrates that this may be how chromosomes come into contact with one another in order to determine the level of similarity  5 necessary i n order for synapsis and recombination to occur. The meiotic "bouquet" arrangement is a well-established feature o f early prophase i n many organisms such as yeast, and was also found to be involved i n chromosome pairing i n both mouse and human male meiosis (Scherthan et al, 1996). Scherthan et al (1996) found that telomere movements were associated with the onset o f synaptic chromosome pairing using F I S H analysis o f tissue sections from both human and Mus musculus testis preparations. The centromere, universally denoted as the primary constriction, is essential for chromosomal attachment to the spindle and for proper segregation (Sullivan et al, 1996). Centromeres are genetically important elements present i n all eukaryotic chromosomes. These chromosome regions enable proper sister chromatid separation and chromosome segregation at mitosis and meiosis.  Figure 1-1:  Diagram o f meiosis.  A single c h r o m o s o m e p a i r w i t h a s i n g l e c r o s s o v e r is  s h o w n . T h i s leads to the f o r m a t i o n o f 4 distinct gametes.  7  1.2.1.2 Recombination: Crossing over is defined as reciprocal genetic exchange between homologous chromosomes. It is initiated by double-strand breaks and is a fundamental step in the process of meiosis. Chiasmata, the physical manifestation of crossing over, are thought to play a role in holding the homologues together during meiosis. In humans, at least one recombination event per chromosome arm is necessary for proper disjunction (Carpenter, 1984; Lawrie et al., 1995). As outlined in Figure 1-2, a crucial event in current models for recombination is the invasion of an intact DNA duplex by a single-stranded (ss) end of a broken DNA molecule (Haber, 1999). Many key proteins involved in recombination, such as RecA, topoisomerases, helicases, and DNA repair molecules, are highly conserved from yeast to man (Pittman and Schimenti, 1998). Analyses of recombination in humans have the limitation that the cells of interest are relatively inaccessible. A testicular biopsy is necessary in order to study males and analysis of females is even more difficult since recombination occurs in the fetal ovary (Hassold et al., 2000). New immunofluorescence techniques have made it possible to examine meiosis and recombination in more detail. Using an antibody to MLH1, a DNA mismatch repair protein that localizes to the sites of crossovers in pachytene-stage germ cells, Baker et al. (1996) observed recombination on all chromosomes. The observed number and location of the foci, defined as areas of high concentrations of MLH1, correlated with previous cytogenetic analyses of chiasmata suggesting that MLH1 contributed to the formation and/or processing of meiotic recombination events (Baker et al., 1996; Hassold et al, 2004). Anderson et al. (1999) conducted immunostaining studies of male mouse pachytene preparations using antibodies to both MLH1 and SCP3, a component of the lateral element of the SC. Their  8 observations were consistent with the expected number, placement and the non-random distribution resultingfrompositive interference found in previous studies. Positive interference is defined as a reduction in the likelihood of a second exchange occurring near to the first one. Finally, Barlow and Hulten (1998) used a similar method to look at MLH1 foci in spermatocytes and in oocytes of a female fetus. Overall -50 autosomal foci were seen per nucleus in males while the total number of foci was much higher (mean = 95) in females. In both the mouse and human studies, distal foci were much more common in males. Both of these findings are consistent with data on recombination differences in females and males. There is approximately 90% more recombination in females, but males have more recombination near telomeric regions (Donis-Keller et al, 1987; Robinson, 1996; Hubert et al., 1994). Recent studies using MLH1 have found that -50 crossovers per cell are found in spermatocytes in comparison to more than 70 in oocytes (Tease and Hulten, 2004). This same group also suggests that the longer synaptonemal complex in the oocytes may contribute to the increased recombination in females.  Figure 1-2: Recombination heteroduplex, formed when a single strand (represented by the dotted lines) from a broken duplex "invades" an intact double- stranded DNA duplex.  9 Distribution of Meiotic Exchanges: '  Interference causes the meiotic recombination events to be distributed nonrandomly  along the chromosome. In yeast, loss of a gene encoding the SC protein Zipl resulted in a loss of interference accompanied by an increase in aneuploidy (Sym and Roeder, 1994). The approximate sex-averaged relationship between genetic and physical maps in humans is >1 cM/Mb and 0.5 cM/Mb in mice (Weissenbach et al, 1992; Copeland et al, 1993; Pittman and Schimenti, 1998). Recombination frequencies per megabase in humans are extremely low compared to the yeast S. cerevisiae (370 cM/Mb) (Symington et al, 1991). Crossovers are not distributed randomly across the chromosomes in any organism. In yeast, the regions that undergo high levels of recombination are generally found near promoter regions and correspond with the positions of double strand break sites (Ohta et al, 1994; Wu and Lichten, 1994). Recombination "hotspots", or regions of high recombination, are also present in mammals and an extensively studied one is the MHC in mouse (Steinmetz et al, 1982). Both male and female hotspots have been reported in humans as well (Robinson and Lalande, 1995; Hubert et al, 1994; Paldi et al, 1995; Jeffreys et al, 2000; Lercher and Hurst, 2003; Arnheim et al, 2003). C. elegans is the only species in which a gene affecting the distribution of meiotic recombination events has been identified (Rose and Baillie 1979; Zetka and Rose 1995). Recombination is very important for the proper disjunction of meiotic chromosomes. In this project I analyzed patterns of recombination along chromosome 15. I wanted to extend previous analyses of patterns of recombination along normal vs. nondisjoined chromosomes, as well as at the common Prader-Willi syndrome (PWS) and Angelman syndrome (AS) deletion breakpoints in the proximal region. The details of these studies will  10 be discussed in Chapter 7. Recombination may contribute to many chromosome rearrangements and high levels have been observed at deletion and duplication breakpoints. 1.2.1.3 Chromosome cohesion and segregation: Throughout the process of recombination, both sister chromatids and homologues need to maintain cohesion. This cohesion is later released so chromosomes (at MI) and chromatids (at Mil) can segregate. Both premature and late separation of chromatids/ chromosomes can lead to chromosome nondisjunction. There are many structures involved in this intricate process, regulated by a number of genes and proteins. Chiasmata help keep the homologues attached to each other while cohesins keep the sister chromatids together. At metaphase I, the homologous chromosome pairs line up along the metaphase plate and the microtubules of the spindle fibers attach to the sister kinetochores. They attach to one homologue, pulling both sister chromatids toward one pole of the cell, while the other sister kinetochores pull the other homologue toward the opposite pole. In order for this to happen, the kinetochores, defined as a proteinaceous structure found at the a-satellite sequences of the centromere, on sister chromatids must lie side-by-side as outlined in Figure 1-3. Balanced tension between opposite forces on the two active kinetochores by the spindle is necessary in order for meiosis to progress, meaning that all chromosomes must be lined up properly on the microtubules (Wolstenholme and Angell, 2000). The spindle assembly checkpoint proteins monitor this. In males an imbalance will result in arrest of meiosis and cell death (LeMaire-Adkins et al, 1997 - mice; Handel et al, 1999; Roeder and Bailis, 2000 - human). In females however, an imbalance can occur without arrest and has even been proposed to be the main mechanism of trisomy (Angell, 1997; LeMaire-Adkins et al, 1997; reviewed in Hunt and Hassold, 2002). This may be particularly true in older  11  women as a result of a breakdown of the proteins responsible for cohesion along the chromosome arms, leading to univalents or unpaired chromosomes that segregate independently at metaphase.  Figure 1-3: The kinetochores of sister chromatids are side-by-side at MI, and face the same pole. This results in both sister chromatids moving to the same pole in anaphase I. Adapted from Paliulis and Nicklas, 2000.  12 The final stage in MI is anaphase I, where the cohesion between the chromosome arms breaks down allowing the chiasmata to resolve. Cohesin must remain associated with the centromeres, however, holding the sister chromatids together until anaphase II. The recombined homologues then separate and migrate toward their respective poles. At metaphase II, the chromosomes line up along the metaphase plate, but this time the sister kinetochores are back-to-back so that they capture microtubules from opposite poles and then move to opposite poles in anaphase II (Figure 1-4). They do not have a pairing partner and there are no chiasmata holding the arms of the sister chromatids together. It is the cohesion at the centromere that keeps the chromosome intact until anaphase II.  Figure 1-4: The kinetochores of the sister chromatids are back-to-back at M i l , and face opposite poles. This results in the sister chromatids moving to opposite poles in anaphase II. Adapted from Paliulis and Nicklas, 2000.  Sex-Specific Differences: Meiosis differs in males and females in several ways, some of which have already been mentioned, and are summarized in Table 1-1. 1) Timing: in females meiosis begins for all oocytes in fetal life while meiosis begins in puberty in males and is an ongoing process with a new meiosis beginning every day. This means that females are born with all of the follicles that they will ever have (Faddy and Gosden, 1996) - no more can be produced - while males produce sperm daily. Meiosis takes up to 40 years in females while ~64 days is all that is necessary to complete the entire process of spermatogenesis - only one week is needed to progress from prophase I through the second division (Hunt and Hassold, 2002). 2) Meiotic arrest: Meiosis is arrested in dictyotene in females until ovulation and then again in meiosis II until fertilization, while male meiosis is continuous and there are no points of arrest. 3) Meiotic checkpoints: Interestingly, males have checkpoints throughout meiosis where integrity is monitored. If recombination has not occurred (pachytene checkpoint) or the tension on the spindle is not correct (the spindle assembly checkpoint), male gametes will arrest and die. There is now evidence from both human and mouse that suggests that these checkpoints are less stringent in mammalian oogenesis (Yin et al., 1998; Hodges et al., 2002). 4) Recombination: Sex-specific differences in recombinationfrequencyare also found in humans and other organisms. It is thought that this may be due to the more condensed state of the male chromosomes since double strand breaks, the initiating sites of recombination, occur preferentially in regions of open chromatin. Early studies noted sex differences in recombination in mammals, with females generally having higher rates than males (Dunn and Bennett, 1967). 5) Number of products: lastly, in females there is only  14 one product, while in males, four sperm are produced for every diploid cell that begins meiosis.  Table 1-1: The main differences between males and females in meiosis.  Timing  Meiotic Arrest  Meiotic Checkpoints  Recombination  Number of products  Males  Females  Meiosis begins at puberty and is continuous throughout life No meiotic arrest  Meiosis begins in fetal life and ends before birth  Pachytene and spindle assembly checkpoints monitor integrity Genetic map is shorter than in females - more condensed chromosomes 4 products (sperm)/diploid cell beginning meiosis; millions produced per fertilization event  Meiosis is arrested in dictyotene until ovulation; then again in M i l until fertilization Checkpoints are not stringent Genetic map is 90% longer than in males 1 product (egg)+3 polar bodies/diploid cell beginning meiosis; one egg released per month/fertilization event  15 Models of human meiosis derived from other organisms: In yeast, the main steps of meiosis may be slightly different than that seen for humans. However, in Saccharomyces cerevisiae meiosis and mitosis have been extensively studied. The study of meiotic mutants in this organism has increased considerably our understanding of the process of meiosis. Studies in yeast have identified a number of different classes of mutants including: those affecting recombination (e.g. the rec, mei and rad classes of genes, spol 1, DMC1, Msh4, Msh5 and Mlhl) (Lee et al., 2005), the synaptonemal complex (e.g. Scp3, ZIP1, redl, merl and hopl), cell division control (the cdc class of genes, A M and spd) (Tonami et al., 2005), meiotic checkpoint control (MAD1, 2 &3, BUB1 &3 and MSP1) (Rancati et al, 2005), and sister chromatid cohesion (Mei-S332 and Rec8) (Tang et al, 1998). Homologies of yeast genes have been found in Drosophila melanogaster, mice and humans. In a recent study by Baker et al. (2004), a mitotic spindle assembly checkpoint protein BubRl, encoded by Bublb, was knocked out in mice to determine its physiological role. After producing a series of mice with reduced BubRl expression they observed that many of the mice had developed tumors and the median lifespan of the mutants was ~6 months compared to >15 months in the wild-type mice. The mutant mice showed multiple age-associated phenotypes early in life such as cataracts, loss of subcutaneous fat and impaired wound healing. In culture, the mouse embryofibroblastshad problems with the spindle assembly checkpoint and had increased amounts of aneuploidy. Premature sister chromatid separation was also observed, which is a hallmark of a defective assembly checkpoint. Both males and females were also infertile due to meiotic chromosome segregation defects. Specifically in females, many abnormal metaphase II configurations  16 were seen in arrested oocytes, similar to that associated with reproductive aging in humans. Therefore, a human homologue of BubRl may be involved in regulating aging and infertility as well. 1.2.2 Mitosis: Mitosis is a process of cell division by which diploid cells give rise to diploid daughter cells, or haploid cells give rise to haploid daughter cells, producing exact copies of themselves. The major difference between mitosis and meiosis is that there is no pairing of the homologous chromosomes. This also means that there is not normally exchange of genetic material through recombination in mitosis. For this reason, mitosis is said to be quite similar to MIL Table 1-2 outlines the differences between meiosis and mitosis. 1.2.2.1 Timing: Mitosis takes place on a daily basis in most somatic cells of our developing body in order to replace cells that are lost through apoptosis and in order for growth to take place. Mitosis is necessary for hair growth, nail growth and every developmental step after fertilization for creation of an embryo. Mitosis is also necessary for the steps leading up to meiosis in both males and females. Specifically, in males spermatogenesis involves many mitotic divisions, possibly as many as 20-25 per year. At puberty spermatogonia will already have undergone - 3 0 mitotic divisions when they begin to mature into primary spermatocytes which enter into meiosis I (Turnpenny and Ellard, 2005). In females oogonia originate from primordial germ cells by a process that involves 20-30 mitotic divisions. At 3 months of intrauterine life, the oogonia begin to mature into primary oocytes that start to undergo meiosis (Turnpenny and Ellard, 2005).  17  1.2.2.2 Recombination: Recombination can and does occur during mitosis but it is a rare event compared to the amount of recombination in meiosis. However, it is also more difficult to ascertain and measure. Double strand breaks (DSBs) are necessary in order to initiate recombination, but they are also the most deleterious form of DNA lesion as they can result in chromosomal breakage and rearrangements (Morrison and Takeda, 2000). Two major repair pathways are in place to deal with DSBs - non-homologous end-joining and homologous recombination (Kanaar et al., 1998; Haber, 1999). Non-homologous end-joining repairs broken DNA ends without the requirement for extensive sequence homology. Homologous recombination however, is more accurate and requires an intact homologous chromosome or a sister chromatid to repair the break. 1.2.2.3 Chromosome cohesion and segregation: In metaphase of mitosis, the spindle fibers attach to the kinetochores that appear at the centromere of each chromatid, and one kinetochore of each sister chromatid is attached to an opposite pole. The sister kinetochores separate and each chromatid moves to its respective side of the cell. In order for sister chromatids to separate, cohesin, which has been holding them together, must breakdown. This is accomplished by a protease called separase that becomes active in late metaphase (Revenkova et al., 2004).  18  Table 1-2: The main differences between meiosis and mitosis. Meiosis Timing/cells involved  Recombination  Chromosome condensation and segregation  Ploidy of cells at start of process Ploidy of cells at end of process  Mitosis  Frequent in many somatic In fetal life in females in cells to replace dead cells. ovaries to create oocytes; continuously in males in testes to create spermatocytes. Rare event. Occurs in all nuclei of both oocytes and sperm - at least one event between each pair of chromosomes. Spindle fibers attach to Meiosis I: spindle fibers kinetochores of each sister attach to kinetochores of chromatid - to opposite each homologue - sister poles. chromatids remain attached. Homologues to opposite poles. Meiosis II: spindle fibers attach to kinetochores of each sister chromatid, as in mitosis and move to opposite poles. Diploid or haploid cells. Diploid cells. Haploid germ cells.  Same as start of process.  19 1.3  ANEUPLOIDY:  Although, meiosis is a precise and highly organized process that is necessary for sexual reproduction, errors appear to be common in human female meiosis. One such outcome is aberrant chromosome segregation or nondisjunction. Nondisjunction leaves gametes with a gain or a loss of chromosomes resulting in conceptions with a gain or a loss of a chromosome or chromosomes (aneuploidy). This common chromosome anomaly found at human conception occurs in at least 5% of all clinically recognized pregnancies. Most aneuploidies do not survive to term meaning that trisomy (the gain of a whole chromosome) and monosomy (the loss of a whole chromosome) are the leading causes of pregnancy loss (Hassold and Hunt, 2001). It is estimated that 15-20% of all pregnancies result in a loss. Chromosome abnormalities of various kinds occur quite frequently in humans and can be found in 10 to 30% of all fertilized eggs. Specifically, 20% of human eggs are aneuploid while 2-5% of spermatocytes have an abnormal number of chromosomes (Jacobs, 1992; Martin et al., 1996; Brandriff et al., 1984). In other organisms, malsegregation of a chromosome is less common. In Saccharomyces cerevisiae the frequency of aneuploidy is as low as 1 in 10,000 per meiosis, while in Drosophila melanogaster X-chromosome nondisjunction in females ranges from 1 in 1,700 to 1 in 6,000 (Hassold and Hunt, 2001). In another well-studied mammal, the mouse, the overall incidence of aneuploidy among fertilized eggs is 1-2% (Yamamoto et al., 1973). Studying the frequency of chromosome abnormalities in humans is difficult, as all developmental stages cannot be studied. Data comes mainlyfromstudies of clinically recognized pregnancies and studies of gametes. Many laboratories have cytogenetically analyzed spontaneous abortions (SAs) to determine the amount of aneuploidy found in early  20 losses (6-20 weeks gestation). Over 35% of all aborted karyotyped fetuses/embryos have been found to be aneuploid, and the chromosomes involved is very different than that seen in liveborns (Hassold and Hunt, 2001). Trisomies for almost all chromosomes have now been found in SAs, the most common being trisomy 16, which accounts for 1/3 of all trisomies (Hassold etal, 1996). 1.3.1 Trisomy: Trisomy, first described in man in 1959 (Lejeune et al, 1959), is the presence of a third copy of one chromosome in the nucleus and is the most common chromosome abnormality in humans. At least 50% of pregnancy losses are chromosomally abnormal and more than half of these are trisomic. Trisomy 16 is estimated to occur in more than 1% of clinically recognized pregnancies, making it the most commonly occurring trisomy in humans (Hassold et al, 1996). Full trisomy 16 normally results in miscarriage in the first trimester of pregnancy (Warburton, 1987). Only trisomies for chromosomes 13, 18,21 and X can survive to term, although most do not (Warburton, 1987). Trisomy 21 is the most common trisomy found in liveborns making up -1/700 births, and is the most common cause of mental retardation (MR) in humans (Griffin, 1996; Mikkelsen, 1988). The most common origin of a trisomic conceptus differs depending on the chromosome involved. In general, trisomy most often results from ND of chromosomes in meiosis I during oogenesis (Lamb et al, 1996; Hassold and Hunt, 2001). Early somatic errors can also result in mosaic trisomy, but are less common. Trisomy 16 almost exclusively results from a maternal MI error (Risch et al, 1986; Hassold et al, 1995). However, for trisomy 18, the majority of errors occur at meiosis II and for 47, X X Y and trisomy 2 approximately 50% are due to a paternal error (Abruzzo and Hassold, 1995;  21 Robinson et al, 1999). Monosomy X or Turner syndrome is thought to be largely (70-80%) due to ND in the male by producing nullisomic sperm for the sex chromosomes (Hassold and Pettay, 1992). Whether or not this loss of the other sex chromosome occurs in sperm or postzygotically is unknown and is a topic of debate. It seems likely, therefore, that many factors that affect nondisjunction are chromosome-specific. 1.3.1.1 Trisomy mosaicism: Although the majority of trisomies are lost as early SAs, many can survive to term in a mosaic state. Chromosomal mosaicism is the presence of two different cell lines, with two different chromosomal make-ups, in an individual who has developed from a single fertilized egg. Mosaicism that is most commonly found at prenatal diagnosis involves one cell line that has a normal chromosome complement, 46, X X or 46, X Y , and another cell line that is trisomic; called trisomy mosaicism (Strachan and Read, 1999). The trisomic cell line in a trisomy mosaic can arise either through a meiotic mechanism or a somatic mechanism (Kalousek and Vekemans, 2000). Marker studies have suggested that most mosaics begin as trisomic zygotes and subsequently lose an extra chromosome, thereby gaining a normal cell line (Hassold, 1985). However, Robinson et al. (1997 and 1999) found that most are somatic in origin and origin is dependent on the chromosome involved. The number of cells present at the time of the event determines the phenotype of the individual. In general, a meiotic origin is correlated with high levels of trisomy in the individual while a somatic origin is correlated with lower levels (Robinson et al, 1997). The abnormal cell line may be found in multiple tissues but sometimes is only found in one (Stavropoulos et al, 1998). Mosaicism is found in 1 -2% of all chorionic villous sampling (CVS) procedures done (Mikkelsen and Ayme, 1987; Ledbetter et al, 1992; Wang et al, 1993; Kalousek and Vekemans, 2000), and  22 often the trisomy can only be found in the placenta; a phenomenon referred to as "confined placental mosaicism" (CPM) (Kalousek and Dill, 1983). The effect that trisomy mosaicism has on an individual varies tremendously, such that the fetus may be abnormal, small for gestational age or be completely normal (Kalousek and Vekemans, 2000). 1.3.1.2 Origin of mosaicism: If a conceptus began life as a trisomy and subsequently lost the third chromosome as a result of anaphase lag, this is referred to as trisomic rescue (Figure l-5a). A mitotic nondisjunction event leading to a cell that has 47 chromosomes is also possible (Figure 15b). A very early mitotic mistake could lead to generalized mosaicism as possibly up to half of the cells will be trisomic. If this event occurs a little later, there is a better possibility of CPM. At the 64-cell stage of an embryo, only 3-5 cells are dedicated to the formation of the embryo proper while the rest of the cells will make up the extraembryonic tissue (Gardner and Lyon, 1971). It seems more likely then that the mistake would happen in the cells destined to become the placenta. However, confined embryonic mosaicism may also exist such that only the embryo contains the trisomic cells. It is important to realize that trisomy mosaicism may result in an underestimation of the rate of meiotic errors leading to trisomy. A rescued trisomy may only have trisomic cells in the germline for example and go undetected as an error. Mosaicism in the germline can go undetected and can lead to recurrent trisomy and recurrent pregnancy loss. However, Robinson et al, (2001) found that this is rarely the explanation for recurrence and has only been demonstrated to occur for recurrence of trisomy 18 or 21. It may be that other aneuploidies simply lead to follicular atresia when present in the female germline (Robinson et al, 2001). Germline mosaicism is  thought to account for the majority of the recurrence for trisomy 21 (Pangalos et al., 1992; Bruyere et al, 2000).  /  \  las  dj) dD  m  Mosaic -17 / 46 chromosomes  Figure l-5a: Mosaicism can result from anaphase lag in meiosis. This diagram shows loss of one of the three copies of the chromosome at the 2 meiotic division resulting in two cell lines; one with 46 chromosomes and one with 47. This is also called "trisomic rescue". The extra chromosome is isolated in a separate cell and is eventually lost. This diagram taken from http://www.medgen.ubc.ca/wrobinson/mosaic/index.htm. nd  24  Mosaic 47  46 chromosomes  Figure l - 5 b : Mosaicism can also arise during a nondisjunction event in mitosis. A cell with two copies of a chromosome gives rise to one cell with three copies and another with only one copy of that chromosome. Here the cell with the three copies of the same chromosome will continue to grow, but the cell with only one copy of the chromosome will not. This diagram taken from http://www.medgen.ubc.ca/wrobinson/mosaic/index.htm.  25 1.3.1.3 Uniparental disomy: Uniparental disomy (UPD) is the presence of a normal chromosome pair, both copies of which have come from the same parent. UPD can result in clinical conditions in humans as a result of either producing homozygosity for recessive mutations or abnormal patterns of imprinting (Robinson, 2000). Isodisomy refers to regions of the chromosome derived from identical sister chromatids while heterodisomy refers to regions of the chromosome derived from homologous chromosomes. This can result in a number of ways and is often found as a consequence of trisomic rescue. If in a trisomic conceptus of maternal origin the chromosome that is lost during the rescue is the paternal chromosome, then the resulting individual or conceptus has maternal UPD for that chromosome - both chromosomes are from the mother and there is no paternal contribution. The distribution of isodisomic and heterodisomic regions depends on the stage of origin. If the error occurred at MI, there will be heterodisomy at the centromere, as the chromosomes are from the two different homologues. If the error occurred at M i l , there will be isodisomy at the centromere, as the chromosomes are sister chromatids. There can be regions of both isodisomy and heterodisomy along the chromosome, however, as a result of the recombination that took place in gametogenesis (Robinson, 2000). UPD is generally a benign situation with no adverse phenotypic consequences to the individual (Ledbetter and Engel, 1995; Kotzot, 1999). However, there are a few chromosomes for which UPD has detrimental consequences. These chromosomes contain imprinted genes, which are differentially expressed depending on the parent-of-origin and this means that both the maternal and the paternal contribution are required in order for proper development to take place. A good example of a chromosome that contains imprinted  26 genes is 15. Maternal UPD 15 results in Prader-Willi syndrome (PWS) while paternal UPD 15 results in the entirely different Angelman syndrome (AS). This is due to the loss of paternally expressed genes in mat UPD 15 and the loss of a maternally expressed gene in pat UPD15 (Robinson et al, 2000). Other chromosomes for which UPD is known to cause i  phenotypic abnormalities due to loss of expression of imprinted genes are chromosomes 6, 7, 11 and 14 - more than 40 imprinted genes have been described throughout the genome in humans (Jorde et al., 2003). 1.3.2  Monosomy: Monosomy is the absence of one copy of a chromosome pair resulting in a total of 45  chromosomes. Autosomal monosomies are early embryonic lethals - so early in fact that monosomy 21 was once the only autosomal monosomy that had been described in fetal deaths (Warburton, 1987). Sex chromosome monosomy (45,X), however, is one of the most common chromosome abnormalities found in spontaneous abortions (Warburton, 1987). The vast majority of 45,X do not survive (99.5%) and they make up -10% of all spontaneous abortions (Warburton, 1987). Even partial monosomy in the form of large deletions cannot readily be tolerated. For each ND event that produces a gamete that is disomic rather than monosomic for a particular chromosome, there should be a complementary gamete that is nullisomic for that chromosome. Upon fertilization with a normal gamete with the complete chromosome complement, this should result in the same number of trisomies and monosomies. Evidence from analyses of human sperm after fertilization or penetration of hamster eggs indicates that monosomies are just asfrequentas trisomies (Brandriff et al, 1984). However, trisomies are observed far morefrequentlythan monosomies in spontaneous abortions and pregnancy  27  losses. This is most likely due to the fact that monosomies are far less tolerated during development and therefore monosomic embryos are lost much earlier during pregnancy. More recent karyotype studies of human diploid embryos fertilized in vitro (Jamieson et al., 1994), as well as fluorescence in situ hybridization (FISH) studies used to examine human preimplantation embryos (Murine et al., 1995) support this theoretical concept. 1.3.3 Segmental aneuploidy: Segmental aneuploidy is represented in the form of unbalanced translocations. A duplication or a deletion that results in either a loss or gain of a segment of a chromosome, such that there are three copies or one copy of that region leads to segmental trisomy and segmental monosomy respectively. Unequal recombination can result in both deletions and duplications and is a major mechanism of genome evolution (Figure 1-6). During meiosis, chromosomes with repetitive sequences can pair incorrectly so that one repeat element of one chromatid pairs with a different element in another chromatid. Recombination can then lead to an increase in the number of repeats in one chromatid and a decrease in the other.  Figure 1-6: Unequal recombination between misaligned sister chromatids or homologous chromosomes containing highly homologous sequences leads to both deletion and duplication products. This is seen here, where there is either one copy or three copies of the sequence represented by the arrows rather than two.  28 1.3.3.1 Deletions: A chromosomal deletion occurs when the chromosome breaks in two places and the intervening piece is lost. This involves loss of genetic information and results in "partial monosomy" for that chromosome. Microdeletion syndromes are caused by large deletions that knock-out numerous genes creating a specific recognizable phenotype. Some examples are PWS, AS and cri-du-chat - deletions on chromosome 15ql l-ql3 (paternal), 15ql l-ql3 (maternal) and 5p, respectively (Strachan and Read, 1999). Recurrent deletions are of special interest as the breakpoints are usually in the same place on the chromosome in different people (Strachan and Read, 1999; Robinson et al, 1991). Repetitive sequences, duplicons (large blocks of duplicated genes of DNA sequence), and other sequences have been implicated as the catalyst for such breaks and this idea is explored in Chapter 7 for the common deletion breakpoints on chromosome 15. 1.3.3.2 Duplications: Chromosomes that contain reiterated segments are said to have duplications. Typically, these duplicated segments are present in an end to end arrangement and are called tandem duplications. The number of duplicated elements in a region can change through unequal crossing-over, as described above. The major catalyst of illegitimate exchange is repetitive sequences, especially Alu-like repeats. There is an Alu sequence every 6 kb in the human genome (Pittman and Schimenti, 1998). Whether or not a phenotype is associated with a duplicated region depends on many factors. The size of the duplicated region, the function of the genes, and the location of the new segment are some factors to be considered. Duplicons have recently been shown to be involved in abnormal recombination resulting in deletions and other rearrangements (Amos-Landgraf et al, 1999). Duplicated sequences can  29 diverge to perform related but specialized developmental functions. The human P-globin gene family is an example of specialized genes that developed as a result of such duplication. Redundant genes and sequences can evolve into new genes with unique but similar or related functions. Of all of these forms of aneuploidy, the focus of this project is trisomy. 1.4  FACTORS THAT MAY AFFECT RISK FOR TRISOMY: Despite the immense importance of meiotic nondisjunction to human reproduction  and genetic disorders of newborns, little is known of its etiology. The association between increasing maternal age and trisomy, however, has been well documented. Altered recombination has also been shown to be associated with most trisomies studied to date. Other factors have been studied to look for associations or causal roles in ND including mitochondrial mutations, replication timing, centromere size, gene mutations and environmental factors such as smoking, diet, and oral contraceptives. No clear association with these latter factors has been found to date. 1.4.1  Review of Epidemiology: Are some women predisposed to trisomy? This is a question with important clinical  implications as we could then ask whether we can predict women at increased risk for trisomy and thus use the estimated risk of a trisomic pregnancy as an indication for prenatal diagnosis. This is the main focus of this project. Many different groups have studied this question extensively, and opinions have changed over the years (Warburton et al, 1987; Warburton et al, 2004). Early studies suggested an association between the occurrence of a trisomic spontaneous abortion (SA) and trisomy in either a prior or in subsequent live births (Alberman et al, 1975; Alberman, 1981). However, Warburton (1985) could find no such association. Considering all of the available data to date, Warburton et al. (1987) again  30 concluded that there was no predictive value of a chromosomally abnormal SA on subsequent pregnancies. This group also suggested that the search for common factors leading to a general increase in the rate of meiotic nondisjunction among trisomy-prone couples was likely to be fruitless. So why then almost two decades later are we still investigating this question? Some studies have indeed shown that women having a trisomy 21 birth at a young age (<30 years) belong to a group with an increased risk for a subsequent trisomic pregnancy (Stene, 1970; Hook, 1981; Mikkelsen and Stene, 1979; Stene etal, 1984). This was considered to be due to mosaicism for trisomy 21 in a small number of couples (Pangalos et al, 1992; Bruyere et al, 2000). There are also many cases of younger women having multiple trisomies and it has been suggested that this group may contain women at risk for a subsequent trisomic pregnancy (Warburton and Kinney, 1996). Most recently, Warburton et al (2004) reconsidered the recurrent trisomy data and concluded that there is a significant increased risk for a trisomic pregnancy following a trisomic pregnancy. Their data came from karyotype results from prenatal diagnoses. They compared the number of trisomies at prenatal diagnosis with the expected number, given maternal age-specific rates and calculated a standardized morbidity ratio (SMR). For women having had a previous trisomy 21 pregnancy, the SMR for having a second trisomy 21 (homotrisomy) varied depending on maternal age. When both trisomies occurred under the age of 30 years, the SMR was 8.2 (p=0.002), confirming previous findings of a large increased risk for recurrence of the same trisomy after a trisomy 21 at this young age. When both trisomies occurred at 30 years of age or older the SMR was 2.1 (p=0.005), indicating a doubling of risk among older women. A doubling of risk, SMR=2.3 (p=0.007), was also found for all ages for having a different  31 trisomy after having a trisomy 21 (heterotrisomy). For the other trisomies that can survive to term (13, 18, X X X and XXY) the SMR for having the same trisomy recur was 2.5 (n.s.). When all viable trisomies were combined, i.e. including trisomy 21, the SMR for heterotrisomy was 1.6 (p=0.04). After a nonviable trisomy the SMR was 1.8 (p=0.04). This significantly increased risk for heterotrisomy supports the hypothesis that some women have a risk for ND higher than do others of the same age. 1.4.1.1 Aging: Increasing maternal age is the only well documented factor known to be associated with trisomy risk. As maternal age increases, risk for trisomy increases as well exponentially after the age of 30 (Warburton, 1987). Penrose recognized its association with increasing maternal age long before it was determined that Down syndrome was caused by trisomy 21 (Penrose, 1951). This increased risk is very dramatic, with women under the age of 25 years having a 2% risk of trisomy while in women over 40 years of age this risk increases to 35% (Hunt and Hassold, 2002). In the 1980's, it was suggested that it was the ability to recognize and abort trisomic fetuses that declined with age, not necessarily that the trisomic pregnancies themselves increased in frequency with maternal age (Ayme and Lippman-Hand, 1982). This model does not explain why it is solely trisomies of maternal origin that are subject to the maternal age-effect, and it was never supported by data. This means that it cannot be the uterine environment but it is rather a problem with the egg itself. There is no known influence of race, geography, or socio-economic status on maternal agespecific rates of trisomy (Hassold and Hunt, 2001). However these have not been well studied. A recent study has shown an association between low socio-economic status and an increased risk for the mother to have a clinically recognized DS pregnancy (Christianson et  32 al., 2004). This same group also found that the case mothers were more likely than control mothers to be Hispanic. 1.4.1.2 Recombination: While it is obvious that maternal age is important for risk of trisomy, the question is whether other factors may cause variability in risk at a given maternal age. This is the idea behind the "two-hit" theory of recombination and trisomy (Lamb et al., 1996). With this theory Lamb proposed that there are susceptible configurations of crossovers that may not be easily resolved in the older oocyte. The first "hit" is the susceptible meiotic configuration and the second "hit" is disruption of a meiotic process that increases the risk for ND. It is this second hit that is the source of the maternal age effect on trisomy (Lamb et al., 1996). In younger mothers, the probability of the second hit is lower whereas with increasing age this probability increases. Meiotic-specific proteins may degrade over time, spindle fibers may break down, and mitochondrial mutations may accumulate. Therefore, there is a higher risk of a trisomic conceptus resulting from an oocyte of an older woman if there is also altered recombination. Recombination is thus not the cause of the ND but is merely correlated with ND. It is easy to see that by this same logic there must be relatively young women that either did or did not have altered recombination and have trisomic conceptuses. What else other than recombination then is making them more prone to trisomy at an earlier age? I envision a "multiple-hit" theory of ND in humans such that recombination or some other factor may set up a susceptible meiotic configuration in fetal life that increases the risk slightly for trisomy. Premature chromosomal aging as evidenced by telomere shortening and asynchronous replication may then further contribute to increasing the risk.  33 A crossover, or the exchange of genetic material between homologous chromosomes, is generally necessary for proper segregation of homologues in organisms in which recombination is normally found. One exchange per chromosome arm is required minimally in order for the proper segregation of human chromosomes to occur (Carpenter, 1984; Hawley, 1988; Lawrie et al., 1995). Back-up systems exist in Drosophila (distributive segregation) and a few mammalian species in which the sex chromosomes are achiasmate (Hawley et al, 1993; Simchen and Hugerat, 1993; Wolf, 1993). It has been proposed that either similar back-up systems operate in humans or that humans do not require such a system as normal meioses rarely involve nonexchange bivalents (Koehler and Hassold, 1998). There are many mutations in yeast and flies that reduce or eliminate exchanges and result in high frequencies of missegregation during the reductional division (Jones, 1987; Hawley, 1988). Thefirstdirect evidence of an association between reduced recombination and human trisomy came from studies by Warren et al. (1987) in which 34 Down syndrome (DS) families were analyzed. They reported reduced levels of recombination along chromosome 21 in meioses leading to trisomy 21. Ten years later Lamb et al. (1997) also found supporting evidence after using centromere mapping techniques, i.e. polymorphic genetic marker analysis, to compare recombination in normal meioses and in meioses leading to trisomy. Specifically they estimated that ~35% of trisomy 21 involved meioses with achiasmate bivalents. However, they also found chiasmate bivalents and the majority of these had a single exchange in distal 21q. This led to the idea that the placement of the exchanges may also be crucial for proper disjunction of chromosomes. For chromosome 21 it appears that distal exchanges are less efficient at segregating chromosomes than more  34 medially placed exchanges. The concept that alterations in recombination are determinants of human trisomy has become a central one to the study of causes of human trisomy. MI derived trisomies of chromosome 15 (and UPD15), 16 and 18 as well as sex chromosome aneuploidies, are all associated with a reduction in recombination (Lamb et al, 1997; Koehler et al, 1996; Thomas et al, 2000; Robinson et al, 1998; Bugge et al, 1998; Koehler etal, 1998). However, the effect varies considerably among the chromosomes. For example, achiasmate events do not appear to be important for trisomy 16 (Hassold et al, 2000), while they make up 20-40% of MI trisomies 15, 18 and 21 and sex chromosome aneuploidies (Hassold et al, 2000). The placement of exchanges on the other hand appears to be important for trisomies 16 and 21 but has not been found for any other trisomies (Koehler et al, 1998). For this reason, although trisomy 21 and recombination has been studied extensively, it cannot be used as a model for all human ND, as there may be chromosome-specific effects. The relationship between maternal age and altered recombination is difficult to reconcile because of the great discrepancy between time of action (Abruzzo and Hassold, 1995). From the time in fetal life when recombination occurs until conception can be as much as 40 years. Henderson and Edwards proposed the "production line" hypothesis in 1968. It suggested a link between recombination and human nondisjunction such that declining levels of recombination were the cause of the maternal age effect on trisomy (Henderson and Edwards, 1968). They hypothesized that meiotic chromosomes of older women were held together by fewer chiasmata, and that the chromosomes were therefore more likely to nondisjoin at MI. This was based on their observation of an increased incidence of univalents in MI oocytes from aged female mice. The results of direct tests of  35 the hypothesis are incompatible with its predictions - it is now clear that this model cannot explain the maternal age effect on human trisomy (Hassold et al, 2000). Specifically, oocytes from both young and old mice were obtained and the incidence of univalents (Speed and Chandley, 1983; Sugawara and Mikamo, 1986) and the amount of recombination (Beermann et al, 1987) did not support the theory. However, in a study of control women, Kong et al. (2004) found that there was an increase in recombination rate with age among their liveborn offspring and that women with high recombination rates tended to have more children. They suggest that the increased recombination events may provide protection from age-related meiotic breakdown. Recombinationfrequencyhas been shown to be associated with maternal age for chromosome 15 nondisjunction (Robinson et al, 1998) and more recently recombination pattern was shown to be associated with maternal age for chromosome 21 nondisjunction (Lamb et al, 2005). Specifically, lower recombination is observed among nondisjunction events taking place in younger women. The nondisjunction process is similar in women of different ages, and the increase in trisomy with age may be the result of a decrease in efficiency for processing certain types of exchange configurations in the older ovary (Hassold et al. 2000). A central goal of this project is to determine what factors may cause the decrease in efficiency in the older oocyte. 1.4.2  Ovarian aging: It has been proposed that it is not how old the woman is but rather how close she is to  menopause that determines her risk of having an aneuploid conceptus (Brook et al, 1984). In particular, the total number of follicles and developing follicles decreases with increasing maternal age so that when an oocyte is recruited for ovulation there are fewer choices and a  36 less mature or suboptimal oocyte may be the one chosen for fertilization (Gougeon et al, 1994; Reuss et al, 1996; Scheffer et al., 1999). As menopause is estimated to occur when the total number of follicles reaches -1000, any decrease, either artificially or naturally, will decrease the number of follicles and lead to relatively earlier menopause. Studies in mice showed that after a unilateral ovariectomy, there is an earlier rise in aneuploidy rates with age as compared to non-ovarectamized mice (Brook et al, 1984). In studies of women with a Down syndrome child, such women were found to be more likely than controls to have one ovary as a result of either ovarian surgery or congenital absence of one ovary (Freeman et al., 2000). These data support the "limited pool hypothesis" (Warburton et al, 1989; Kline and Levin, 1992) that states that it is the low number and quality of the few remaining oocytes that results in the increase in aneuploidy seen with increasing maternal age. Further support for the "limited pool hypothesis" comes from the finding by Kline et al. (2000) that menopause occurred one year sooner in mothers of a trisomy. They recruited women into three groups; women whose index spontaneous abortion (SA) was trisomic, women whose index spontaneous abortion was chromosomally normal and control women who had chromosomally normal live births. Earlier, Phillips et al. (1995) found no such evidence in their study of mothers of trisomy 21. This may be due to methodological limitations and differences in definitions of menopause. Specifically, Phillips et al. (1995) selected controls only for cases that had reached menopause, and their case subjects age of menopause was based on recall only. Kline et al. (2004) revisited this question and did not see a difference in antral follicle count or hormonal signs of aging among women who had experienced a trisomy as compared to controls. One goal of my thesis is to examine whether  37 premature aging as reflected in telomere length and other genetic measures can lead to aneuploidy. 1.4.3  Chromosome structure:  Chromosomes have two major structural features; the primary constriction of the centromere and the chromosome ends or the telomeres. Each has their own unique repetitive sequence and their own unique function in meiosis and mitosis. Normal centromere and telomere function are indispensable to proper chromosome segregation, but their role has been less well delineated than that of recombination because experimental approaches have been more limited (Gaulden, 1992). Not only are centromeres and telomeres important for chromosome and chromatid segregation but telomeres are also a measure of aging and short telomeres have been found to be associated with premature aging syndromes and infertility (Vaziri et al, 1993; Dorland et al, 1998). 1.4.3.1 Centromeres: The centromere is a specialized region of the chromosome that is formed by DNA alphoid sequences, other repetitive sequences and proteins. The most abundant DNA element in the human centromere is the a-satellite, which constitutes as much as 3-5% of chromosomal DNA (Lo et al, 1999). The human centromere is made up of 171 bp long repeating monomers, which are AT-rich and contain the DNA motif recognized by kinetochore proteins in vitro (Masumoto et al, 1993), and is where centromere proteins (CENPs) bind. Since centromere function is required for proper segregation of chromosomes at MI and for chromatid separation at Mil, variation in centromeric sequence or size might be expected to affect ND levels (Abruzzo et al, 1996). In particular, an abnormally small or  38 large alphoid array size within a homologous chromosome pair, or a large discrepancy between homologous alphoid array sizes, may result in improper pairing and alignment of the homologues, disrupting recombination, and leading to ND. There is considerable variation in alphoid DNA array length between homologous chromosomes. For example, the size of the X chromosome a-satellite array ranges from 1300 to 3700 kb (Mahtani and Willard, 1990). Therefore, a pair of parental chromosomes may differ in their a-satellite array size for the X chromosome by 2400 kb. It has also been shown that small array size variants are unusually common in chromosome 21, with a prevalence of 3.7% in adult euploid controls and 6.85% in DS patients (Lo et al, 1999). If the size of the centromeric heterochromatin does not exceed a minimum threshold value, it is possible that cohesion becomes compromised, either because the alphoid DNA may lack adequate mechanical strength to hold the chromatids together or because small alphoid arrays may bind fewer centromere-associated proteins. This in turn could lead to ND, because failure to maintain sister chromatid cohesion results in premature sister chromatid separation (at MI) (Maratou et al, 2000). Sumner (1991) has suggested that there are functional constraints on the size of the centromeric repeats. If they are too large, chromatids may fail to separate in time, and if they are too small, chromatids may fail to hold together. Maratou et al. (2000) also found that the risk for nondisjunction is related to the size of the alphoid array. In their study of the role of alphoid DNA size variation in trisomy 21 formation, they found that the alphoid array on one of the chromosome 21 homologues was small. The study looked at 23 families with a trisomy 21 child and initially compared combined alphoid array size to 38 controls. However, in a subset of families (n=12) it was possible to look at individual homologues providing  39 evidence that the risk for ND is related to the size of the alphoid array on one of the homologues being small. The importance of heterochromatic regions in mediating pairing and segregation was first suggested by studies of the effects of homologous duplications on the segregation of the achiasmate chromosome 4 in Drosophila oocytes (Hawley et al, 1993). The more chromosome 4 heterochromatin carried by this duplication, the higher the observed level of induced chromosome 4 ND. It is possible that ND may be increased in women with larger regions of heterochromatin at the centromere, similar to that seen in Drosophila and may be particularly true for achiasmate meiotic configurations. Although aberrant alphoid size may predispose a chromosome to ND, additional ND susceptibility factors are probably required, acting in combination with alphoid size for ND to occur. As with recombination, an altered or suboptimal size of a-satellite DNA may be another factor that when combined with a maternal age-related processing defect is unable to allow proper segregation of such a susceptible bivalent. Thus, the too small or too large alphoid size may be another "hit" in a "multiple hit" hypothesis of age-dependent ND of chromosomes in humans. Loss of control of replication of the centromeres may also result in problems at meiosis and mitosis, as has been suggested by Litmanovitch et al. (1998). This group found that the DNA at centromeres (a-satellite sequences) tends to replicate asynchronously, indicating a loss of replication control, in individuals with cancer and therefore presumed to be predisposed to somatic nondisjunction (N=10), when compared to controls (N=10). Using FISH with probes for the a-satellite regions of chromosomes 10, 11, 17 and X , chromosome pairs with asynchrony showed an increased rate of aneuploidy as well. Replication timing at  40 three loci including the centromere of chromosome 15, in mothers of trisomy compared to controls, was investigated in this project and a complete discussion can be found in Chapter 4. 1.4.3.2 Telomeres: Telomeres are specialized structures at the end of eukaryotic chromosomes that are important for chromosome stability (Rufer et al., 1998). They are composed of an array of repetitive DNA sequence (TTAGGG in the human) and associated proteins. Telomeres are believed to be involved in positioning the chromosomes during mitosis, the maintenance of chromosomal integrity, and the protection of unique DNA sequences. Decreased telomere size is associated with errors of chromosome segregation in mitosis. As cells near senescence, chromosomes with short telomeres show a tendency to form dicentric chromosomes, most likely as a mechanism for end protection (DeLange, 1995). Without this new structure being formed, the ends of the chromosomes, including essential genes, could be lost through endonuclease digestion. In yeast, loss of a single chromosomal telomere was found to be sufficient to induce cell cycle arrest (Sandell and Zakian, 1993). Disruption of telomere-telomere attachments during meiosis may also impair eventual synapsis and recombination (Walker and Hawley, 2000). Telomere length declines with age in most mitotic tissues, at a rate of - 50-200 bp per cell division, with the exception of the male germline where the expression of telomerase prevents any age-dependent telomere erosion (Aviv et al. 2003). The average telomere length in <10 year olds is 20kb but declines to 5kb in 60-70 year olds (Schwartz et al, 1993). This inverse relationship between telomere length and age in humans is a linear one, however, at each age there is a large amount of variability in telomere length among  41 individuals (Rufer et al, 1999). That is, different individuals of a given age may show marked genetic variation in telomere length (VonZglinicki et al, 1995). On the basis of the regular shortening and association with age, telomeres have been connected with replicative aging both in vitro and in vivo and have been called the "mitotic clock" (Saretzki and Von Zglinicki, 2002). In human diseases, such as Progeria and Down syndrome, that display premature aging, individuals have been found to have shorter than average telomeres (Vaziri etal, 1993). Short telomeres have been found to be associated with aneuploidy and chromosome loss in many cancers and somatic tissues (Butler et al, 1998). As well, it has been shown that the rate of aneuploidy in cultured lymphocytes increases with advancing age (Jacobs et al, 1963). It is possible that short telomeres may lead to the missegregation of chromosomes and ultimately to aneuploidy. There are a number of possible explanations for short telomeres. First, mutations in genes involved in telomerase activity or its structure may lead to shortened telomeres. Overexpression of the telomere-binding protein TRF1 has also been shown to cause telomere shortening (vanSteensel et al, 1997). Second, clones have sparked controversy about the reprogramming of telomere lengths in cloned animals after Dolly the sheep was reported to have shortened telomeres (Shiels et al, 1999). Finally, Dorland et al (1998) found significantly longer telomeres in women undergoing IYF with poor oocyte production when compared to fertile women of the same age group. They suggested that fewer cell divisions may have occurred due to a decrease in growth hormones. An increase in growth hormones may therefore lead to a higher rate of cell divisions and thus shorter telomeres.  42 1.4.4  Replication timing:  Normally both alleles of a gene replicate at the same time, or synchronously, in the cell cycle. Asynchronous replication is however seen for the genes on the X chromosome, with those on the inactive X replicating later than their counterparts on the active X chromosome, as well as imprinted genes, olfactory receptor genes and other monoallelically expressed genes (Ensminger and Chess, 2004). A higher rate of allele asynchrony and aneuploidy was found in a study comparing the cells of older women and mothers of a Down syndrome offspring, to young healthy female controls and fathers of DS offspring (Amiel et al, 2000). Specifically, they found that an asynchronous pattern was significantly more frequent at RB-1 and 21q22 in middle-aged women and in mothers of a Down syndrome offspring compared to the control group. They also used a-satellite sequences from chromosome 8 and 18 to demonstrate an increase in those trisomies in somatic in mothers of Down syndrome offspring and in middle-aged women compared to controls. 1.4.5  Gene mutations:  MTHFR and MTRR: In addition to structural aspects of chromosomes that can affect their segregation, there is some limited evidence that gene variants may play a role. Based on evidence that abnormal folate and methyl metabolism can lead to DNA hypomethylation and abnormal chromosomal segregation, mothers of a DS child were proposed to have an increased frequency of certain polymorphisms in methylenetetrahydrofolate reductase (MTHFR, C677T) and methionine synthase reductase (MTRR, A66G) (James et al, 1999). Both of these common point mutations result in reduced enzyme activity in heterozygotes and homozygotes and are known risk factors for neural tube defects (van der Put et al, 1998).  43 Women with the MTHFR, C677T polymorphism were found to have a 2.6 to 2.9 fold increased risk of having a child with DS (James et al, 1999; Hobbs et al, 2000). The same group analyzed maternal polymorphisms at the other genes in the folate pathway and they found an increase in mutant homozygotes among DS mothers at MTRR, and furthermore, the combined presence of both mutations seemed to increase risk further. However, a subsequent study failed to show the same association for trisomy 21 (Peterson et al, 2000) and a study of over 200 trisomies other then trisomy 21 could show no obvious increase in MTHFR or MTRR polymorphisms (Hassold et al, 2001). Apos4 alleles: Young mothers of DS children have been shown to have a fivefold greater risk than normal of developing Alzheimer disease (AD) later in life (Potter and Geller, 1996). This may suggest that these women either were mosaic for trisomy 21 or had an ongoing predisposition to chromosome missegregation that is reflected in their trisomy 21 offspring and their own increased risk for AD (Geller and Potter, 1999). Some studies provide evidence for a shared genetic susceptibility for DS and AD. Almost all individuals with DS above the age of 40 years have the neuropathological changes characteristic of AD (Rumble et al, 1989). Some forms of AD are caused by mutations in the APP (B-amyloid precursor protein) gene on chromosome 21, the presenilin-1 (PS-1) gene on chromosome 14 and the presenilin-2 (PS-2) gene on chromosome 1 (Petersen et al, 2000). The presenilin proteins may be important in chromosome segregation, as they have been shown to localize to kinetochores and centrosomes (Li et al, 1997). However, apoE mutations may explain this relationship. Apolipoprotein E (apoE) is a gene located on chromosome 19. The rare allele E4 has been identified as a risk factor for early-onset and late-onset AD in both familial and  44 sporadic cases (Strittmatter et al, 1993; Chartier-Harlin et al, 1994). Avramopoulos et al. (1996) demonstrated a higher frequency of the APOE s4 allele in young mothers of DS children due to maternal II errors. Petersen et al. (2000) demonstrated an increased frequency of allele 1 of the intron 8 polymorphism in PS-1 in mothers who had a trisomic offspring resulting from a meiosis II error as well. Mitochondrial mutations: Another factor proposed to contribute to problems with chromosome segregation is abnormalities of the mitochondria. Mitochondrial disfunction resulting from decreased ATPase6 and Tfam expression during meiotic maturation of oocytes has been suggested to cause nondisjunctional errors (Lee et al, 2003). Mitochondrial DNA (mtDNA) mutations have also been found to increase exponentially with increasing age. The curve showing the amount of mtDNA mutations with time or age looks very similar to the trisomy risk with age curve. The accumulation of mtDNA errors with time in both the somatic tissue supporting the oocyte in the ovary, as well as within the oocyte itself can lead to energy deficits and other problems that can make proper chromosome disjunction difficult (Schon et al, 2000). 1.4.6 Environmental: There has always been an interest in the possibility that our environment, what we eat and what we are exposed to, may influence our reproductive success or failure. Smoking, alcohol, caffeine and many other such factors have been shown to affect the health and well being of the developing fetus. But can these same exposures actually lead to nondisjunction at meiosis and increase risk for trisomy? The results are conflicting with some studies claiming decreased risk for a DS birth for smoking mothers (Kline et al, 1993), smoking not associated at all with DS (Torfs et al, 2000) and maternal smoking increasing the odds of a  45 DS child in younger mothers and a certain subset of meiotically-derived cases (Yang et al., 1999). High alcohol and caffeinated coffee consumption has been associated with a reduced risk of having a DS conceptus (Torfs et al, 2000). Yang et al. (1999) found that oral contraceptive use in combination with maternal smoking increased the risk for a DS conceptus while oral contraceptive use alone was not a significant risk factor. There has been conflicting evidence about the effects of radiation on reproduction and pregnancy outcome. Sperling et al. (1994) assessed whether an increased prevalence of trisomy 21 in West Berlin in January 1987 might be due to the Chernobyl reactor accident and they believed there to be a link. However, other groups have not found any genetic connection at all to exposure to radiation (Neel, 1998). In a study by Czeizel et al. (1993), exposure to an insecticide called trichlorfon was suspected as the cause for an increase in DS and other congenital abnormalities as well as an increase in thefrequencyof twins in Hungary in 1989 and 1990. In this case-control study, there were 15 live births in a small Hungarian village and 11 were affected by congenital abnormalities, 4 had DS and 6 were twins. All likely causes of such clusters were excluded and consumption of contaminated fish from fish farms was confirmed in most of the pregnant women, but all of the mothers of DS. When the chemical treatment of farmed fish was banned in 1991, the cluster ceased. This was the first published link between an environmental factor and congenital anomalies, particularly DS. It seems likely that our environment and in particular the environment in which the oocytes are developing is important for risk for trisomy. The effects may not be obvious and may be a "hit" in the proposed multiple-hit hypothesis for trisomy risk. Bisphenol A (BPA) is an estrogenic compound widely used in the production of polycarbonate plastics and epoxy resins. It can be found in baby bottles, water bottles, lining  46  tin cans and can also be found in dental sealants. In 2003, Hunt et al. reported an inadvertent environmental exposure in their mouse colony to BP A. This was followed by a highly significant increase in meiotic chromosome abnormalities including ND. By then intentionally damaging the caging materials and water bottles they were able to repeat those same results, leading them to conclude that BPA is a potent meiotic aneugen. The meiotic effects are dose dependent and are induced at environmentally relevant doses of BPA. This study provided the first direct link between mammalian meiotic aneuploidy and an accidental environmental exposure. Studies like this one lend support to the fact that environmental exposures can lead to changes in the environment of the body and the oocyte and lead to nondisjunction. The following project was undertaken in order to try to determine genetic mechanisms associated with nondisjunction in humans. I looked at features of chromosomes and meiosis associated with aging such as, telomere length, replication timing at genes and centromeres, somatic aneuploidy rates and recombination, specifically in maternal nondisjunction. I hypothesized that I would find an association between increased signs of premature aging in women experiencing chromosome nondisjunction. A family with recurrent aneuploidy (trisomy 21) was used as a case study of a possible predisposition to chromosome nondisjunction in meiosis. Finally, I investigated the association between a type of assisted reproductive technology called intracytoplasmic sperm injection (ICS!) and telomere length, as a measure of epigenetic integrity and newborn health.  47 Chapter 2: Materials and Methods 2.1  SUBJECTS AND SAMPLE COLLECTION  This study involved several distinct populations as listed and discussed separately below. When necessary, all participants signed a consent form approved by the Clinical Research Ethics Boards of the University of British Columbia and the BC Children's and Women's Hospital. 2.1.1  Trisomic pregnancy mothers and controls:  Samples  Cases: A total of 38 women (24 to 44 years of age; average age 37.1 years) with at least one trisomic pregnancy were identified by referrals from physicians at the Recurrent Pregnancy Loss Clinic (N=27) and the Medical Genetics Clinic at the Women's Health Center at BC Children's and Women's Hospital (N=8), and from other centers in BC (N=3). The majority of the cases were women who had experienced more than one trisomic pregnancy (N=22), while the remainder was made up of women who had either a single trisomy as well as at least one other pregnancy loss (N=14) and women who had only a single trisomy with no other known losses (N=2). Obvious causes of trisomy, such as a translocation in the mother, were ruled out by karyotyping. All karyotyping of samples was done in the Cytogenetics Laboratory at BC Children's and Women's Hospital. Controls: A total of 46 women aged 22 to 58 years of age (average age 38 years) with no history of pregnancy loss were ascertained through posted advertisements. Some of the women in this group have never had any children (N=7), some had an unknown number of children (N=16), and the remainder had at least one successful pregnancy (N=23).  48 All participants were volunteers and signed consent forms approved by the Ethics Boards at the University of British Columbia (UBC) and BC Children's and Women's Hospital, ethics approval number CO 1-0460. Details on additional samples used for studies in chapters 6 and 7 are given in those chapters. 2.2  TISSUE CULTURE AND HARVEST Heparinized blood was obtained from a subset of cases and controls as part of our  studies of trisomy risk (Chapters 3, 4 and 5). For each heparinized blood sample collected, five cultures were set up by adding 1 ml of whole blood to 10 ml of culture media in a 50 ml flask (details). Each culture contained 8.3 ml of RPMI culture media (Gibco - Invitrogen) 1.5 ml (15%) fetal bovine serum (StemCell Technologies), 1 ml each of L-glutamine and penicillin/streptomycin (1%) (StemCell Technologies), 100 pi of heparin (10,000 i.u./ml) (0.1%) and 0.2 ml of phytohemagglutinin (M form) (Gibco - Invitrogen) (2%). The culture flasks were placed in a humidified incubator at 37°C with 5% circulating C O 2 for three days. One to two hours before harvesting, 0.2 ml of KaryoMAX colcemid solution (Gibco - Invitrogen) was added to each flask. Cultures were then transferred to Falcon tubes and centrifuged at 1500 RPM in a Jouan CR412 bench-top centrifuge at room temperature (RT) for 5 minutes. After removal of the supernatant, prewarmed (37°C) hypotonic (0.075M KCL - Gibco - Invitrogen) solution was added to the pellet while vortexing. The tubes were then incubated in a 37°C water bath for 30 minutes. The cells were centrifuged again at 1500 RPM for 5 minutes, and the supernatant was removed. Fix (3:1 methanol and acetic acid) was added dropwise while vortexing until the pellet turned brown. Fix was then added up to 8 ml and the tubes were left in a fumehood for 15 minutes. The previous steps (spin, remove supernatant, add fix)  49 were repeated until the pellet was clean and white. Fix was added up to 8ml and tubes were stored at -20°C until needed. 2.3  MOLECULAR METHODS  2.3.1  DNA extraction: DNA was extracted from blood samples by a salt extraction technique (Laitinen et al.  1994). Seven milliliters of whole peripheral blood was transferred to a 50 ml conical Falcon tube. Cold erythrocyte (EC) lysis buffer (155 mM NH C1; 10 mM K H C 0 ; 0.1 mM 4  Na2EDTA;  3  pH 7.4) was added to each tube for a total volume of 45 ml. The tubes were then  incubated at 4°C for 2-3 hours or until the blood mixture was no longer cloudy. The tubes were centrifuged at 2000 RPM at 4°C for 15 minutes. After removing the supernatant, the pellet was washed by resuspending in 15 ml of cold EC lysis buffer and centrifuged at 3000 RPM. This was repeated until the pellet was clean. The cell pellet was suspended in 5 ml of SE buffer (75 mM NaCl; 25 mM Na EDTA; pH 8.0), mixed with 50 pi of proteinase K (20 2  mg/ml) and 500 pi of 10% sodium dodecylsulfate (SDS), and incubated at 37°C overnight. An additional 5 ml of SE buffer was added followed by a 5-minute incubation at 55°C. Three milliliters of 6M NaCl was added and vortexed for 30 seconds. This mixture was then centrifuged at 3000 RPM at RT for 15 minutes and the clear supernatant was poured into a new 50 ml conical tube. Two volumes of ice-cold 95% ethanol were added to precipitate the DNA. The DNA was spooled out with a sealed glass pipette, briefly washed in 70% ethanol, and dissolved in up to 500 pi of TE buffer (lOmM Tris/HCl pH 8.0; ImM EDTA) for storage.  50 2.3.2  Microsatellite marker typing:  Most primers were ordered from Research Genetics, but many were also synthesized at the NAPS unit at UBC. Each PCR tube contained 0.15 pi each of both the forward and reverse primers, 7.2 pi of distilled H 0,1.0 pi of 10X buffer with MgCl , 1.0 ul of dNTPs, 2  2  0.02 pi of taq and 0.5ul of DNA for a total reaction of 10.0 ul. Polymerase chain reaction (PCR) amplification was performed on an MJ Research thermocycler with 35 cycles of 30 seconds at 94°C for denaturation, 55°C for annealing, and 72°C for elongation. The PCR products were then visualized by one of two methods: (1) silver-stained polyacrylamide gels or (2) automated fluorescence analysis. For silver staining, approximately 5 pi of the product was mixed with an equal volume of urea loading buffer, denatured for 5 minutes at 94°C and loaded on a 6% polyacrylamide gel. For automated fluorescent analysis, PCR was performed with a forward primer labeled with the ABI Prism Dyes HEX or 6-FAM. The amplification products were sized using capillary electrophoresis on an ABI Prism 310 genetic analyzer. Fluorescence was detected by ABI Prism data-collection software and analyzed by use of GeneScan software. 2.3.2.1 Origin of Trisomy: Origin of trisomy was determined as described previously (Robinson et al. 1995 and 1997). Briefly, the observation of a marker amplifying three distinct allelesfromtrisomic tissue provides clear confirmation of a meiotic origin of the extra chromosome. Also, meiosis I and meiosis II errors can be distinguished from each other using centromeric markers, and although the paternal blood sample was not available, parental origin can be assigned with reasonable confidence if the maternal DNA is available and multiple markers are typed along the involved chromosome pair (Robinson et al. 1999).  51 2.3.2.2 Recombination mapping - genetic mapping: See Methods Chapter #7 2.3.3  Telomere length assay: Telomere length was determined by Southern blot analysis of terminal restriction  fragments (TRFs). The TeloTAGGG Telomere Length Assay kit from Roche Diagnostics (Montreal) was used to quantify TRF length. The kit uses a digoxigenin (DIG)-labeled telomere probe that is visualized by hybridizing the labeled blot with an anti-DIG-antibody coupled with alkaline phosphate and metabolizing the alkaline phosphate with CDP-Star, a chemiluminescence substrate. Chemiluminescence was measured using a Bio-Rad Fluor-S Multilmager and analyzed with Quantity One software. Digestion: The DNA from the trisomic pregnancy mothers and controls were used in this study to compare telomere length between the two groups. For each sample, 0.5 pg of DNA was digested with 10 U of Hinfi and 10 U of Rsal in IX digestion buffer in a total volume of 10 pi. The kit included two control DNA samples of known telomere length that were also digested as above. The digestion mixture was incubated overnight at 37°C, and a minigel was used to check for complete digestion. Gel electrophoresis loading buffer was then added to each sample. Electrophoresis and Southern blot transfer: A 0.8% agarose gel in IX TAE buffer (0.04 M Tris-acetate, 0.001 M EDTA, pH 8.0) was used to separate the digested DNA samples. A DIG-labeled molecular weight marker was run at either end of the lanes on the gel for analysis. The gel was run at 5 V/cm for 3.5 hours, or until the loading dye had moved ~ 10cm. The gel was then submerged in 0.25 M  52 HC1 for 10 minutes, rinsed in distilled water two times, submerged in denaturation solution (0.5 M NaOH, 1.5 M NaCl) for 15 minutes twice, rinsed in distilled water twice and submerged in neutralization solution (0.5 M Tris-HCl, 3 M NaCl, pH 7.5) twice. The DNA in the agarose gel was then transferred to a nylon membrane by capillary transfer in 20X SSC buffer (3 M NaCl, 0.3 M sodium citrate, pH 7.0) overnight. The nylon membrane was then washed in 2X SSC buffer and the DNA was fixed to the membrane by UV-crosslinking, and stored at 4°C. Hybridization and washing:  The nylon membrane was prehybridized in 24 ml of prewarmed DIG Easy Hyb solution at 42°C for 45 minutes. The membrane was then hybridized with 1.8 ul of DIGlabeled telomere probe in 9 ml of prewarmed DIG Easy Hyb solution at 42°C for 3 hours. 150 ml of a stringent wash buffer I (2X SSC, 0.1% SDS) was used to wash the membrane after hybridization for 5 minutes at room temperature (RT), and then 150 ml of a prewarmed stringent wash buffer II (0.2 M SSC, 0.1% SDS) was used to wash the membrane twice for 15 minutes at 50°C. The final wash was for 5 minutes at RT with 150 ml of IX washing buffer. This was followed by an incubation in 100 ml of IX blocking solution for 30 minutes at RT and finally the blot was incubated in 100ml of IX blocking solution with 6 U of AntiDIG-AP (anti-DIG antibody coupled with alkaline phosphate) for 30 minutes at RT. After washing with IX washing buffer three times, the membrane was incubated with 100 ml of IX detection buffer, containing the alkaline metabolizing agent CD?-Star for 5 minutes. The membrane was placed in a hybridization bag with 3 ml of substrate solution on the DNA side of the membrane for 5 minutes, after which time the excess liquid was squeezed out of the bag and the bag was sealed.  53 Detection and analysis: The telomere length was determined by measuring chemiluminescence using a BioRad Fluor-S Multilmager. A digital image of the membrane was taken after ~ 3 minutes of exposure to light. Quantity One software was then used to analyze the smears on the membrane by determining the most intense region of the smear and relating that to the size standards run on either end of the gel. A plot of the chemiluminescence intensity is given with a peak at the most intense point. A band is then assigned manually to the most intense point and a kb value is assigned to the band. This represents the average length of the telomeres in that sample. 2.4  CYTOGENETIC METHODS  2.4.1  Quantitative FISH (Q-FISH):  Metaphase preparation: Metaphase slides were prepared for many of the mothers of trisomy and the controls for the telomere length study. Fixed and stored cell pellets were resuspended infreshfix (3:1 methanol: acetic acid) to the appropriate consistency and concentration (wax paper appearance) and 1 drop of the solution was dropped onto a precleaned glass slide. The slides were thenrinsedtwice with a Pasteur pipette full of fix and air dried. Slides were "aged" overnight before hybridization mixture was added. Rehydration, fixation and dehydration: All of the following steps were performed on a rocking platform at room temperature. Slides (cells) were rehydrated in IX phosphate buffer solution (PBS) for 15 minutes and then fixed in 4% formaldehyde in PBS (21.6 ml of 37 % formaldehyde in 178.4 ml PBS) for 2 minutes. Slides were then washed 3 times in IX PBS for 5 minutes. Slides were treated with  54 pepsin (1 mg/ml) prepared infreshacidified water (pH 4.0) and warmed to 37°C for 10 minutes. Slides were then washed 2 times in IX PBS for 2 minutes,fixeda second time with 4% formaldehyde in PBS for 2 minutes and washed again 3 times in IX PBS for 5 minutes. Finally, slides were dehydrated in pure ethanol in three steps of 5 minutes each at 70%, 90% and 100%, and then air-dried. Hybridization: The hybridization mixture contained peptide nucleic acid (PNA) probes for the telomeres and centromeres of all chromosomes. The mixture contained a Cy3 - (CCCTAA)3 telomere probe and a FITC-labeled PNA probe specific for a consensus sequence of alphasatellite DNA (Taneja et al, 2001). Each slide had two spots of 10 pi each consisting of 0.5 ul 0.2M TRIS, 0.32 pi MgCl , 7 pi deionized formamide (dF), 0.5 pi of PNA telomere probe 2  labeled with Cy3™ (yellow), 0.34 pi PNA centromere probe labeled with FITC (fluorescein isothiocyanate - green - 530 nm), 1 pi NEN and 0.34 ddH 0. Slides were then placed on an 2  80°C hot plate for 3 minutes in order to denature both metaphase chromosomes and probes. Finally, slides were placed in a slide box in a beaker lined with moist paper towels and sealed with parafilm for 1 to 2 hours. Washing and dehydration: After hybridization, slides were washed to remove excess probe. Slides were first washed for 2 x 15 minutes in 70% formamide in 10 mM TRIS + 0.1% BSA (pH=7.0-7.5). This was followed by a second wash in IX TBS with 0.8% tween 3x5 minutes. Slides were then dehydrated in three steps, each 5 minutes long; 70%, 90% and 100% ethanol. Slides were air-dried and then an antifade (Vectashield) containing DAPI counterstain was added for visualization.  55  Microscopy: Digital images of interphase and metaphase cells after FISH with Cy3 - and FITC labeled P N A probes were acquired with a Microlmager MII400-12 camera (Xillix; Vancouver) attached to an Axioplan 2 microscope (Zeiss, Jena, Germany) equipped at the excitation site with a multiple filter wheel (Pacific Scientific, Rockford, IL), a DAPI/Cy-3 dual band emitter filter at the emission site (Chroma Technology, Brattleboro, VT), and a focusing motor drive (ZSS 43-200-1.2; Phyptron, Grodenzell, Germany). Microscope control and image acquisition was performed with dedicated software (SSM; Xillix). Images were acquired with a Plan-Apochromat X63, N A 1.4 objective lens (Zeiss) and a mercury/xenon lamp (200 W; Optiquip, Highland Mills, N Y ) . Images of each fluorescent dye were acquired and stored for further analysis. A n algorithm developed by Steven Poon (Poon et al, 1999) runs in his software program TFL-TELO. The program requires two spectral fluorescence images to generate a length estimate for each telomere in a metaphase chromosome spread; one image shows the chromosomes which are stained with the D N A dye, DAPI, and the other image shows the associated telomeres hybridized to the Cy3 fluorescent probe. The algorithm performs telomere segmentation, telomere fluorescence measurement and chromosome segmentation for each metaphase spread (Figure 2-1). It is assumed that the length of the telomere is directly related to its integrated fluorescence intensity (IFI) value as the fluorescence probes used are assumed to hybridize quantitatively to telomere repeat sequences. The IFI values given by TFL-TELO are in a table format that can be copied and transferred to Excel for analysis.  56  Figure 2-1: Computer generated image of a metaphase preparation after hybridization with PNA probes for the telomeres and counterstained with DAPI. An algorithm segments the telomeres and allows the TFL-TELO program to determine the amount of fluorescence at each telomere segment. The numbers assigned to the chromosomes are arbitrary. Two levels of calibration were used to ensure a reliable quantitative estimation of telomere length in various samples. To correct for daily lamp variations, images of fluorescent beads (orange beads, size 0.2 pm; Molecular probes) were acquired and similarly analyzed with the Image Analysis computer program. Then to relate fluorescence intensity to number of (TTAGGG) repeats, we hybridized and analyzed plasmids with a defined n  (TTAGGG)n length (Hanish et al 1994). From these experiments the slope and the intercept of the linear relationship between the calculated IFI values and the telomere length estimates  57 were established (Zijlmans et al. 1997). As outlined by Zijlmans et al. (1997), there is a restriction of this calibration method in that the actual telomeres are outside of the range of the (TTAGGG) length of the plasmids. The assumption is made that the linear correlation n  obtained betweenfluorescenceintensity and telomere insert size in plasmid DNA is maintained in the higher range.  2.4.2 Interphase FISH for Replication Timing: Slides were prepared for many of the mothers of trisomy and the controls for replication timing studies. The same method was used to prepare these slides as was used for metaphase slide preparation. Denaturation, dehydration and hybridization: Slides were denatured with prewarmed 70% formamide in 2X SSC in a coplin jar at 73°C for 5 minutes. Slides were dehydrated for 1 minute each in 70%, 85% and 100% ethanol and air-dried before adding the hybridization mixture. The hybridization mixture consisted of 7 pi of hybridization mixture, 2 pi of dd^O and 1 pi of probe. Probes for 21q22 (LSI - red; Vysis) and chromosome 15 centromere (red; Vysis) were used to score replication timing for each sample. The hybridization mixture was denatured in 73°C for 5 minutes and then two 10 pi spots were added to each slide and covered with a cover slip. The cover slip was sealed with rubber cement and the slides were placed in a humidified slide box and placed in a 37°C waterbath overnight. Washing: The rapid wash procedure in the Vysis protocol was used. A coplin jar of 0.4X SSC with 0.3% NP-40 was prewarmed and used to wash the slides for 2 minutes after 1-3 seconds  58 of agitation. A second wash solution of 2X SSC with 0.1 % NP-40 was used to wash the slides for 5 seconds to 1 minute at RT after 1-3 seconds of agitation. Visualizing and Scoring: The slides were air dried in darkness and then 23 pi of a DAPI counterstain in an antifade (DAPI II; Vysis) was added to each slide and covered with a cover slip. For each sample at least 200 interphase nuclei were scored for the number of signals found for each probe. Specifically a dot hybridization assay was used to determine replication timing. Two single dots (singlets=S) or two double dots (doublets=D) either meant that the nucleus had not replicated yet or had replicated synchronously (both loci had replicated at the same time in the cell cycle). A singlet and a doublet (SD) in the same nucleus indicated that the locus had replicated asynchronously (Figure 2-2). Only non-overlapping nuclei with clear signals of the same size were scored. The proportion of nuclei that had replicated asynchronously was calculated by taking the # SD nuclei / total # nuclei x 100. As it could not be determined whether or not the nuclei containing singlets had even replicated at all, in some cases an X chromosome probe for HPRT labeled with FITC was used simultaneously to determine which nuclei had truly replicated. HPRT'has been extensively studied and SD nuclei are seen -40% of the time. Only those nuclei that contained an SD signal for HPRT were then scored for the other probe used, either the chromosome 15 centromere probe or the 21q22 probe.  59  Interphase nuclei  S - single dot  D - double dot  Unreplicated  Synchronous  Replicated  Synchronous  SD - single dot/ double dot Asynchronous  Figure 2-2: Diagram of the dot hybridization assay used to assess replication timing in interphase nuclei. Two single dots or two double dots for a single locus indicate a synchronous pattern of replication, while a single dot and a double dot found within the same nucleus indicates replication asynchrony.  2.5  STATISTICAL ANALYSIS  T-test: The t-test tests the significance of the difference between the means of two samples. T-test for independent samples: The t-test for independent samples is used when the two samples are independent of each other in that they are separate samples containing different sets of individual subjects. The individual measures in group A are in no way linked with or related to any of the individual measures in group B, and vice versa. The two samples must be randomly drawn from normally distributed populations and the measures of which the two samples are composed are equal-interval.  60 T-test for correlated samples: The t-test for correlated samples is used when the two samples are correlated in some way. The two sets of measures are arranged in pairs and are thus potentially correlated. This procedure may also be called the repeated-measures or within-subjects t-test, because it typically involves situations in which each subject is measured twice, once in condition A and then again in condition B. The only requirement of the correlated-samples design is that each individual item in sample A is intrinsically linked with a corresponding item in sample B. Correlation:  The primary measure of linear correlation is the Pearson product-moment correlation coefficient, symbolized by the lower-case Roman letter r, which ranges in value from r=+l .0 for a perfect positive correlation to r=-l .0 for a perfect negative correlation. The midpoint of its range, r=0.0, corresponds to a complete absence of correlation. The line that is calculated that best fits the bivariate distribution of data points is called the regression line, or line of regression. Significance of the Difference Between Two Correlation Coefficients: Using the Fisher r-to-z transformation, a value of z is calculated that can be applied to assess the significance of the difference between two correlation coefficients, r and r , found a  b  in two independent samples. Chi-Square Goodness of Fit/ Fisher's Exact Probability Test:  The Chi-Square "Goodness of Fit" test is used to test if a sample of data came from a population with a specific distribution. The purpose of it is to determine whether the observed frequencies (counts) markedly differ from thefrequenciesthat we would expect by chance. This test can only be used if all values are equal to or greater than 5. The chi-square  61 statistic becomes inaccurate when used to analyze contingency tables that contain exactly two rows and two columns, and that contain less than 50 cases. Fisher's exact probability is not plagued by inaccuracies due to small N's. If the sample sizes are small and the numbers are below 5, then an alternative and preferable test is the Fisher Exact Probability test.  62  Chapter 3: Telomere length in women who have experienced a trisomic pregnancy •  The patients used i n this study were collected by referrals from M a r y Stephenson at the Recurrent Pregnancy Loss C l i n i c (N=27) and various clinicians at the Medical Genetics C l i n i c at the Women's Health Center at B C Children's and Women's Hospital (N=8), and from other centers i n B C (N=3).  •  A l l Southern experiments were carried out i n the lab o f Wendy Robinson at the B C Research Institute for Women's and Children's Health. M a r i a Penaherrera, Chiho Hatakeyama, Ruby Jiang and myself were all involved i n the D N A extraction, digestion and Southern transfer o f all samples, as well as the data collection. I alone analyzed the data.  •  The origin o f trisomy studies were done by Ruby Jiang and Chiho Hatakeyama.  •  A l l Q-FISH experiments were carried out i n the lab o f Peter Lansdorp at the Terry Fox Laboratories at the B C Cancer Agency, and I did all o f the tissue culturing, harvesting, metaphase preparation, F I S H experiments, data collection and analysis.  AIM:  The aim o f this portion o f my thesis is to determine i f telomere length is decreased i n women experiencing a trisomy. This w i l l test the hypothesis that women experiencing a trisomy are prematurely aged and w i l l thus have shorter telomeres on average than women o f the same age who have not experienced a trisomy.  63  3.1  INTRODUCTION: Telomeres are specialized structures at the end o f eukaryotic chromosomes that are  important for chromosome stability (Rufer et al., 1998). They are composed of an array o f repetitive D N A sequence ( T T A G G G in the human) and associated proteins. In healthy human cells, telomere length has been found to range from a mean of 5 kb (60-70 year olds) to a mean o f 20 kb (<10 years) (Schwartz et al., 1993). Telomeres are believed to be involved in positioning the chromosomes during mitosis, maintenance o f chromosomal integrity, and the protection of unique D N A sequences. Specifically, decreased telomere size is associated with errors of chromosome segregation in mitosis. A s cells near senescence, chromosomes with short telomeres show a tendency to form dicentric chromosomes, most likely as a mechanism for end protection (DeLange, 1995). In yeast, loss o f a single chromosomal telomere was sufficient to induce cell cycle arrest (Sandell and Zakian, 1993). Disruption o f telomere-telomere attachments during meiosis may also impair eventual synapsis and recombination (Walker and Hawley, 2000). Telomere length declines with age in all mitotic tissues, at a rate o f ~ 50-200 bp per cell division, with the exception o f preimplantation embryos and the male germline where the expression o f telomerase prevents any age-dependent telomere erosion ( A v i v et al., 2003). This inverse relationship between telomere length and age i n humans is striking. However, individuals o f a given age may show marked genetic variation i n mean telomere length (Rufer et al., 1999). Furthermore, different telomeres within a single cell vary in size. In human diseases, such as Progeria and D o w n syndrome, that display premature aging, individuals have been found to have shorter than average telomeres (Vaziri et al., 1993). O n the basis o f the regular shortening and association with age, telomeres have been connected  64 with replicative aging both in vitro and in vivo and have been called the "mitotic clock" (Saretzki and Von Zglinicki, 2002). Short telomeres have been found to be associated with aneuploidy and chromosome loss in many cancers and somatic tissues (Butler et al., 1998), because telomeres play a key role in chromosome segregation in both mitosis and meiosis (Sandell and Zakian, 1993). As well, it has been shown that the rate of aneuploidy in cultured lymphocytes increases with advancing age (Jacobs et al., 1963). It is possible that short telomeres lead to the missegregation of chromosomes and to this increased somatic cell aneuploidy observed with age. Although meiosis is initiated in females in early development and these cells do not undergo mitosis, the telomeres contained within the oocytes may be short as a reflection of the length that the telomeres were when meiosis began in this individual. Specifically, an individual that was born with relatively short telomeres may also tend to have shorter telomeres than age matched controls later in life. Individuals with premature aging syndromes also display decreased fertility suggesting that there might be a relationship between fertility and accelerated telomere shortening (Dorland et al., 1998). In addition, oocyte maturation is dependent on the health of the entire follicle: the egg plus the somatic support cells that surround it. Accelerated aging of the follicular support cells, which do undergo limited mitotic expansion, could also affect oocyte quality. Telomere length is established early in development and we would therefore expect some correlation between telomere length in different tissues from the same individual (Saretzki and VonZglinicki, 2002; Cawthon et al, 2003). Telomere length measured in blood is correlated with length in other tissues (Saretski and Von Zglinicki, 2002; Butler et  65 al, 1998). Thus, I expect that the length that was measured in peripheral blood will reflect the length in ovarian cells. I hypothesize that women experiencing a trisomic pregnancy will have shorter than average telomeres for their age as compared to controls with no history of trisomy. I predict this effect to be more striking in women experiencing trisomy at a relatively young age. In this study I analyzed telomere length by Southern analysis of peripheral blood in two groups of women in order to test this hypothesis. One group of women had experienced a trisomic pregnancy while the other group had not. Methods and Materials: Samples  Cases: As described in chapter 2, a total of 38 women (24 to 44 years of age; average age 37.1 years) with at least one trisomic pregnancy were identified by referrals from physicians at the Recurrent Pregnancy Loss Clinic (N=27) and the Medical Genetics Clinic at the Women's Health Center at BC Children's and Women's Hospital (N=8), and from other centers in BC (N=3). The majority of the cases were women who had experienced more than one trisomic pregnancy (N=22), while the remainder were made up of women who had a single trisomy plus at least one other pregnancy loss (N=14) or women who had only a single trisomy with no other known losses (N=2). Obvious causes of trisomy, such as a translocation in the mother, were ruled out by karyotyping. Controls: A total of 46 women aged 22 to 58 years of age (average age 38 years) with no history of pregnancy loss were ascertained through posted advertisements. Some of the women in this group have never had any children (N=7), in some there was no pregnancy history other than no history of SAs (N=16), and the remainder had at least one successful pregnancy (N=23).  66 All participants were volunteers and signed consent forms approved by the Ethics Boards at the University of British Columbia (UBC) and BC Children's and Women's Hospital, ethics approval number C01-0460 (Appendix I). 3.2  RESULTS:  Molecular Studies: Out of the 38 women experiencing a trisomic pregnancy or loss, 27 had one or more losses or pregnancies for which products of conception could be obtained and analyzed for the origin of the trisomy. Of these 27 women, 26 were determined by molecular testing to have at least one trisomic pregnancy that had a maternal meiotic origin. Only case ND24 had a single trisomic pregnancy that had a paternal postzygotic origin and this woman was therefore excluded from all analyses. As maternal meiotic errors predominate among trisomies (Hassold et al., 1995), I assumed that the other 11 samples would have this origin and they were therefore included in the analyses. Telomere Length Assay - Southern analysis: The methods for the Southern analysis are given in Chapter 2, and all results are found in Table 3-1 and Table 3-2 for cases and controls respectively. Figure 3-1 shows a typical Southern blot image where each lane represents one sample after hybridization with a telomere-specific probe. To assess the reproducibility of the telomere length assay most samples were analyzed more than once on separate Southern blots. The repeat measurements were fairly well correlated (r=0.51) and a t-test for correlated samples determined that there was no significant difference between the two groups of measurements (p=1.0). When there was a difference of more than 2 kb between measurement 1 and measurement 2, the sample was analyzed a third time and all values were averaged to determine the final value to be  67 used for statistical analysis. Figure 3-2 summarizes the results from the telomere length assay after Southern blot for the patient samples and the control samples. The average mean TRF length for all of the mothers of trisomy is 8.6 kb (n=37, avg. age 37.1 years) compared to 8.4 kb (n=46, avg. age 38 years) for the controls (p = n.s.; t-test). There were a range of TRF lengths in the patient group from 6.0 kb to 15.6 kb. This range was similar in the control group with the shortest TRF length of 3.5 kb to the longest at 13.3 kb. • S  • mm  IP  21.2kb  ~  8.6 kb 6.1 kb  S  % y  -M  "'  *  l i t  4  "  T  2.7 kb 2  te  3!  e  1.1 kb  I S t $ p « S r TWO 7 A 3 & G I t  £ i  I  1 € 1 -2  3  (S  2 s  i  e  -E  1  i  i  |  I  c?  a  o  i  U  ^  fc.  I  •3  <  < i  I s  Figure 3-1: A Southern blot showing typical smears resultingfromhybridization of a telomere repeat (TTAGGG)-specific probe to digested genomic DNA. Each lane represents one sample. The darkest point in the smear indicates the average TRF (terminal restriction fragment) length in each sample and its size is estimated using the molecular weight ladders shown in lanes 1 and 20.  :*S*»f<.  68  Southern Blot Telomere length with age in Cases (N=37) and Controls (N=46) 18  R = -0.18 R = -0.16  16 14  £  12  c _  10  mei  _  fl)  6  o  4  u r-  •  • *—"  8  •  •—•  —^2  • .  #  *  ...  ;  «  -  . . .  -  . *  • •  2 0 20  25  30  35  40  45  50  55  60  Age (years)  Figure 3-2: TRF length by age after measurement by Southern blot in all cases (N=37 in red) and controls (N=46 in blue). The correlation coefficients are not significantly different (p=n.s.; Fisher r-to-z transformation). Telomere Length Assay - Q-FISH analysis: The methods for the Q-FISH analysis are given in Chapter 2, and all results are found in Table 3-1 and Table 3-2 for cases and controls respectively. Figure 3-3 shows a metaphase spread preparation after telomere probe hybridization. To assess the reproducibility of the telomere length assay 3 samples were analyzed more than once on separate FISH slides. The repeat measurements were well correlated (r=-0.96) and a t-test for correlated samples determined that the difference was not statistically significant (p=0.19). For the three repeated samples, the two values were averaged to determine the final value to be used for statistical analysis. Figure 3-4 summarizes the results from the telomere length assay by Q-FISH. The average telomere length for all of the mothers of trisomy is 4.6 kb (n=14, avg. age 37.1 years) compared to 4.4 kb (n=l 1, avg. age 32.5 years) for the controls (p = ns; t-test). There  69 was a range of telomere lengths in the patient group from 2.6 kb to 6.2 kb. This range was similar in the control group with the shortest telomere length of 3.5 kb to the longest at 6.3 kb. Figure 3-2 shows that when corrected for age, the telomere length in these two groups measured by Southern analysis follows an expected decrease with age and the trend lines are almost identical. The correlation coefficients ( r ) are not significantly different (z=0.06; n.s.). Figure 3-4 shows that when Q-FISH was used to measure telomere length, the control samples appear to have the expected decrease in telomere length with age as a group, while the patient group appears to have a slight increase in telomere length with age as a group.  Figure 3-3: A metaphase preparation after Q-FISH with PNA probes for the telomeres. A computer program TFL-TELO measures the amount of fluorescence emitted from each telomere and calculates a value that can be converted to a kilobase length. All of the telomere lengths are then averaged to arrive at an average value for telomere length for that sample.  70  Q-FISH telomere length with age in Cases (N=13) and Controls (N=11)  0  J  ,  ,  ,  20  25  30  35  40  Age (years)  Figure 3-4: Telomere length by age after measurement using Q-FISH in cases (N=13 in red) and controls (N=l 1 in blue). Table 3-1: List of patient samples and pregnancy history (N=37 for Southern and N=13 for Q-FISH). SA = spontaneous abortion, T = term birth, E = ectopic pregnancy, SB = stillbirth and TA = termination. Sample name  Trisomy Ascertained on  Pregnancy History  Telomere Length (kb) - Southern (repeat msmts)  ND1  Age at blood draw (Age@ first trisomy) 24 (22)  47,+18  9.8(9.2, 9.1, 11.2)  ND13  34 (33)  47,+16  7.3 (7, 7.6)  4.3  ND16  42 (40)  47,+16  8.8(11.2, 8.2, 7.1)  4.0  ND20  37 (36)  47,+21  7.5 (6.0, 8.9)  4.0  ND21  37 (33)  47,+13  40  47,+16  9.5(8.8, 8.6, 11.8, 9.5, 8.6) 9.2 (8.4, 10.1)  3.6(3.3,3.8)  ND23 ND25 ND26  41 (41) 33  47,+22 47,+16  ND27 ND28  30 33  ND29 ND30  31 (29) 35  47,+21 mT16 (mosaic trisomy) 47,+16 47,+21  TA,SA,SA,SA (47,XY + 18), SA(ch4 structural) SA,T,SA,SA (47,+3),SA (+16) T(XY),T(47,XY+21x2), TA(47,+ 18),TA(47,+ 16) T,TA(47,+21), SA (47,+15) SA,SA,SA (47,XY+13), SA (46.XX) T ( 4 6 , X Y ) , SA (47,+16), T (46,XY) T,SA,SA,SA (47,XX+22) SA,SA,SA (47,XX+16),SA (46,XX) 47,+21 (history ofT21) 47,+16 (mosaic)  Telomere Length (kb)Q-FISH (repeat msmts) 4.2  ND31  38 (36)  47,+4  47,+16 and previous SA T (46,XX),SA, T A (47.XX+21) SA,SA (47,XX+21),SA (47,XX+4)  7.3 (8.5, 6) 9.3(10.7, 7.8) 8.8 (9.2, 8.5) 10  5.7 2.6  4.6  6.2 (6.4,6.1) 9.3 8.3 (9, 7.6)  5.4  71 ND32 ND33 ND36  29 (28) 39 41 (36)  47,+21 47.+21 47,+21  ND37  39 (38)  47,+7  ND38 ND39  38 (37) • 38 (35)  47,+13 47,+21  Cl  35 (36)  47,+22  RSA-1  36 (34)  triploidy  RSA-10  41 (39)  47,+22  RSA-27  39 (38)  47,+15  RSA-37  40 (39)  47,+4  RSA-51  37 (38)  47,+20  RSA-110  34 (33)  47,+16  RSA-133  44 (39)  48,+16,+18  RSA-158  40 (38)  47,+9  RSA-189  42 (39)  47,+16  RSA-192  42 (38)  polyploidy  RSA-198  40 (39)  47,+16  RSA-218  41 (40)  47,+16  RSA-225  39 (37)  47,+12  RSA-251  36 (35)  47,+13  RSA-256  35 (32)  i21q  RSA-260  35 (37)  47,+22  RSA-263  39 (37)  47,+21  TA(47,XX+21), ongoing 47,+21 (2 SAs) T(47,+21),TA (47,+21),T(46) SA,T,SA (47,+21),SA(47,+7) SA,TA(47,+13) TA(47,+21),T(twin boys),TA(47,XY+21) T (46,XX),SB (46,XX),SA (47,+22) SA,SA,SA(47,XY,+7), SA(48,XX,+15,+16),SA( 69,XXY) T,SA(47,XX,+15),SA,SA (47,XY,+22),T SA,SA,SA(70,XXY,+2),S A(46,XY A N D 47,XY,+ 15) T,T,SA,SA,SA(47,XY,+1 3),SA(47,XX,+22), T,SA(48,+14,+21),SA(47, +4) T,SA,SA(47,XY,+16),SA (47,+20),T SA(47,XX,+10),SA(47,X X,+16),E T,SA,SA(47,XY,+14),SA (47,XY,+15), SA,SA(48,XY,+16,+18) T,SA,SA(47,XX,+ 16),SA ,SA(47,XX,+9) TA,T,SA(47,XY,+16),SA ,SA, SA(46,X,+21),SA(47,XX, +16) TA,SA(47,XY,+8),SA,SA (71,XXXY,+14) SA(48,XX,+12,T(12;12)+ 16),SA(47,XY,+16) T,SA,SA(46,XX),SA(47, +17), SA(47,+15),SA(47,+16) T,T,TA,SA(47,XX,+16),S A,SA(47,XX,+12) T,SA(46,XX),SA(47,XX, +4),SA(47,XY,+13) SA(47,XY,+16),T,SA(46, XX,i21q) SA,SA+E,SAX2(47,XY,+ 22;47,XX,+22) SA(+22),T(+18),T(+21)  7.6 (8,7.1) 7.1 (7.4,6.8) 7.5 (6.9, 8)  4.7 5.5  9.3 (9.1,9.4)  4.7  7.9 8.5 8.2 (8.1,7.8,6.9, 8.2,9.3, 8.7) 9.2 (6.4, 7.7, 13.5)  12.2(11.9, 12.5) 6.6  9.3 (9.1,9.4)  7.1 8.4 7.8  8.2 (8, 8.4) 6.0  10.1 9.3 8.6  8.1 (7, 9.1) 6.9 15.6 9.6 7.2  6.2  72 Table 3-2: List of control samples (n=39 for Southern and n=l 1 for Q-FISH). Sample name C2 C3 C4 C5 C6 C7 C8 C9 CIO Cll C12 C14 A3 S2 RH1  Age at blood draw 36 22 27 32 38 32 31 38 32 43 28 23 30 29 45  RH2  42  RH3  39  RH4  40  RH5  39  RH6  40  RH7  39  RH8  40  RH9  43  RH10  44  RH11  38  RH12  39  RH13  41  RH14  43  RH15  46  RH16  58  RH17  40  P Y 3 mom 90-11-002 86-01-040  39 39 40  Pregnancy History  Control with children No children No children Control with children Control with children No children No children Control with children No children Control with children No children No children No children Control with children Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Reproductively healthy - children after 37 years Control with children No losses or trisomies No losses or trisomies  Telomere Length (kb) Southern (repeat measurements) 9.2 (9.9, 8.4) 9.0 (9.9, 8.4) 7.9 (9, 6.7) 9.6(11.7,9.7,7.3) 6.3 (7.4,5.1) 10.8 (9.9, 11.7) 7.3 (6.9, 7.3, 7.8, 7.2) 8.7 (9, 8.3) 6.8 (7.3, 6.3) 8.4 (7, 9.5, 8.6) 10.3 7.2 (7.2,6.9, 6.5,7.2,8, 11.6) 9.0 (8.3, 9.6) 7.7 (7.1,8.3) 9.0 (8.7,9.3) 7.4 (7, 7.8) 11.0(13.2, 8.8) 8.6(8.5, 8.7) 9.0 (9.2, 9.7, 8.2) 10.6(13.1,9.7,9.1) 12.0(11.7,12.9, 11.4) 10.2 (9.4, 12,9.1) 7.1 7.4 7.2 6.8 6.9 6.3 7.7 3.5 6.2 (6.4, 5.9) 5.6  Telomere Length (kb) Q-FISH (repeat measurements) 3.7 4.0 (2.7, 5.2) 4.5 3.5 4.3 3.5 4.0 6.3 4.3  5.1 3.5 (3.1,3.9)  73 89-02-002 90-10-043 91-11-013 93-08-010 87-08-016 91-03-004 86-02-015 86-02-004 90-01-029 89-10-048 89-10-045 85-04-017 88-04-030 88-03-023  37 33 39 39 42 34 41 41 38 40 37 37 37 37  No No No No No No No No No No No No No No  losses losses losses losses losses losses losses losses losses losses losses losses losses losses  or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies or trisomies  7.6 6.8 (6.1,7.5) 11.8 8.8 10.4 8.1 8.1 7.3 8.1 10.9 11.3 8.5 13.3 6.7  Comparison of Southern analysis and Q-FISH: The telomere length measurements obtained by the two methods were quite different and did not correlate well with each other (r=-0.18) (Figure 3-5). In order to determine which method is more reliable and reproducible, multiple samples from a wide range of ages were analyzed and then the two methods were compared to each other. As mentioned above, Southern analysis seemed to be reproducible and repeated samples correlated well with each other, while the same was not seen for samples analyzed by Q-FISH. Figure 3-6 shows that the Southern TRF lengths are significantly longer (p<0.0001; t-test) than the telomere lengths acquired by using Q-FISH, but the correlation coefficients when comparing telomere length with age are not significantly different between the two methods (p=n.s.; Fisher r-to-z transformation).  74  Correlation between Southern and Q-FISH for all samples  R = -0.18  •  •  o  •  •  9  •  * •  •  •  *  •  •  •  •  2  4  3  5  6  Q-FISH  Figure 3-5: Correlation between the Southern method of analysis and Q-FISH for telomere length measurement in samples measured with both methods (N=25).  Telomere length by age for all samples by S o u t h e r n (N=95 in red) and Q-FISH (N=28 in blue) 18 - ~ 16  n  R = -0.06 R = 0.06  & 14 £ 12 O) c 10  4  m  i  0)  O £  • * •« - * »—*  8  6 4 2 0 10  20  30  40  50  60  70  80  Age (years)  Figure 3-6: Telomere length measurements for all samples measured by Southern analysis (N=95 in red) and Q-FISH (N=28 in blue). TRF lengths are significantly longer (p<0.0001; t-test) but correlation coefficients are not significantly different between the two methods (p=n.s.; Fisher r-to-z transformation).  75 High and low risk cases vs. Reproductively Healthy (RH) controls: Our power to detect a significant effect may have been compromised by the heterogeneous nature of our cases and controls. I therefore divided cases into Tow' and 'high' risk groups in two ways: by number of pregnancy losses or by maternal age at the trisomy pregnancy. I also divided the controls into those with (labeled RH controls) or without a proven history of fertility at the age of 37 or later. All of the data presented in this section is based on Southern analysis as there was insufficient data obtained by Q-FISH to divide the groups in this way. When all cases (N=37) are compared to this RH group (N=17) no significant difference in telomere length is detected (p=n.s.; t-test) (Figure 3-7). There is also not a significant difference in telomere length between women who have had more than one trisomic pregnancy (recurrent trisomy - N=22) and women in the control group who have had successful pregnancies after the age of 37 (N=17), (p=n.s.; t-test) (Figure 3-8). However, the RH control group (42.1 years old) is significantly older than the recurrent trisomy mothers (38 years old) (p=0.0001; t-test), and yet their telomeres are the same average length (8.6 kb). The length versus age regressions were also not significantly different between the two groups, but there is a trend towards longer telomeres in the RH group.  76 Telomere length with age in all cases (N=37 in red) and RH controls (N=17 in blue) 18 _  R = -0.18  16  R = -0.31  n 14 Ui  12  c 10  _  _  I  _  O  8 6 4 20  25  30  35  40  45  50  55  60  Age (years)  Figure 3-7: TRF length by age in all cases (N=37 in red) and "reproductively healthy" (RH - children over age 37 years) controls (N=17 in blue). The correlation coefficients are not significantly different (p= n.s.; Fisher r-to-z transformation).  Telomere length with age in High-risk cases (N=22 in red) and RH controls (N=17 in blue)  _  18 16 -  R = -0.03 R = -0.31  14 -  •  5 12c 10 _ a) 8 E  •  #  •  • : '.rfc: ^ •  —'  •  •  " 4 -  6  • * • —  O (J) H 2  0 -i 20  1  1  1  i  i  i  i  i  25  30  35  40  45  50  55  60  Age (years)  Figure 3-8: TRF length by age in high-risk cases (women experiencing >1 trisomy) (N=22 in red) and RH controls (N=17 in blue). The correlation coefficients are not significantly different (p = n.s.; Fisher r-to-z transformation).  77 Although reproductive history is undoubtedly important when defining groups of susceptible women, more important may be the age at which they had their first trisomic pregnancy. As trisomy is known to increase with maternal age, particularly after the age of 35,1 divided our case women into those who experienced their first trisomy under the age of 35 and those who did so at or after this age. I expect that those women under 35 years of age are more susceptible to trisomy as their risk should be lower at this age. This may be due to a premature aging process that was proposed earlier and which should be reflected in their telomere length if our hypothesis is correct. The "young" group was not found to have significantly (p=n.s.; t-test) shorter telomeres than the "older" group (Figure 3-9). When I compared the "young" mothers to controls, there was no significant difference in telomere length. However, based on the regression lines, there does appear to be a trend towards shorter telomere length in the women who were young at age of first trisomy. Telomere length by age in young (N=10 in red) and old (N=27 in purple) cases 18  1  16  I  10  cu  8  E  6  14  _  ° CU  H  4  2 0 20  25  30  35  40  45  Age at blood draw (years)  Figure 3-9: TRF length by age in women experiencing their first trisomy at <35 years old (N=10 in red) and women experiencing their first trisomy at or after 35 years of age (N=27 in purple). The correlation coefficients are not significantly different (p= n.s.; Fisher r-to-z transformation).  78 3.3  DISCUSSION:  Southern analysis vs. Q-FISH analysis in estimating telomere size: Southern analysis is expected to yield a greater value for average telomere length than Q-FISH analysis, as was observed in this study. This is because the digestion of the genomic DNA results in a terminal restriction fragment (TRF) that includes a portion of the subtelomeric region as well as the telomere itself (Figure 3-10). There will be a decreased local concentration of the short fragments and therefore decreased detection of the probe at the shorter lengths compared to the larger TRFs as a result (Poon and Lansdorp, 2001). This may explain the large values for TRF estimates obtained by the Southern method. Krejci and Koch (1998) also suggest a number of faults with the Q-FISH procedure and analysis, mainly related to calibration, which may be the reason for the shorter repeat size estimates obtained by this method when compared to other methods. Thus, the Q-FISH values should be relatively shorter but consistently so.  •  s CD  E o  Unique Sequence  Telomere Associated Repeats (TAR) Terminal Restriction Fragment (TRF)  -< ( T T A G G G ) - telomeric repeats - 3-20 kb in humans n  Figure 3-10: Diagram of the structure of the telomere. Southern analysis measures the length of the telomeric repeats (TTAGGG) as well as a small portion of the telomere associated repeats (TAR). This structure is referred to as the terminal restriction fragment (TRF).  79 In analysis of telomere length by Southern, there is less accuracy in length designation for large sized alleles due to the shorter migration of fragments in the larger size range. As well, the larger fragments will allow more probe to bind resulting in the smear at the larger lengths to appear darker. The software program used to analyze the data may therefore miscalculate the darkest point of the smear leading to an overestimate of the average telomere length in the sample. However, this bias should occur in a consistent fashion for all samples. Although Southern analysis is considered the "gold standard" for telomere length measurement, there are a number of drawbacks related to this method. It is a fairly labour intensive protocol that requires a significant amount of time to complete. The telomere length produced by this method is only an average of all of the telomeres in the cell, and includes the subtelomeric portion as well. It only requires a small amount of DNA (0.5 pg), however, and the protocol used in this study did not involve radioactive detection of the probe. One disadvantage of Q-FISH is that it requires a cultured sample in order to get metaphase chromosomes, which is not always possible to obtain. It requires a fluorescence microscope for detection of the probe and specialized software to analyze the telomere images. The computer analysis is labourious, but the FISH technique itself is not. A major advantage for certain types of studies of telomere lengths is that telomere length measurements for individual chromosomes is possible. As I was more interested in an overall decrease in average telomere length as a function of aging and increased cell turnover, individual telomere length was not looked at in this project.  80 Telomeres and Trisomy Risk: The present results do not support my hypothesis that the telomeres in mothers of trisomic pregnancies are shorter than those found in normal control mothers, as average telomere length was identical for both groups. One major assumption of my hypothesis is that the telomere length found in somatic cells will reflect the telomere length in the ovaries. It is also assumed that an individual with short telomeres at birth will have shorter than average telomeres later in life when compared to someone of the same age. Telomere length is established early in development and a correlation has been observed for blood and various other somatic tissues (Saretzki and VonZglicki, 2002; Butler et al., 1998). However, oocytes have not been analyzed. While I can exclude telomere length as a major risk factor for trisomy, I cannot exclude a small effect that requires a larger sample size to detect or a role in a proportion of cases. Our data does suggest several interesting trends that would be worth following up in a larger study. The subgroup of our controls who had children past the age of 37 (Reproductively Healthy - RH), appear to have the longest average telomere length of all of the groups (8.6 kb) despite being the oldest (avg. age=42.1 yrs). The "high-risk" group with multiple trisomies is younger in age (avg. age=38 yrs), and yet their telomeres are similar in size (8.5 kb) when compared to this "reproductively healthy" group (p=n.s.; t-test). Furthermore the women having trisomy pregnancies at a younger age (<35) did tend to have slightly shorter telomeres. It will thus be important to consider these groups separately in a larger study. It is likely that there are many factors that may contribute to an increased risk for trisomy. It has been suggested that there might be two "hits" to maternal age-related trisomy  81 (Lamb et al, 1996). The first hit would involve the establishment of a suboptimal chiasmate (crossover) configuration in the fetal oocyte; this event would be age independent. The second hit would involve abnormal processing of the susceptible bivalent at metaphase I and would be the age-dependent component of the process. From this explanation, the nondisjunction process is similar in women of different ages, and the increase in trisomy with age is a result of a decrease in efficiency for processing certain types of exchange configurations in the older ovary (Hassold et al, 2000). It is therefore possible that telomere length may be another hit in a multiple-hit model of risk for trisomy, but that alone its effect is difficult to detect without a larger sample size, which would allow for further subdivision by maternal age and pregnancy history. Based on the results presented in this project, telomere length is not a clinically useful factor to assess risk, although it cannot be fully excluded as playing role in a subset of cases. I can conclude from my findings that women at high-risk for having a trisomic pregnancy, based on their history of having >1 trisomy, tend to have shorter telomeres than women who have had no pregnancy losses and have had children after the age of 37, although the difference is not significant. My results are consistent with the idea that women having no pregnancy losses and having children after the age of 37 are at the lowest risk for having a trisomy and this may be a result of their slower aging reflected in their long telomeres.  82 Chapter 4: Replication timing and somatic aneuploidy in women who have experienced a trisomic pregnancy  •  The patients used in the study were collected by referrals from physicians at the Recurrent Pregnancy Loss Clinic (N=27) and the Medical Genetics Clinic at the Women's Health Center at BC Children's and Women's Hospital (N=8), and from other centers in BC (N=3).  •  All FISH experiments were carried out in the lab of Wendy Robinson and I did all of the tissue culturing, harvesting, interphase preparation, FISH experiments, data collection and analysis.  •  Some early data collection and experiments, in order to determine reproducibility of the dot hybridization assay, were done with the help of Maria Pefiaherrera.  •  The HPRT probe (BAC # CTD-2565I18; RP11-671P4) used was labeled with the help of Brenda Lomax in the Cytogenetics Laboratory at Children's and Women's Hospital.  A I M : The aim of this portion of the project is to determine if replication asynchrony and aneuploidy are increased in women experiencing a trisomy pregnancy, as a test of the hypothesis that these women are prematurely aged and are thus exhibiting the levels of replication asynchrony and aneuploidy that are found in women of an older age.  83 4.1  INTRODUCTION: The human genome is divided into a series of well-defined zones that replicate in a  programmed manner throughout S phase of the cell cycle in a manner that appears to be related to gene expression (Amiel et al, 1999). The majority of genes show a biallelic mode of expression and these have been found to replicate highly synchronously. However there are a few classes of genes that have a monoallelic mode of expression - olfactory receptor genes, genes subject to X chromosome inactivation and imprinted genes - and these all replicate in an asynchronous manner (reviewed in Selig et al, 1992; Yeshaya et al, 1998; Amiel et al, 1998). In general, expressed loci replicate early in S phase while silent ones replicate late (Amiel et al, 1999). However, for many imprinted genes, it is the paternal allele that always replicates earlier than the maternal allele (Kitsberg et al, 1993). An increase in replication asynchrony for normally synchronously replicating genes has been associated with a) age, b) having a DS offspring, and c) nondisjunction in lymphocytes (as reviewed in Chapter 1 and below). FISH analysis provides a reliable and accurate method for detecting asynchronous replication, and was first described by Selig et al. (1992). In a non-synchronized population of cells, a high percentage of doublet signals in interphase nuclei is indicative of early replication while more single dots is characteristic of late-replicating DNA (Mostoslavsky et al, 2001). For a more thorough description of the FISH method refer to Chapter 2.4.2. FISH has the advantage that it can be used on small populations of cells and is less labour intensive than BrdU labeling (Simon et al, 1999). BrdU labeling requires that a replicating population of cells be exposed to a "pulse" of BrdU for ~ 4 hours. BrdU is an analog of thymidine (T) and will thus be incorporated into the DNA in place of T during replication,  84 but only in those regions of the DNA that is replicating. Recent studies have shown a relationship between asynchronous replication and Turner syndrome (Reish et al, 2002), Xeroderma pigmentosum patients and carriers (Amiel et al., 2004), Roberts syndrome (Barbosa et al, 2000) and the presence of deletions (Amiel et al, 2002). Amiel and colleagues showed loss of temporal control of replication in both older women and young mothers of children with Down syndrome (DS) as compared to age matched controls (Amiel et al, 2000). This group analyzed replication timing at RB-1 on chromosome 13 and 21q22 and concluded that there is an increased rate of allele asynchrony in mothers with DS children and in women of older age. They suggest that this replication asynchrony may be part of the explanation for their predisposition to nondisjunction. Loss of replication control of DNA leading to asynchronous replication of homologous a-satellite DNA loci (found at the centromeres) in some individuals was also found to be associated with nondisjunction in somatic cells (Litmanovitch et al, 1998). This group looked at lymphocytes from normal females and women with a cancer, and found that there was an increase in replication asynchrony in the women with a predisposition to cancer. As well, those chromosome pairs with the highest level of asynchrony also showed increased rates of aneuploidy. Amiel et al. (2000) also analyzed aneuploidy using a-satellite probes for chromosomes 8 and 18 and concluded that there is an increased rate aneuploidy in mothers with DS children and in women of older age. Aneuploidy in cultures of human peripheral lymphocytes was first reported by Jacobs et al. (1961) and was found to be associated with chronological aging. In females, monosomy for chromosome X is particularly commonly found with increasing age (Fitzgerald et al, 1975; Fitzgerald and McEwan, 1977; Nowinski et al, 1990). An increased  85 rate of aneuploidy in mitotic cells has also been found in individuals predisposed to meiotic nondisjunction, such as parents of a DS child and couples with recurrent spontaneous abortions (Staessen et al, 1983; Ford, 1984; Hecht et al, 1984; Juberg et a/.,1985). However, Nowinski et al. (1990) found no correlation with reproductive history and aneuploidy in mitotic cells, but rather a correlation was found between increasing aneuploidy and age and aneuploidy and gender - specifically, females had a higher rate of aneuploidy. I hypothesize that there is an aging of chromosomes resulting in loss of chromatin integrity with time. These changes in chromatin structure with age result in changes in gene expression, altered replication timing, and increased chance of chromosome missegregation in somatic cells. It is possible that the same effects occur in meiotic cells. If risk of trisomy correlates with the age-related changes in chromatin structure, and these changes show interindividual variability then I expect that women who have experienced a trisomy may have an increased tendency to these signs of somatic cell aging than age-matched controls. In this study replication timing at three loci, the centromere of chromosome 15 (15Z4), 21q22 and HPRT on the X chromosome, was measured in two groups of women to test whether or not replication asynchrony and aneuploidy was increased in women experiencing a trisomic pregnancy. One group had experienced at least one trisomic pregnancy and the other had not. Methods and Materials: Patient samples The samples in this study are from a subset of the trisomic pregnancy mothers and controls as outlined in Chapter 2.1.1.  Replication asynchrony Allele asynchrony was measured in nuclei from peripheral blood samples of women who had experienced at least one trisomic pregnancy compared to control women with no history of pregnancy loss at two different loci. Replication timing was analyzed at the two different loci using probes for the centromere of chromosome 15 (15Z4 labelled with Texas red) and the 21q22 probe (labeled with Texas red) (Vysis) for each sample studied. These probes were chosen to determine if previous studies looking at 21q22 and finding replication asynchrony could be duplicated, and 15Z4 was chosen in order to analyze a centromere. In order to calculate the amount of asynchrony in each sample, 200 nuclei, if possible, were scored and the frequency of nuclei with both a single and a double dot (SD) was calculated. A total of 7 cases and 19 controls were analyzed at the centromere of chromosome 15 (15Z4) and 11 cases and 17 controls were analyzed at 21q22 on chromosome 21. The number of cases and controls looked at for each probe differs as some of the cell pellets for the cases had run out. A dot hybridization assay described by Selig et al. (1992) was used in interphase nuclei that allowed me to calculate the number of nuclei that were replicating asynchronously and compare that to the number that replicated synchronously. I looked at the number of singlets and doublets for each probe in an interphase nucleus and scored it accordingly. Two singlets, indicating that the locus has not yet replicated, are scored as SS, two doublets appear following replication and are scored as DD and a singlet and a doublet found within the same nucleus is scored as SD. This latter pattern of replication indicates that the two alleles have replicated asynchronously - specifically, one allele has already replicated while the other has not.  87 HPRT is a gene located on the X chromosome that replicates asynchronously in each cell cycle. It has been extensively studied to determine the replication timing differences between the active and inactive X chromosomes (e.g. Subramanian and Chinault, 1997), and is known to show an SD signal in a significant portion of cells (-40%). The copy on the active X replicates early while the copy on the inactive X replicates late. Cohybridization of cells with HPRT can thus be used to verify that the cells analyzed are actively replicating. For a subset of study and control women (five women from the control group and two women from the patient group for a total of seven), two FISH probes were used simultaneously, an FITC-labeled HPRT probe located on the X chromosome and one of either the centromere for chromosome 15 or 21q22 labeled with Texas Red. This was done for only a subset of women in order to examine the efficacy of using the HPRT probe by comparing with values obtained without the cohybridization with HPRT as an internal control.  Aneuploidy In order to determine if replication asynchrony was also associated with an increase in somatic aneuploidy, as both have been found with increasing age, the amount of somatic aneuploidy was determined in each sample by counting a minimum of 100 nuclei and scoring the number of signals for probes on chromosomes 15 (a-satellite), 21 (21q22) and X (HPRT). Thefrequencyof one, two, three and four signals was scored for each sample in this way, and the amount of somatic aneuploidy determined. This was done for all loci in 4 control women and 2 case women.  88  4.2  RESULTS:  Chromosome 15 centromere (15Z4)  Replication asynchrony: All data collected at 15Z4 for both cases and controls is found in Table 4-1. The average amount of replication asynchrony at 15Z4 in cases (N=7) with an average age of 36.3 years was 20.7% (Table 4-2). In the controls (N=19) with an average age of 34.6 years, the amount of replication asynchrony was slightly lower at 20.3% (ns; t-test). When both groups are plotted for replication asynchrony with increasing age, the control group shows the expected increase, while the case group shows that there is a decrease with age (Figure 41). The correlation coefficients are not significantly different (ns; Fisher r-to-z transformation) but the trends are in opposite directions. It is the younger women in the case group that appear to have a higher than normal amount of replication asynchrony for their age. The older women have comparable values for their age. I wanted to determine if accelerated aging, as demonstrated by an increase in replication asynchrony, is more important in younger rather than older women, as this seems to be indicated by the trend seen in Figure 4-1. In order to determine whether or not the younger women in the case group have significantly more replication asynchrony at the chromosome 15 centromere than age-matched controls, the data was analyzed for women <35 years of age and > 35 years of age at first trisomy separately. The data is shown in Table 4-3 and no significant difference was found for the women experiencing a trisomy at a younger age compared to those women having a trisomy after the age of 35.  89 Table 4-1: The amount of replication asynchrony observed at 15Z4 in cases (N=7) and controls (N=19). Sample Name  Controls A3 A3 S2 C2 C2 C3 C3 C4 C4 C5 C5 C6 C6 C7 C7 C8 C8 C9 C9 CIO CIO C12 C14 RH1 RH2 RH14 RH15 RH16 RH17 Cases ND22 ND22 ND23 ND23 ND28 ND28 ND29 ND29 ND31 ND31 Cl Cl ND37 ND38  # single dots (SS)  # double dots (DD)  # single dot and double dot (SD)  Total # nuclei analyzed  %  %  %  Asynchrony  SS  DD  30 30 29 36 36 22 22 27 27 32 32 38 38 28 28 21 21 38 38 32 32 28 23 45 42 43 46 58 40  127 160 110 101 141 159 182 110 149 143 153 159 184 145 172 132 172 135 187 151 182 120 138 135 128 70 72 6 29  25 8 53 24 12 16 11 39 9 20 17 10 5 19 11 27 5 28 4 14 5 37 32 15 34 11 21 4 9  50 39 41 75 70 30 18 42 62 37 36 33 25 35 45 44 45 48 34 35 34 47 23 40 38 21 32 4 12  202 207 204 200 223 205 211 201 220 200 206 202 214 199 228 203 222 211 225 200 221 204 193 190 200 102 125 14 51  24.8 18.8 20.1 37.5 31.4 14.6 8.5 20.9 28.2 18.5 17.5 16.3 11.7 17.6 19.7 21.7 20.3 22.7 15.1 17.5 15.4 23 11.9 21.1 19 20.6 25.6 28.6 23.5  62.9 77.3 53.9 50.5 63.2 77.6 86.3 54.7 67.7 71.5 74.2 78.7 86 72.9 75.4 65 77.4 64 83.1 75.5 82.4 58.8 71.5 71.1 64 68.6 57.6 42.9 56.9  12.4 3.9 26 12 5.4 7.8 5.2 19.4 4.1 10 8.3 5 2.3 9.5 4.8 13.3 2.3 13.3 1.8 7 2.3 18.1 16.6 7.9 17 10.8 16.8 28.6 17.6  -  160 196 122 149 106 142 111 165 139 162 142 170 13 39  16 3 41 12 29 21 44 5 24 6 20 11 5 6  25 14 41 51 69 48 52 39 37 36 40 21 3 12  201 213 204 212 202 209 207 209 200 204 202 202 21 57  12.4 6.6 20.1 24.1 34.2 23 25.1 18.7 18.5 17.6 19.8 10.4 14.3 21.1  79.6 92 59.8 70.3 52.5 68 53.6 78.9 69.5 79.4 70.2 84.2 61.9 68.4  8 1.4 20.1 5.7 14.4 10 21.3 2.4 12 2.9 10 5.4 23.8 10.5  Age  40 40 33 33 31 31 38 38 35 35 39 38  90  Table 4-2: Results for replication timing asynchrony, % aneuploidy (single or triple signals) and HPRT as an internal control to ensure nuclei replication. * Replication asynchrony is gnificantly different at 21q22 between cases and controls (p=0.04; t-test). No other si s i g n m c a n i u n i ei e l i t e s w c i e u i % asynchrony Group (N) 20.3 (19) Controls Chromosome  10.7  % single signals (N) 4.9 (N=4)  % triple signals (N) 4.4 (N=4)  %SS  % DD  68.7  15(15Z4)  Cases  20.7 (7)  70.6  10.6  3.7 (N=2)  4.8 (N=2)  Chromosome  Controls  24.9(17)*  61.6  13.5  8.9 (N=4)  4.5 (N=4)  21 (21q22)  Cases  28(11)  56  15.7  4.5 (N=2)  4.0 (N=2)  Chromosome X  Controls  -  6.8 (N=4)  5.1 (N=4)  (HPRT)  Cases  -  8.6 (N=2)  3.5 (N=2)  HPRT with  Controls  24.5 (4)  56.5  18.5  15Z4  Cases  17.5 (2)  65.2  17.2  HPRT with  Controls  26.8 (5)  54.4  18.4  21q22  Cases  23.5 (2)  66.1  15  •  Replication asynchrony with age at 15Z4 in Cases (N=7) and Controls (N=19)  35 30 c 25 o _ 20 o c >. 15 (0  =n  R2  •  R = 0.3049  •  2  *  —•—  •  r*^ * • •  10  •  5 0  10  20  30  40  50  60  70  Age (years)  Figure 4-1: Replication asynchrony with age at 15Z4 in cases (N=7 in red) and controls (N=19 in blue).  91 Aneuploidy: The amount of somatic aneuploidy at 15Z4 in cases (N=2) was compared to controls (N=4). The average amount of single signals was 3.7% for the cases, while the controls had 4.9% of their nuclei with a single signal. The average amount of triple signals in the cases was 4.8%, while it was 4.4% in controls. All data can be found in Table 4-5, while the mean values are found in Table 4-2. Chromosome 21 (21q22)  Replication asynchrony: All data collected at 21q22 for both cases and controls is found in Table 4-4. The average amount of replication asynchrony at this locus in cases (N=l 1) (average age of 36.1 years) was 28%. In the controls (N=l7) with an average age of 35.4 years, the amount of replication asynchrony was significantly lower at 24.9% (p=0.04; t-test). When both groups are plotted for replication asynchrony with increasing age, again the control group shows the expected increase, while the case group shows that there is a decrease with age (Figure 4-2). The correlation coefficients are not significantly different (p=ns; Fisher r-to-z transformation). In order to determine if this increased replication asynchrony is more important in younger rather than in older women, the groups were divided based on age at first trisomy. When younger case women (< 35 years of age at first trisomy) were compared to control women of the same age for replication asynchrony at 21q22, there was a significant difference in the amount of replication asynchrony (p=0.001; t-test). The control group (N=9) had an average age of 29.2 years and an average amount of replication asynchrony of 23.4%, compared to the case group (N=5) with an average age of 30.8 years and 30.4%  92 replication asynchrony. There was no significant difference between the older women (> 35 years of age) in the case and control groups.  Table 4-3: When replication asynchrony is analyzed at each locus based on age at first trisomy, there is a significant difference seen when comparing the younger members of each group (<35 years of age) at 21q22 (*p=0.001; t-test). No other significant differences are seen. Chromosome IS (15Z4)  Chromosome 15 (15Z4)  Chromosome 21 (21q22)  Chromosome 21 (21q22)  Group  % asynchrony (N)  Controls <35 years old  18.1 (10)  Cases <35 years old  25 (2)  Controls >3S years old  22.7 (9)  Cases >35 years old  19(5)  Controls <35 years old  23.4 (9)*  Cases <35 years old  30.4 (5)*  Controls >35 years old  26.6 (8)  Cases >35 years old  26.5 (6)  Table 4-4: The amount of replication asynchrony observed at 21q22 in cases (N=l 1) and controls (N=17). Sample Name  Controls A3 A3 A3 A3 S2 C2 C2 C3 C3 C4 C4 C5 C5 C6  Age  30 30 30 30 29 36 36 22 22 27 27 32 32 38  # single dots (SS)  # double dots (DD)  # single dot and double dot (SD)  Total # nuclei analyzed  % Asynchrony  147 128 118 139 110 106 136 137 163 144 151 155 117 118  12 23 29 15 53 65 15 17 17 19 12 15 14 37  47 51 48 58 41 31 49 45 46 42 49 21 75 52  206 202 195 212 204 202 200 199 226 205 212 191 206 207  22.8 25.2 24.6 27.4 20.1 15.3 24.5 22.6 20.4 20.5 23.1 11 36.4 25.1  %  %  SS  DD  71.4 63.4 60.5 65.6 53.9 52.5 68 68.8 72.1 70.2 71.2 81.2 56.8 57  5.8 11.4 14.9 7.1 26 32.2 7.5 8.5 7.5 9.3 5.7 7.9 6.8 17.9  93 C6 C7 C7 C8 C8 C8 C8 C9 C9 C9 C9 CIO CIO C12 RH5 RH14 RH15 RH16 RH17 Cases ND21 ND21 ND21 ND21 ND22 ND22 ND23 ND23 ND25 ND25 ND26 ND26 ND28 ND28 ND29 ND29 ND31 ND31 ND32 ND32 Cl Cl ND37 ND38  38 28 28 > 21 21 21 21 38 38 38 38 32 32 28 35 43 46 58 40  124 107 121 46 130 123 154 116 125 121 111 112 104 120 40 38 17 40 59  19 40 23 2 16 37 13 24 21 28 30 38 46 37 21 25 1 21 13  61 52 62 8 53 45 42 69 60 54 70 55 54 47 33 31 4 26 18  200 199 206 56 199 205 209 209 206 203 211 205 204 204 94 94 22 87 90  30.5 26.1 30.1 14.3 26.6 22 20 33 29 26.6 33.2 26.8 26.5 23 35 33 18.2 30 20  62 53.8 58.7 82.1 65.3 60 73.7 55.5 60.7 59.6 52.6 54.6 51 58.8 42.6 40.4 77.3 45.9 65.6  9.5 20.1 11.2 3.6 8 18 6.2 11.5 10.2 13.8 14.2 18.5 22.5 18.1 22.3 26.6 4.5 24.1 14.4  37 37 37 37 -  107 130 117 136 95 97 120 107 123 120 105 125 107 117 101 110 129 116 109 78 130 106 18 42  46 30 24 23 30 47 33 38 25 28 32 23 48 18 36 21 18 32 56 41 28 32 7 5  52 46 64 50 79 62 50 59 55 62 64 54 45 70 66 80 59 52 47 69 61 68 9 6  205 206 205 209 204 206 203 207 203 207 201 203 200 205 203 211 206 200 212 188 219 206 34 53  25.4 22.3 31.2 24 38.7 30 24.6 28.5 27.1 30 31.8 26.6 22.5 34.1 32.5 37.9 28.6 26 22.2 • 36.7 27.9 33 26.5 11.3  52.2 63.1 57.1 65 46.6 47.1 59.1 51.7 60.6 58 52.2 61.6 53.5 57.1 49.8 52.1 62.6 58 51.4 41.5 59.4 51.5 52.9 79.2  22.4 14.6 11.7 11 14.7 22.8 16.3 18.4 12.3 13.5 15.7 11.3 24 8.8 17.7 10 8.7 16 26.4 21.8 12.8 15.5 20.6 9.4  40 40 41 41 33 33 33 33 31 31 38 38 32 32 35 35 39 38  94  Replication asynchrony with age at 21q22 in Cases (N=11) and Controls (N=17) 40  n 15 55  10 5  0  4— 0  1  1  10  20  — — i  30  '  '  40  50  1  60  70  Age (years)  Figure 4-2: Replication asynchrony with age at 21q22 in cases (N=l 1 in red) and controls (N=17 in blue). Aneuploidy: The amount of somatic aneuploidy at 21q22 in cases (N=2) was compared to controls (N=4). The average amount of nuclei with single signals was 4.5% for cases and 8.9% for controls. Triple signals were found in 4.0% of case nuclei and in 4.5% of control nuclei. A l l data can be found in Table 4-5, while the mean values are found in Table 4-2.  Chromosome X (HPRT)  Replication asynchrony: The HPRT locus was used as an internal control for replication status of nuclei analyzed in this study. In my hands HPRT was found to replicate asynchronously (an SD pattern of replication) in 39.7% of nuclei analyzed (data not shown).  95 Aneuploidy: Cases (N=2) had a mean of 8.6% of their nuclei with single signals, while controls (N=4) were observed to have a mean of 6.8%. Triple signals were found in 3.5% of nuclei in cases and 5.1% of nuclei in controls. All data can be found in Table 4-5, while the mean values are found in Table 4-2. Table 4-5: Somatic aneuploidy data for cases (N=2) and controls (N=4) at all 3 loci. FISH efficiency, as determined by the number of 0 signals is included. Sample Name Controls RH14 RH14 RH14 RH15 RH15 RH15 RH16 RH16 RH16 RH17 RH17 RH17 Cases ND37 ND37 ND37 ND38 ND38 ND38  % single signals  % triple signals  FISH eff.  114 137  4.4 5.8  7.9 8.0  1.7 0.7  0 0 0  102 114 105  3.9 8.8 4.8  6.9 5.3 4.8  0 2.6 2.8  7 3 5  2 0 0  106 101 107  3.8 5.0 10  6.6 3.0 4.7  3.6 1.9 1.8  60 92 89  1 1 3  0 0 0  63 111 98  3.2 16 6.1  1.6 0.9 3.1  1.6 0.9 3.9  4  91  4  0  99  4.0  4.0  2.9  5 2  4 12  90 83  5 4  0 0  99 99  4.0 12  5.1 4.0  4.8 2.0  2 2 0  3 5 5  92 93 91  5 3 3  0 0 1  100 101 100  3.0 5.0 5.0  5.0 3.0 3.0  2.0 1.9 0  0  7  93  4  0  104  6.7  3.8  0  Probe used  #0 signals  #1 signal  #2 signals  #3 signals  #4 signals  21q22 HPRT (X) 15Z4 21q22 HPRT (X) 15Z4 21q22 HPRT (X) 15Z4 21q22 HPRT (X) 15Z4  2 1  5 8  99 118  9 11  1 0  0 3 3  4 10 5  91 98 95  7 6 5  4 2 2  4 5 11  93 93 91  1 1 4  2 18 6  3  21q22 HPRT (X) 15Z4 21q22 HPRT (X) 15Z4  Total  HPRT subset - test for replication  A probe for HPRT, a gene on the X chromosome, was used in combination with both 15Z4 (N=7) and 21q22 (N=6) in a subset of women in order to identify the replicating cells. The data can be found in Table 4-6. The amount of replication asynchrony at 21q22 and 15Z4 was not significantly different in either cases or controls when compared to the rest of  96 the group in which the HPRT internal control was not used (Table 4-2). The frequency of SD, SS and DD nuclei did not differ when HPRT was used as an internal control. It is therefore not necessary to use it in the future.  Table 4-6: The amount of replication asynchrony observed at 15Z4 and 21q22 in cases (N=2) and controls (N=5 for 21q22 and N=4 for 15Z4) using HPRT as an internal control; i.e. HPRT was always observed as an SD pattern before the nucleus was scored for the other probe. Sample Name  Controls RH5 RH14 RH14 RH15 RH15 RH16 RH16 RH17 RH17 Cases ND37 ND37 ND38 ND38  4.3  Probe used (HPRT always SD)  # single dots (SS)  # double dots (DD)  # single dot and double dot (SD)  Total # nuclei analyzed  %  %  %  Asynchrony  SS  DD  21q22 21q22 15Z4 21q22 15Z4 21q22 15Z4 21q22 15Z4  40 38 70 17 72 40 6 59 29  21 25 11 1 21 21. 4 13 9  33 31 21 4 32 26 4 18 12  94 94 102 22 125 87 14 90 51  35 33 20.6 18.2 25.6 30 28.6 20 23.5  42.6 40.4 68.6 77.3 57.6 46 42.8 65.6 56.9  22.3 26.6 10.8 4.5 16.8 24.1 28.6 14.4 17.6  21q22 15Z4 21q22 15Z4  18 13 42 39  7 5 5 6  9 3 6 12  34 21 53 57  26.5 14.3 11.3 21.1  52.9 61.9 79.2 68.4  20.6 23.8 9.4 10.5  DISCUSSION: The results from this study are consistent with the hypothesis that accelerated  chromatin aging may be present in young women who have experienced a trisomic pregnancy. While our numbers of samples were limited, and no significant differences between the groups at any of the three loci could be found, an interesting trend is seen. The younger women have a higher amount of replication asynchrony when compared to control women of the same age, while the older women have a similar amount of replication  97 asynchrony as controls of the same age. This suggests that abnormalities of the chromatin reflected in replication timing may play a role in trisomy occurring in younger women. The dot hybridization FISH assay determines if two alleles of a gene or a locus are replicating synchronously or asynchronously during S phase of the cell cycle. Afractionof cells will show replication asynchrony of the alleles at any locus even if they are truly replicating highly synchronously. This is not due to temporal differences in replication timing but rather to a limitation of the FISH assay that mistakenly identifies about 10% (from 7.8 to 11.6%) of doublets as singlets (Selig et al., 1992). At the chromosome 15 centromere, there was no difference found in the amount of replication asynchrony between the two groups of women. At the locus on chromosome 21 (21q22) a significant difference was observed (p=0.04; t-test) between case and control women. This is even more pronounced when the group is divided based on age. The younger members of the case group (< 35 years old) have a significant increase in replication asynchrony when compared to the younger controls group members (p=0.001; t-test). It is interesting to note how similar the trend lines for both groups appear at each locus. In a study of genes in the immune system, Mostoslavsky et al. (2001) determined that 40% of the nuclei showed an SD pattern, while normal control genes were said to be replicating relatively synchronously when 20% of the nuclei had an SD pattern. Similar results were also found by Singh et al. (2003) who estimated that asynchronously replicating genes present the SD pattern in 20-40% of cells and synchronously replicating genes presented this SD pattern in ~10-15% of cells. Both groups in my study appear to be replicating synchronously at all loci studied. There is however, a significant increase in replication asynchrony at 21q22 in the case group compared to controls. Amiel et al. (2000)  98 found a significantly higher rate of allele asynchrony in mothers of Down syndrome children and middle-aged women when compared to young control women. In my study, the younger mothers of trisomy did however, have a higher amount of allele asynchrony than the control women of the same age. Previous studies have also found an increase in aneuploidy in the somatic cells of mothers of DS (Amiel et al, 2000) and in women with cancer (Litmanovitch et al, 1998). My numbers were too small to be conclusive about my aneuploidy studies and need to be followed up with larger samples. However, there does not appear to be a trend towards an increase in aneuploidy in the somatic cells of women experiencing a trisomy in my case group. It would be worthwhile to increase the number of samples and collect more data in order to more conclusively determine whether or not there is any association between having a trisomy and having an increased amount of replication asynchrony and/or somatic aneuploidy. Including HPRT in these future studies is not worthwhile as there is no significant difference between the % SS, %DD or %SD when comparing samples in which HPRT was used and not used.  99 i  •  Chapter 5: Recurrent trisomy 21 in a Northern BC family  This family was ascertained through Dr. Laura Arbour and genetic counselor Rosemarie Rupps of the Genetic Clinic at Children's and Women's Hospital.  •  Karyotyping of the women experiencing a trisomy 21 pregnancy was done at the Cytogenetic Clinic at Vancouver General Hospital.  •  All previous karyotyping of the trisomy 21 individuals was done at other centers in BC.  •  I did all of the tissue culturing and harvesting of the blood samples as well as the DNA extraction.  •  I did the microsatellite analysis with the help of Ruby Jiang, a technician in the lab of Wendy Robinson.  •  MTHFR genotyping was done with the help of Luana Avila.  •  Helene Bruyere and Clare Jensen at the Cytogenetics Laboratory at Vancouver General Hospital carried out many of the FISH experiments.  •  I did all of the analysis of the data.  A I M : The aim of this chapter is to determine if the recurrent trisomy 21 seen in the presented family is caused by an inherent problem with chromosome 21. I tested the hypothesis that there is a cryptic rearrangement of chromosome 21 segregating in this family predisposing them to trisomy 21. A version of this chapter has been submitted accepted for publication. Gair, JL, Arbour, L , Rupps, R, Jiang, R, Bruyere, H and W P Robinson. Recurrent Trisomy 21: Four Cases in Three Generations. Clinical Genetics. 1  100 5.1  INTRODUCTION: Trisomy 21(T21) is the most common trisomy found in newborns with an incidence  of 1/700. This frequency increases considerably with maternal age, rising from 1/1587 for 20 - 24 year olds to 1/24 if the mother is over 45 years of age (Hook, 1976). This high rate of nondisjunction is predominantly the result of an error occurring during meiosis in oogenesis (Lamb et al, 1996; Hassold and Hunt, 2001). In addition to the increased risk with age, couples with a previous trisomy 21 pregnancy have a significantly increased recurrence risk above that expected for their age (Warburton et al, 2004). The recurrence of trisomy 21 in multiple pregnancies from the same couple is generally thought to be due to gonadal mosaicism in one or the other parent (Bruyere et al, 2000; Pangalos et al, 1992; Warburton et al, 2004). Chance may also account for some instances of recurrence particularly when the mother is older (e.g. >35 years). However, other explanations, such as a mutations affecting chromosome segregation at meiosis (Hunt and Hassold, 2002), variation in recombination rates (Lynn et al, 2000; Schon et al, 2000) or rates of ovarian aging have been proposed (Kline et al, 2000). Cryptic translocations involving the pericentromeric regions of acrocentric chromosomes could also result in abnormal pairing at meiosis leading to increased nondisjunction in carriers (Cockwell et al, 2003). There has been little direct support for such hypotheses however, as pedigrees presenting convincing evidence for an inherited predisposition to nondisjunction over multiple generations have not been reported. I hereby present a pedigree (Figure 5-1) with familial trisomy 21. Four cases of Down Syndrome have been born within three generations of this family. There is no cytogenetic evidence for a translocation involving chromosome 21 segregating in this family, nor does advanced maternal age explain the findings, as all mothers were under the age of 30  101 at the time of the affected pregnancies. Nonetheless, a possible cryptic abnormality of chromosome 21 could be segregating in this family. To test this hypothesis I carried out a number of studies looking at chromosome 21 segregation and structure.  Mat a g e : -30  I-3  47,XX,+21  ( M M )  47,XX,+21 Mat age : 29  ( ^ )  ( M M ^  47,XX,+21 Mat age : 21  IV 47,XY,+21  o  Mat age : 18 T21 affected  Figure 5-1: Recurrent trisomy 21 in a family  from Northern British Columbia.  102 Methods and Materials: This family was ascertained through the BC Provincial Medical Genetics Clinic through a Down syndrome (DS) child born to an 18 year old female member. Her family history of DS (Figure 5-1) was extensive and unusual in that three other members of her family, within three generations, had children with DS. DNA from nine family members was obtained for genetic studies of the recurrence of trisomy 21. The DS individuals had all been karyotyped andfreetrisomy 21 was found. The mothers of the DS individuals were also karyotyped and found to have normal female karyotypes. Blood was drawn for DNA and chromosome analysis from multiple family members, after informed consent.  j  Cytogenetic and FISH Studies The metaphase chromosomes of three individuals, II-3 and III-13 and II-1, were repeat G-banded to further exclude the presence of a translocation that could explain the recurrence of T21. FISH (fluorescence in situ hybridization) analysis of stimulated peripheral blood lymphocytes was performed on two members of this pedigree, II-3 and III13, to further exclude a cryptic translocation between the centromere of chromosome 21 and another acrocentric chromosome. A Vysis probe for 21q22, labelled with rhodamine, as well as a Cytocell probe for the centromere (a-satellite) of chromosome 13/21, labeled with Texas red, was hybridized to metaphase spreads. Somatic nondisjunction for chromosome 21 was measured by using the Vysis 21q22 probe in interphase nuclei to look for an increase in aneuploidy for chromosome 21. Probes for the centromere of chromosome 15 as well as the HPRT - a locus.on the X chromosome were also analyzed for somatic aneuploidy (BAC CTD-2565I18). Approximately 100 interphase nuclei/sample were analyzed to look for signals from chromosome 21q22,15Z4  103 and HPRT. The number of nuclei containing three signals was compared to the number containing two, and percent aneuploidy was calculated as the number of nuclei with three signals divided by the total number analyzed. The number of loci containing one signal was also calculated. Molecular Studies Molecular analysis using 20 highly polymorphic markers spanning chromosome 21 was performed in order to determine the origin of the extra chromosome 21 in the Down syndrome individuals and the haplotypes segregating in this family (Table 5-1). DNA was extracted from peripheral blood by conventional methods. PCR amplification was performed on an MJ Research thermocycler with 35 cycles of 30 seconds at 94°C for denaturation, 55°C for annealing, and 72°C for elongation. The PCR products were then visualized by one of two methods: (1) silver-stained polyacrylamide gels and (2) automatedfluorescenceanalysis. For silver staining, approximately 5 pi of the product was mixed with an equal volume of urea loading buffer, denatured for 5 minutes at 94°C and loaded on a 6% polyacrylamide gel. For automatedfluorescentanalysis, PCR was performed with a forward primer labeled with the ABI Prism Dyes HEX or 6-FAM. The amplification products were sized using capillary electrophoresis on an ABI Prism 310 genetic analyzer. Fluorescence was detected by ABI Prism data-collection software and analyzed by use of GeneScan software. Dosage of alleles can be estimated by relative peak height in heterozygous individuals.  104 Table 5-1: Microsatellite markers spanning chromosome 21 were analyzed in all members of the pedigree and the genotype data is found below. Marker  cM  2503J9TG* CEN 0.6 D21S1904 0.6 D21S215* 5.77 D21S369* 5.87 D21S120* 6.52 D21S258* 6.53 12.4 D21S409 D21S1899 16.13 D21S11* 17.4 D21S1257 26 D21S265 29.5 D21S1262 44.6 D21S259 52.7 D21S1809* 53.2 D21S1440* 56.8 D21S1224* 63.8 D21S212* 68.5 D21S171 68.6 D21S112 D21S1446 69.3  1-2  1-3  11-2  11-3  11-4  11-5  111-13  11-1  111-11  acd  aa  ac  acd  ac  aa  aa  bee  acd  cc aab bed bb be ab ce bb ui bb cd ac ab ac dd be ab ui ad  be ac cd be ab ab ad ab ui bb bd dd ab ab be be bd ui bb  be aa cc be ab bb ac bb ui bb dd ad ab ac cd ce bd ui bd  cc abc edd bb bb ac cd ab ui bb be ad aa be bd bb ab ui ab  be aa cc be ab ab ae bb ui bb dd cd bb ac cd be bd ui bd  be aa be be ab ab ae bb ui bb dd ad ab ac cd ce bd ui bd  ab aa be be ab ab ad bb ui bb bd ab ac cc bd be bb ui bd  acc abc acd abb bbc aaa bee bbc ui abb abd bdd acc bbb add ade bec ui bed  bec aaa  ,  bbc abd abc ade abb ui bbc edd edd bbb aac cde bee bbd ui bed  "Distance in cMfromchromosome 2 lp telomere for female meiosis. * Analyzed on ABI 310 Genotyper. Heparinized blood samples available for FISH. http://cedar.genetics.soton.ac.uk/pub/chrom21 /gmap A  Definitive haplotype phase is unknown for individual 1-2 and her partner. It was therefore estimated using the genotype informationfromthe four children II-2, II-3, II-4 and II-5 by i) minimizing the number of crossovers by assuming a crossover in only one of the four transmissions was always more likely than that the three remaining cases had recombined between the same two markers and ii) when two of the four transmissions must have had a crossover (i.e. from D21S1262 to D21S259 and from D21S1440 to D21S212) then assuming the combination of crossovers which would avoid more than three crossovers on one chromosome (which would be highly unlikely for chromosome 21). MTHFR genotyping  For the methylenetetrahydrofolate reductase (MTHFR) polymorphism studies, the study group consisted of six women from the pedigree, all considered to be high-risk for  105 producing a trisomy 21 child (i.e. 1-2, II-2, II-3, II-4, II-5 and III-13). The genotype and allele frequencies for the MTHFR C677T polymorphism were compared to published controls obtained from Arbour et al. (2002). The genotype and allele frequencies for the MTHFR A1298C polymorphisms were compared to controls taken from a control group used in another study. These five control mothers were all women who had no history of trisomy or pregnancy loss and were genotyped for the A1298C polymorphism. Telomere length was measured and analyzed, as described in Chapter 2, for four of the mothers in this family (II-3, 11-4,11-5 and III-13). 5.2  RESULTS:  Exclusion of translocation: After G-banding at a 400-450 band resolution performed at the Cytogenetics Laboratory at Vancouver General Hospital in Vancouver, BC, both II-3 and III-13 showed a normal female karyotype (46,XX) and II-1 showed a female karyotype with an extra chromosome 21 (47,XX,+21). Specifically, no translocation involving chromosome 21 was apparent. Although a higher G-band resolution would be desirable, a low number of quality metaphases were present and attributed to transport time between the blood collection location and Vancouver. Repeat blood samples could not be obtained. To further exclude a translocation, FISH with the probes for the centromere of chromosome 13/21 and 21q22 was performed in two mothers of trisomy (II-3 and III-13). Six signals on each metaphase spread were observed. One signal at the centromere was seen on each of the two chromosomes 13 and two signals were seen on each of the two chromosomes 21 (Figure 5-2) in all but one cell analyzed. In a single cell of 10 cells analyzed from 11-3, trisomy 21 was noted by FISH. As this could not be confirmed, this  106 could be a culture artifact or inconsequential low-level mosaicism. This demonstrates that the centromere of chromosome 21 is unlikely to be translocated to another chromosome. However, a translocation between chromosomes 13 and 21 cannot be ruled out as the centromeric a-satellite probe used is cross specific for 21 and hybridizes with 13. Also noted in Figure 5-3 was an apparently amplified signal at the 21 centromere in II-3 as well as a diminished signal at 21q22 in this same sample. An inversion involving a break within the region that the probe for 21q22 hybridizes to and a break in 21 ql 1.2 somewhere between D21S369 and D21S258 could produce such a signal pattern. This would result in a split signal at 21 q22 such that a portion of it would be at 21 q22 and the other at 21 q 11.2.  Figure 5-2: FISH with probes for the 13/21 centromere and 21q22 was performed on samples from individuals II-3 and HI-13. This image is from sample II-3. Two signals on each of the chromosomes 21 and one signal on each of the chromosomes 13 was observed. No translocation to another chromosome could be detected.  Figure 5-3: An amplified signal at the 21 centromere (white arrow) in II-3 as well as a diminished signal at 21q22 (red arrow) on one of the chromosomes 21 in this same sample was observed, prompting investigations into a possible inversion.  108  Exclusion of inversion: In order to confirm or exclude an intrachromosomal inversion with 21q22 breakpoint, FISH with the same 13/21 a-satellite probe and the 21q22 probe was done on this same individual (II-3) separately. No split signal for 21q22 was seen and normal signals were seen for all chromosomes (Figure 5-4). However, this does not conclusively exclude the possibility that there is an inversion involving the centromere that can explain the molecular results for this individual.  Figure 5-4: When each FISH probe (21q22 on the left metaphase and 13/21 centromere on the right metaphase) was hybridized to samples from II-3, no split signals were observed.  109 Somatic Aneuploidy: The levels of both monosomy and trisomy of chromosomes 21,15 and X were measured in cultured lymphocyte nuclei of two members of this family (II-3 and II-4) (Table 5-2) and compared to controls. The amount of triple signals was not significantly different at any of the three loci, most importantly, T21 was not found to be increased in the lymphocytes in these individuals. The only significant difference was seen at 15Z4. The two family members had an increased amount of single signals for this locus compared to controls (p=0.008; t-test). However, this is not significant if multiple comparisons are accounted for, if a non-parametric test is used or if all samples are lumped together and compared. Therefore, there appears to be no difference between the mothers and controls for amount of somatic aneuploidy, with sample size being an obvious limiting factor.  Table 5-2: Somatic aneuploidy in controls (N=4) compared to 2 of the mothers from the Group  Probe  % single signals  % triple signals  Controls (N=4)  15Z4  3.7*  4.8  T21 family (N=2)  15Z4  7.9*  2.6  Controls (N=4)  21q22  8.9  4.5  T21 family (N=2)  21q22  8.6  5.7  Controls (N=4)  HPRT (X)  6.8  5.1  T21 family (N=2)  HPRT (X)  5.5  6.0  110 Origin of Trisomy: Table 5-1 shows the microsatellite marker typings along chromosome 21 from all available samples in this pedigree. DNA samples were available from only two of the individuals with trisomy 21 (II-1 and III-l 1). In the trisomic individual III-11, there are three distinct alleles amplifying from multiple microsatellite marker loci (2503J9TG, D21S258 and D21S409) confirming a meiotic origin for the extra chromosome 21. Although the father was not available, the marker inheritance pattern was consistent with a maternal origin of the extra chromosome for all markers. Maternal heterozygosity was retained for the marker closest to the centromere on the p arm (2503J9TG) and at D21S258 (at 6.5 cM) on the q arm making an error in maternal meiosis I the most likely origin of the chromosome. Neither parent was available for individual II-1, however, three alleles were observed at proximal markers suggesting also a meiosis I origin. Duplication of the centromeric region in two individuals: During the process of genotyping individuals in this pedigree, three alleles were observed at marker D21S369 in 1-2 and at D21S215 in her daughter II-3. Allelic dosage differences and repeat genotyping confirmed an apparent duplication of these loci in both individuals. Further markers in the region were tested and 2503J9TG (pi 1.1) also showed an extra allele, suggesting the presumed duplication spans the chromosome 21 centromere. This extra copy does not seem to be cosegregating with the other copies of 21pter-21ql 1.1 in these individuals. As a 13;21 cryptic translocation could not be excluded by FISH, proximal chromosome 13 markers were typed to determine if the duplicated 21ql 1.1 markers cosegregated with proximal 13. The duplicated chromosome 21 markers appear to cosegregate with D13S633, which is the closest informative marker to the chromosome 13 centromere  Ill (Table 5-3). The mother 1-2 has transmitted allele "a" to only the daughter II-3, whom also inherited the supernumerary chromosome 21 alleles, while allele "b" was transmitted to her three other daughters tested. A 3:1 segregation pattern with II-3 inheriting the unique chromosome 13 allele from her mother would have 1/8 probability to occur by chance. Thus, the data is suggestive of a cryptic 13;21 translocation accounting for the supernumerary chromosome 21 alleles. All of the duplicated markers are located within a repetitive region present in multiple copies elsewhere in the genome and thus single copy FISH probes including this region are not available. Furthermore, it was not possible to obtain additional blood samples to further confirm such a translocation. Table 5-3: Microsatellite marker results for chromosome 13 markers. D13S633 was informative and was found to cosegregate with the duplicated alleles on chromosome 21 in the same individuals (1-2 and II-3). Marker  CEN D13S141 D13S1236 D13S115 D13S633 D13S232  cM (FCM) Map location  1.37 4.03 7.02 7.14 9.24  19622459 21594180 21740491 21258132 22697762  I-2  1-3  11-2  11-3  11-4  11-5  cc aa bb «ib ab  be ab be ab ac  cc ab be bb aa  be aa bb aa ac  cc ab be bb ba  cc ab be bb ba  Haplotype sharing and recombination: For the haplotype reconstruction (based on minimizing crossovers) presented in Figure 5-5, no obligatory paternal crossovers occurred in any of the transmissions from 1-3 to his four daughters. As the male genetic distance between 2503J9TG and D21S171 is 48.5 cM one would expect only -50% of paternally derived chromosomes 21 to show no crossovers. In contrast, for the transmission of chromosome 21 from 1-2 to her four daughters, a minimum of 10 crossovers is required. An additional crossover is observed in the transmission of chromosome 21 from II-5 to her daughter III-13. All crossovers were confirmed either by using additional nearby markers and/or repeating marker typings for  112 verification. Thus the distribution of 0:1:2:3 crossovers is estimated as 0:1:2:2 in this family. The comparable distribution in a study of segregations from CEPH controls was 33:37:8:0 (Bugge et al., 1998). It thus appears that recombination was unusually frequent on chromosome 21 in this family with at least 4 of 5 segregations showing >1 crossover as compared to only 10% of controls. No evidence for elevated recombination on chromosomes 15 and 16 was observed in this same pedigree (data not shown) suggesting this effect is limited to chromosome 21. No crossovers are observed in the nondisjunction event leading to the trisomy 21 child of II-4, consistent with the known association between reduced recombination and nondisjunction.  d  b b  Figure 5-5: Haplotype reconstruction based on the microsatellite marker typings. Two individuals were found to have 3 alleles at 3 loci (as shown by the red shaded in box). Each shift in colour along either the maternal chromosomes (orange and yellow) or the paternal chromosomes (purple and blue) indicates a crossover. The red box enclosing two alleles is the only region shared by all women in the pedigree.  113 Molecular Studies: MTHFR analysis  Based on the fact that an association exists between abnormal folate and methyl metabolism, DNA hypomethylation and abnormal chromosome segregation, we examined the genotype at the MTHFR gene in this family. In previous studies, mothers of DS children had an increasedfrequencyof certain polymorphisms in methylenetetrahydrofolate reductase (MTHFR, C677T and A1298C) and methionine synthase (MTRR, A66G). Table 5-4 shows the results from the molecular analysis of the two polymorphisms at the MTHFR gene in this family compared to controls. Table 5-4: Genotyping data for both polymorphisms at MTHFR for all family members (N=9) and controls (N=8). Sample name  Genotype MTHFR C677T  MTHFR A1298C  TT CT TT TT TT TT CT CT CC TT  AC AC AA AA AA AA  111-13 11-1 111-11 C2 C3 C4  Mother Father No children Mother Mother Mother Mother Trisomy 21 Trisomy 21 control control control  C5 C9 C10 90-11-002 90-10-043  control control control control control  CT TT CT CT CT  1-2 1-3 11-2 11-3 11-4 11-5  CC TT  AC AC AA AA AC AC AA AA  /  MTHFR C677T  Thefrequenciesof the susceptible genotypes, determined in previous studies (James et al., 1999; Hobbs et al., 2000), homozygous mutants and the mutant alleles were all  114 significantly higher in the study group when compared to published controls (Arbour et al., 2002 ) (Table 5-5). However, all of the women in the study group were members of the same family and would therefore be expected to have correlated genotypes reducing the significance of such an observation. Women either having a trisomy 21 child or considered to be at risk for having a trisomy in this family were included as cases. Appropriate controls were taken from Arbour et al. (2002), as this T21 family is First Nations and I did not have ethnically comparable controls. It is important to note, however, that considerable variation in allele frequencies may still exist among different First Nations populations. Table 5-5: Frequency of maternal MTHFR C677T genotypes in case and control groups. Women with MTHFR polymorphisms (%) Groups  N  CC  CT  TT  Case  5  0  1 (20)  4(80)  Control  11  8(70)  2(21)  1(9)  MTHFR A1298C The frequencies of the susceptible genotypes, homozygous mutants and mutant alleles were not different when comparing the study group to control mothers not experiencing a trisomy or pregnancy loss (Table 5-6).  Table 5-6: Frequency of maternal MTHFR A1298C polymorphisms in case and control groups.  Women with MTHFR polymorphisms (%) Groups  N  AA  AC  CC  Case  5  3 (60)  2(40)  0  Control  5  3 (60)  2(40)  0  115 Telomere length analysis: Telomere length was measured in this family to see if unusually shortened telomeres could be a contributing factor to the high rate of nondisjunction in this family. Figure 5-6 shows that although the average telomere length for the 4 trisomy mothers in this family is slightly shorter (7.1 kb) than the 46 control mothers (8.8 kb), the difference is not significant (p=ns; t-test). The correlation coefficients are significantly different however (p=0.04; Fisher r-to-z transformation), and specifically it appears that the mothers from the T21 family have a more rapid decline in telomere length with age than the control group. However, caution should be taken in this interpretation given the small sample size.  T e l o m e r e Length by A g e in T21 Family Mothers and  Controls  0  10  20  30  40  50  A g e (years)  Figure 5-6: Telomere length as measured by Southern analysis in 4 mothers from this family (N=4 in red) compared to controls (N=46 in blue).  116 5.3  DISCUSSION:  The occurrence of four cases of trisomy 21 in four different members of one pedigree is highly unusual. While some families could show recurrence by chance alone, the number of cases in this family and the relatively young age of the mothers makes other explanations seem more likely. Possible explanations include structural defects or rearrangements of chromosome 21 or gene mutations leading to an increased risk of segregation errors at meiosis. Although the exact cause in this family remains unknown, several indirect lines of evidence suggest that some kind of cryptic rearrangement involving 21ql 1.1 could be present. The most intriguingfindingin this family was the presence of extra copies of three microsatellite loci on the p and proximal q arm of 1-2 and her daughter II-3 (2509J3TG, D21S369 and D21S215). Although D21S215, D21S369 and D21S258 are all in a repetitive region present in multiple copies on several chromosomes, extra copies of these markers have not been noted in large studies of Down syndrome families or normal controls (Laurent et al., 2003). However, cryptic translocations involving the centromeres of the acrocentric chromosomes have been described recently (Cockwell et al, 2003). As a G-banded karyotype in II-3 revealed a normal karyotype, the most likely explanation for the unusual marker typings would be a translocation between chromosome 21 (breakpoint 21ql 1.1 between D21S215 and D21S258) chromosome 13 (breakpoint near the centromere on the q arm). Segregation of markers on proximal chromosome 13 was consistent with this possibility. Such a translocation could explain a tendency for missegregation of chromosome 21 because it would result in pairing problems among the chromosomes involved.  117 Nonetheless, while a cryptic translocation might explain the increased risk for DS in II-3 and for 1-1 (as she might carry the same rearrangement) this does not seem to be the explanation for the other two occurrences of DS in this pedigree. The mothers of the other two DS individuals (II-4 and III-13) do not have molecular evidence for a duplication, nor do they even share the same maternally derived chromosome 21 centromere. The region of chromosome 21ql 1.1 including D21S258 is however potentially shared among all Down syndrome mothers and thus it is possible that there is an inversion or other rearrangement of this region that has both disrupted pairing and led to the further duplication/translocation in a previous generation. The high rate of meiotic recombination observed could have occurred as a compensatory mechanism for poor proximal chromosome 21 pairing. Chromosomal Abnormality: " a)  Cryptic translocation Translocations involving breaks at or very near the centromere may disrupt pairing  and lead to abnormal segregation. Cryptic translocations involving the centromeres of the acrocentric chromosomes have been described recently (Cockwell et al. 2003). FISH with probes for the centromere of chromosome 13/21 and 21q22, shown in Figure 5-2, were used to exclude translocation of the centromere of chromosome 21 in this family to another chromosome. A cryptic translocation between chromosome 13 and 21 cannot be excluded, and in fact is likely in two of the members of the pedigree (1-2 and II-3). This cannot explain the trisomy 21 child that we see for II-4 and III-13 as neither appears to carry the derivative chromosome (i.e. they do not have three alleles at the three markers across the centromere). II-5 and III-13 do not appear to carry the same centromere as the other women in the family. It is possible that the presence of the third chromosome 21  118 centromere has disrupted the centromeres of the other two 21 homologues found in 1-2, so that she passes on a disrupted chromosome 21 centromere to all of her children, regardless of which one they receive. All of her daughters also share a region just distal to the centromere (from D21S215 to D21S258) and this may have been either disrupted or inverted by some process relating to the original translocation. b)  Abnormality of centromere or telomere  The importance of both the centromeres and telomeres of chromosomes in pairing and segregation at meiosis makes them likely candidates for involvement in nondisjunction. The pericentromeric region of one chromosome 21 is shared between two of the mothers of a Down syndrome individual, as seen in Figure 5-5, but not a third. This region contains multiple repeats with shared homology to several other chromosomes, including chromosome 10 and 18 (Laurent et al, 2003). Paralogous copies of D21S215 and D21S369 have been found in databases within sequences assigned to chromosomes 13, 18 and 10, but experimental data could not confirm their presence on chromosome 13. This family has two members (1-2 and II-3) that show three alleles at three loci, 2503J9TG on the p arm and D21S215 and D21S369 on the q arm. We have not observed an unusual inheritance pattern suggestive of a duplication of the marker in any of our previous studies using D21S215 to determine origin of triploidy nor in any of controls. Furthermore, Laurent et al. (2003) reported that while these markers resided in a BAC that cross-hybridized to other chromosomes, there was no evidence of presence of this marker on another chromosome in any of their studied families. The telomere also plays a key role in the pairing and segregation of chromosomes during meiosis and so abnormality of this structure or region may be expected to be involved  119 in a predisposition to nondisjoin chromosomes. This region appears to be shared in all but one of the mothers in this pedigree but a mechanism of involvement for nondisjunction is not apparent. Telomere length, as measured by Southern, was normal in this group of women when compared to control women and no structural rearrangements were noted. c)  Small inversion  Although G-banding was repeated for three members of this family, a small inversion may have gone undetected. Recent reports indicate that the human genome contains many small polymorphic duplications, deletions and inversions when looked at on a fine-scale (Bailey et al, 2002; Giglio et al, 2002; Ji et al, 2000; Shaw and Lupski, 2004). Submicroscopic inversions have been implicated in mediating recurrent translocations (Giglio et al, 2002), recurrent deletions (Gimelli et al, 2003) and they can result in a reduction in recombination between two chromosomes (Koehler et al, 2004). A reduction in recombination near the centromere has been found to be associated with nondisjunction of chromosome 21 (Hassold and Sherman, 2000). However, increased recombination is often found outside of the postulated inversion. Specifically, a pericentric inversion could cause problems at pairing and may also result in abnormally high recombination outside of the inversion. M y data support this, as recombination is high outside of the region of the potential inversion. Genetic Predisposition: There are numerous examples in experimental organisms of mutations that increase rates of nondisjunction. While similar mutations have not been reported in humans, there is some evidence that common polymorphisms may increase risk of nondisjunction.  120 a)  MTHFR polymorphisms  Based on evidence that abnormal folate and methyl metabolism can lead to DNA hypomethylation and abnormal chromosomal segregation, mothers of a DS child have an increasedfrequencyof certain polymorphisms in methylenetetrahydrofolate reductase (MTHFR, C677T and A1298C) and methionine synthase (MTRR, A66G). Women with these particular polymorphisms were found to have a 2.6 to 2.9 fold increased risk of having a child with DS (James et al, 1999; Hobbs et al, 2000). Several subsequent studies could not demonstrate such an association with the MTHFR gene polymorphism and mothers with a DS child (O'Leary et al, 2002; Chadefaux-Vekemans et al, 2002; Bosco et al, 2003; Gueant et al, 2003). The susceptible TT genotype at C677T was found in four (80%) of the women in this family. This genotype has been reported in -9% of Native controls (Arbour et al, 2002), but higherfrequencieshave been found in Amerindian tribe studies (Herrmann et al, 2004). While this polymorphism alone cannot explain the susceptibility to DS in this family because one mother of a DS child did not have this genotype, it may contribute a small amount to the overall genetic risk. b) Mitochondrial Dysfunction  The possibility of a mitochondrial pattern of inheritance is consistent with the DS inheritance in the pedigree. This predisposition to nondisjunction for chromosome 21 is transmitted through females to females. Mitochondrial disfunction resulting from decreased ATPase6 and Tfam expression during meiotic maturation of oocytes has been suggested to cause nondisjunctional errors (Lee et al, 2003). Although this could not be tested, as new  121 blood samples would have been needed, it alone cannot account for the nondisjunction seen in this family. c) APOE s4 andpresinilin-1 polymorphisms  Many studies provide evidence for a shared genetic susceptibility to DS and Alzheimer disease (AD). Almost all individuals with DS above the age of 40 years have the neuropathological changes characteristic of AD (Rumble et al, 1989). Some forms of AD are caused by mutations in the APP (B-amyloid precursor protein) gene on chromosome 21, the presinilin-1 (PS-1) gene on chromosome 14 and the presinilin-2 (PS-2) gene on chromosome 1 (Petersen et al, 2000). The presinilin proteins may be important in chromosome segregation as they have been shown to localize to kinetochores and centrosomes (Li et al. 1997). Apolipoprotein E (apoE) is a gene located on chromosome 19. The rare allele s4 has been identified as a risk factor for early-onset and late-onset AD in both familial and sporadic cases (Strittmatter et al, 1993; Chartier-Harlin et al, 1994). Avramopoulos et al. (1996) demonstrated a higher frequency of the APOE s4 allele in young mothers of DS children due to maternal II errors. Peterson et al. (2000) demonstrated an increased frequency of allele 1 of the intron 8 polymorphism in PS-1 in mothers with a meiosis II error as well. Young mothers of DS children have been shown to have a fivefold greater risk than normal of developing AD later in life (Potter and Geller, 1996). This risk for developing AD may indicate either that they were mosaic for trisomy 21 or had an ongoing predisposition to chromosome missegregation that is reflected in their trisomy 21 offspring and their own increased risk for AD (Geller and Potter, 1999). If this association between apoE genotype and risk for having a DS child is correct, and is due to problems with segregation of chromosome 21,1 would then expect that the  122 young mothers in the pedigree reported here would either be germline mosaics or have a predisposition to nondisjunction in mitosis. However, having DS cases in multiple generations does not indicate mosaicism as the cause, in fact it argues against mosaicism. There are no members of the pedigree with recurrent trisomy 21 to a single couple. There does not appear to be an increase in aneuploidy in somatic cells of these women either, as measured in peripheral blood by FISH with probes for the centromere on chromosome 15, 21q22 and HPRT on the X chromosome. Although we have not directly looked at the APOE and presinilin genes, these polymorphisms are unlikely to be causing the trisomy 21 in the family for a number of reasons, not the least of which is the fact that there are no reported cases of AD in this family. Environmental: Many studies have attempted to correlate maternal smoking, caffeine and alcohol consumption, oral contraception use and other environmental factors with thefrequencyof having a DS pregnancy. The results of these studies are conflicting, some claiming decreased risk for a DS birth for smoking mothers (Kline et al, 1993), some showing no association (Torfs et al., 2000), and some showing increasedfrequencyof a DS child born to younger mothers. The latter study shows that increase in only a subset of meiotically-derived cases; specifically, mothers younger than 35 years with an error at meiosis II (Mil) (Yang et al., 1999). High alcohol and caffeinated coffee consumption has been associated with a reduced risk of for a DS conceptus (Torfs et al, 2000). Yang et al. (1999) found that oral contraceptive use in combination with maternal smoking increased the risk for a DS conceptus while oral contraceptive use alone was not a significant risk factor.  123 A recent study by Hunt et al. (2003) discussed the only environmental factor shown to have a very large affect on risk for aneuploidy in an animal study. An accidental exposure of control laboratory female mice to an environmental source of bisphenol A (BPA) resulted in a sudden increase in meiotic disturbance and aneuploidy, when looking at the oocytes of the exposed female mice. In two independent meiotic studies in two populations of mice, there were increases in abnormalities at meiosis seen. A sudden increase in congression failure, defined as a defect in the alignment of the chromosomes on thefirstmeiotic (MI) spindle, from 1-2% to 40% in mouse oocytes was observed. A significant increase in aneuploidy in control mouse oocytesfroma different studyfroma baseline of 0.5-1% to 5.8% was also observed. BPA is an estrogenic compound widely used in the production of polycarbonate plastics and epoxy resins, and it was released from the cages and water bottles after unintentional and subsequent intentional damage. Disruption of meiosis on this scale due to environmental factors would most likely result in increased aneuploidy for chromosomes other than 21, and as well, pregnancy losses and increased miscarriage within the pedigree would be expected. As this was not reported in this family, a structural or epigenetic alteration on chromosome 21 seems more likely. Conclusion: Although I was not able to demonstrate the exact cause, this pedigree provides support for the possibility that trisomy 21riskvaries among individuals and can be genetically influenced. While trisomy mosaicism has been assumed to be the most likely explanation for an increasedriskof trisomy 21 in some couples (Warburton et al., 2004) this may not be the only explanation. This family appears to be segregating for a cryptic chromosome 13;21 translocation, however this does not entirely explain the susceptibility to  124 D o w n syndrome i n this family. For this reason, I cannot predict who is and is not at risk. Currently maternal serum screening is being offered to all family members, and amniocentesis w i l l also be offered to pregnant family members i f they wish, given the extensive family history.  125  Chapter 6: Telomere length in IVF/ICSI cases  2  •  Dr. Sai Ma ascertained the patients in this chapter and DNA samples were supplied to me blinded for analysis.  •  Luana Avila and myself performed the Southern analyses and I analyzed all of the data.  AIM: The aim of this portion of the project is to determine if telomere length is decreased in individuals conceived as a result of intracytoplasmic sperm injection (ICSI). This tests the hypothesis that such individuals will have shorter telomeres as a result of the procedures by which they are conceived and may be at increased risk of cancer and other conditions of aging.  A version of this chapter has been submitted and accepted for publication. Robinson, WP, Penaherrera, M S , Gair, J, Hatakeyama, C and S Ma. (2005) X-Chromosome Inactivation and Telomere Size In Newborns Resulting From Intracytoplasmic Sperm Injection. A J M A 2  126  6.1  INTRODUCTION: Intracytoplasmic sperm injection (ICSI) is an assisted fertilization procedure used i n  couples with severe male factor infertility, and who can not be helped by conventional in vitro fertilization (IVF). A slight increase in de novo chromosomal abnormalities for ICSI children has been reported, while the major congenital malformation rate is similar for I V F and ICSI (between 3 and 4%) (Devroey and V a n Steirteghem, 2004). It has been suggested that the use o f ICSI or I V F could increase the risk for imprinting disorders considering that the mammalian embryo, cultured in vitro, is susceptible to changes i n imprinting control (Young et al., 2001). Increasing attention has been focused on potential epigenetic disturbances resulting from embryo culture, somatic cell nuclear cloning and assisted reproductive technology (De Rycke et al, 2002; Gosden et al, 2003). Recent publications have noted that a few children are affected by diseases caused by imprinting disorders, such as Angelman syndrome, as a result o f ICSI (Cox et al, 2002). Although the exact mechanisms is unknown, it has been proposed that ICSI disrupts the production or function o f transacting factors necessary for imprinting o f the maternal chromosome 15 (De Rycke et al, 2002). A s well, a total o f seven children have been reported with Beckwith-Wiedemann syndrome ( B W S ) , which is a human overgrowth syndrome, after assisted reproductive technology. Four were born after ICSI with ejaculated sperm and one after ICSI with testicular sperm (DeBaun et al, 2003). Since there is an association between B W S , cancer and H I 9 methylation abnormalities, assisted reproductive technology could be associated with embryonal cancers o f childhood (Devroey and V a n Steirteghem, 2004). There is evidence for an increased risk o f imprinting disorders i n ICSI children and childhood cancers may also be associated with the ICSI procedure. It has been  127 proposed that ICSI may interfere with the establishment of the maternal imprint in the oocyte or pre-embryo, as AS and BWS in ICSI patients are both caused by methylation defects or loss of methylation on the maternal allele (Cox et al, 2002; Devroey and Van Steirteghem, 2004). The mammalian genome undergoes widespread epigenetic reprogramming in germ cells and the early embryo. The reprogramming process includes chromatin remodeling, DNA methylation, genomic imprinting, and maintenance of telomere lengths, histone modifications, epigenetic inheritance and X chromosome inactivation (Schaetzlein et al, 2004; Jeon et al, 2005). A number of studies have shown that telomere length is restored to normal length in cloned mice and cattle embryos derived fromfibroblasts(Scaetzlein et al, 2004). However, cloned sheep derived from epithelial cells had shorter telomeres than controls (Shiels et al, 1999). Cloned calves derivedfrompopulations of senescent donor somatic cells were found to have telomeres whose length was extended beyond those of newborn and age-matched controls (Lanza et al, 2000). I was interested to investigate whether or not the resetting of telomere length might also be disrupted during an assisted reproductive procedure such as ICSI, resulting in children with telomeres that are shorter or longer than normal for their age. In order to determine if telomere length may be affected by IVF and ICSI such that newborns resulting from these procedures maybe at an increased risk for conditions of aging and/or cancer, we measured and compared telomere length between this group of children and those conceived naturally.  128 Methods and Materials Cord blood was obtained, after informed consent, from term placentas in a series of 43 newborns conceived after ICSI. Cord blood was extracted from 16 term placentas from healthy volunteer donors with prior informed consent orfromblinded samples from anonymous healthy deliveries, after UBC ethics approval (C98-0315). Genomic DNA was isolated using standard techniques. An estimation of average telomere length was obtained using the Ze/oTAGGG Telomere Length Assay kit from Roche Diagnostics (Montreal) as detailed in Chapter 2. This assay is based on a standard method to assess telomere length that utilizes Southern analysis of the terminal restrictionfragment(TRF) that is obtained by digestion of 0.5 ug of genomic DNA usingfrequentlycutting restriction enzymes, Hinfi and Rsal. The TRF includes the (TTAGGG)n repeat as well as a portion of the subtelomeric region or telomere associated repeats (TAR). Telomere length was quantified by measuring chemiluminescence using a Bio-Rad Fluor-S Multilmager, and analysis with Quantity One software. This method gives a kilobase value that quantifies the size of the average TRF in each DNA sample. To assess the reproducibility of the telomere length assay, four samples were analyzed two times on separate Southern blots. They were well correlated (r=0.96) and a one-way ANOVA for correlated samples determined that there was no significant difference between the two groups of measurements (F=2.63; p=0.20). 6.2  RESULTS:  Telomere size was measured in the cord blood from 43 ICSI cases and compared to newborn cord blood from 16 control cases since telomere length declines rapidly in the first year of life. Figure 6-1 shows the distribution of telomere length values for the 43 ICSI  129 samples and 16 control samples and Table 6-1 shows the data for each sample. Average telomere length varied considerably in each group, with a mean telomere length of 12.3 (±1.4, range 7.2-21.2) in the ICSI samples and 11.56 (±1.2, range 7.7-15.2) in the controls. There were several very large telomere lengths in the ICSI group. However, there is less accuracy in length designation for large sized alleles due to the shorter migration of fragments in that size range and thus this is not outside the expectation for 'normal'. 25  • 20 t  I  I•  I • 5  0  J  1  ICSI (N=43)  ,  Control (N=l6)  Figure 6-1: Distribution of telomere lengths, measured by Southern blot analysis, in the ICSI newborns (N=43) compared to newborn controls (N=16).  130  Table 6-1: Telomere lengths measured by Southern blot. The average telomere length for the ICSI group is not significantly different than the control group (p=n.s.; t-test). ICSI cordblood samples Sample name Telomere Length (kb) SM1 7.3 SM2 10.2 SM3 11.7 SM4 11.2 SM6 11.3 SM7 11 SM8 8.8 SM9 8.1 SM10 9.7 SM11 8 SM12 7.2 SM13 7.6 SMCB22 9.73 SMCB23 10.1 SMCB24 21.17 SMCB25 11.6 SMCB26 12.2 SMCB27 15.6 SMCB28 13.2 SMCB29 17.5 SMCB30 13.6 SMCB31 12.5 SMCB32 11.5 SMCB33 15.8 SM04-77 15.99 15.1 SM04-75 8.9 SM04-70 12.61 SM04-69 16.4 SM03-28 13.44 SM02-30 13.4 SM02-29 12.26 SM02-19 15.84 SM02-14 SM02-11 11.16 11.96 SM02-09 10.74 SM02-26 9.6 SM02-27 12.99 SM03-30 7.85 SM04-29 17.92 SM04-76 21.2 SM04-54 14.33 SM04-60 11.03 SM04-74 Average 12.3  Control newborn cordblood samples Sample name Telomere Length (kb) PX3 12.6 PX4 9 14.4 PX6 11 PX7 12.9 PX14 12 PX16 12.2 PX17 9.7 PX18 12.8 PX19 PX21 8.9 9.3 PX22 7.7 PX23 15.2 PY1 14.6 PY3 13.2 PY4 9.4 PY7  11.56  131 6.3  DISCUSSION:  While telomere length is generally reestablished correctly, some cloned organisms have been found to have either shorter or longer telomere length then expected (Shi et al. 2003). This may have been due to culture conditions affecting telomerase activity in the early developing embryo. Decreased telomere length is associated with decreased replicative potential (cell life span) and increased cancer risk. To our knowledge, there have not been studies of telomere length in IVF or ICSI conceived humans. Therefore, while not surprising, it is reassuring to find the telomere length is comparable in ICSI conceived infants to that for in vivo conceived infants. Much remains to be understood about the etiology of slightly elevated risk for birth defects and increase in rare imprinting disorders after IVF and ICSI. A concern is that epigenetic mutations may be common among rVF/ICSI conceived infants, but only rarely have an effect apparent at birth. Alternatively, the observed cases of Beckwith-Wiedemann syndrome and Angelman syndrome may be reflective of a susceptibility particular to a very small number of cases and not the IVF/ICSI group as a whole. Follow-up studies looking at neurodevelopment after ICSI/rVF have not shown any differences with each other or as compared to their naturally conceived peers (Bonduelle et al. 2003; Sutcliffe et al. 2003), thus suggesting gross changes in gene expression are unlikely to be common. Our present results are also consistent with normal early blastocyst development in the majority of viable embryos conceived by ICSI. However, epigenetic changes involved in imprinting have not been examined and larger studies are needed in order to look at this in more detail.  132 Chapter 7: Recombination at the common deletion breakpoints and nondisjunction of chromosome 15  •  The samples for this chapter were donated by Dr. Stephen Wood, a former CEPH collaborator. The samples from UPD15 and T15 mothers were obtained by Dr. Wendy Robinson from various collaborators through investigations of Prader-Willi syndrome patients and through cytogenetic analysis of spontaneous abortions.  •  Many members of the Robinson lab including Brian Kuchinka and Maria Penaherrera did much of the analysis previous to my involvement.  •  I continued and expanded on the data set and performed much of the microsatellite genotyping and data analysis.  A I M : The aim of this portion of the project is to determine if recombination is increased at the common deletion breakpoints on chromosome 15. This tests the hypothesis that deletions may result from an increase in recombination, making these areas prone to breaks. I was also interested in what the role of recombination is in nondisjunction. A comparison of normal CEPH female controls with meiosis I and II error females is also included.  7.1  INTRODUCTION: Recombination at meiosis is essential for ensuring normal segregation of  chromosomes into the gametes. While on average there is ~ l c M of recombination per 1Mb of DNA in the human genome, this varies considerably by sex and region of the chromosome. Recombination also appears to occur in "hotspots", regions of high recombination separated by regions of lower recombination. In a previous study in our laboratory, sex-specific recombination hotspots were identified within the imprinted region of chromosome 15ql l-ql3 (Robinson and Lalande, 1995). This region of chromosome 15 was also found to show a reduction in recombination in association with nondisjunction (Robinson et al., 1993 and 1998). In the present chapter, I expand the study of recombination in this region to answer two questions 1) Are recombination hotspots also important in the etiology of recurrent chromosomal rearrangements? 2) Do strong recombination hotspots represent sites critical for chromosome pairing and segregation, or is recombination anywhere in the centromeric portion of the chromosome sufficient to ensure normal chromosome segregation? Recurrins 15qll-ql3  rearrangements  The recurring 4 Mb constitutional deletion at 15ql l-ql3 that gives rise to either Prader-Willi syndrome (PWS) or Angelman syndrome (AS), depending on whether the deletion is paternal or maternal in origin, respectively, is one of the most common deletions that is observed in humans. The overallfrequencyof the deletion is -1/15,000 (Christian et al. 1995). Duplications, triplications and tetrasomy of the 15ql l-ql4 region have also been reported with varying degrees of clinical manifestation, and inv dup(15) accounts for -50% of the small supernumerary marker chromosomes observed in humans (Webb, 1994). Based  134 on family data it was concluded that deletions can result from both inter- and intrachromosomal recombination (Carrozzo et al. 1997; Robinson et al. 1998). Certain regions of the genome, such as 17pl 1.2,15qll-ql4 and22qll.2, seem to be particularly prone to deletions or duplications. While low copy repeat sequences have been found around the breakpoints in each of these regions (Shaw et al., 2002; Pujana et al, 2002; Edelmann et al., 1999), the genome is made up of ~50% repetitive sequences (http://www.ornl.gov/hgmis/project/febj)r/summary_of_sequence.html).  Thus, while the presence of  repeat sequences at the breakpoint sites may be a prerequisite for regional instability, it may not be the only important factor. Clearly the involved repeats must also be at least somewhat recombinogenic - that is double strand breaks must occur in meiosis or mitosis. However, it is not yet clear whether breakpoints associated with such regions are "hotspots" of recombination or whether simply a low level of recombination combined with frequent mispairing of the region is required. To examine this question we have looked in detail at recombination in the proximal 15ql l-ql4 region, which is associated with frequent deletion, tandem duplication, tandem triplication and small supernumerary inversion duplication chromosomes. The recurrent rearrangements of the 15ql l-ql4 region involve six common "hotspots" for breakage. There are three common deletion breakpoints (BP) associated with PWS and AS; two proximal breakpoints (BP1 and BP2) and one distal breakpoint (BP5). Consistent breakpoints are also associated with the inv dup(15) marker chromosome (Christian et al, 1999) and interstitial duplications (Thomas et al, 1999). Small inv dup(15) chromosomes involve the same two proximal breakpoints as the deletions (BP1 and BP2) plus an additional breakpoint BP3, while the larger inv dup(15) chromosomes use BP4 and  135 BP5 located between D15S24 and D15S144 (Pujana et al. 2002) (Figure 7-1), and a 6th breakpoint located distal to D15S144 (Pujana et al. 2002). It has been shown that at each of the six described 15ql l-ql4 BPs, there are clusters containing several copies of the human chromosome 15 low-copy repeat (LCR15) duplicon. The LCR15s have sequence elements in common with the #E7?C2-containing (Hect domain and RCcl domain protein 2) duplicons  (Pujana et al. 2002). Approximately 11 copies of the HERC2 gene have been identified on chromosomes 15 and 16, but most of them are within 15ql l-ql3 (Ji et al. 2000). These duplicons may result in high non-homologous recombination resulting in deletion, duplication and other chromosomal rearrangements (Amos-Landgraf et al, 1999). A diagram of this region of chromosome 15 is shown in Figure 7-1. Early analysis (Robinson et al, 1993) suggested that the deletion breakpoints on chromosome 15 were not sites of high homologous recombination. However, data for the proximal region was minimal, as there were no markers centromeric to BP1 at that time and physical data were limited and of low accuracy. Recent advances in the sequencing of the human genome allow the ability to more accurately measure the level of recombination/physical distance over different regions of the genome.  U9JU30 Supuuduii=3i *>[UTd in paxoq si suaS passaidxa A.\\ems\Bva am pire arqq ui psxoq SJB sauaS psssaidxs AireuiaiBj •iiM.oqs are ZQOZ 7° ^ Birefnj TXIOJJ suooijdnp Surareruoo-^y^/-/ pire sgi-gjyr moq jo suoiyeocq *UM.oqs SIB smrod^ajq xis ny ^ i b - i tbgi ajngjj 3  I HERC2  738CA  "0  D15S18  D15S1035 D15S541/542  D15S156  HERC2  D15S217 D15S12  m .  3  HERC2 D15S24  m D15S1019  2  D15S1043  HERC2  I  9£l  D15S1031 D15S1010 D15S144  5  137 Nondisjunction of chromosome 15  Based on data from nondisjunction of chromosomes 16 and 21, it has been suggested that a chiasma located near the centromere is of particular importance in protecting against nondisjunction (Hassold et al, 1995; Lamb et al, 1997; Sherman et al, 1994). However chromosome 18 does not show altered recombination in the proximal q arm (Bugge et al, 1998). While nondisjunction of chromosome 15 at meiosis I is associated with reduced recombination overall, the effect is modest and only the region very close to the centromere (15ql l-ql3) showed a striking reduction (Robinson et al, 1998). Specifically, a 50% reduction (15.3 to 7.7 cM) in the genetic map was observed in the most centromeric section of chromosome 15 in analysis of the products from meiosis I errors compared to segregations from CEPH normal controls. It is possible that it is not simply distance from the centromere that is important, but rather the presence/absence of recombination at specific 'pairing' sites. Sex-specific meiotic recombination hotspots have been reported in proximal chromosome 15 (Robinson and Lalande, 1995). I suggest that strong female specific hotspots maybe important sites of pairing in female meiosis and that it is a reduction of recombination at specific sites, and not necessarily any site near the centromere, which is important for increasing risk for abnormal chromosome segregation. Methods and Materials Samples CEPH DNA donated from Dr. Stephen Wood was used to determine the amount of recombination in 15ql l-ql3 in both males and females, and specifically at the common deletion breakpoints associated with the ~4 Mb Prader-Willi/Angelman syndrome deletion.  138 The uniparental disomy 15 (UPD 15) samples were obtained through investigations of Prader-Willi syndrome patients and the trisomy 15 (T15) samples were ascertained through cytogenetic analysis of spontaneous abortions (Robinson et al., 1998). A total of 137 UPD 15 mothers and 20 T l 5 mothers were used in these analyses. Of these, 115 of the UPD15 samples have been analyzed previously (Robinson et al., 1998) and 13 of the T15 samples have been analyzed (Robinson et al, 1998; Zaragoza et al, 1994). Therefore analyses of 7 new T15 and 22 new UPD 15 samples are included here. Estimation of Recombination Frequencies Controls (CEPH):  Using microsatellite markers from Research Genetics Inc. (Huntsville, AL), recombination was analyzed along chromosome 15 in both males and females. Markers spanning 15ql 1 -ql 3 were used and are listed in Table 7-1. A schematic of this region of chromosome 15 is shown in Figure 7-1. Regions of crossing over were determined for CEPH family control members, by looking at online data at a parent informative at a certain marker (i.e. they are heterozygous at that locus) and the same chromosome 15 in two children of that parent, one that has a crossover and one that does not. Haplotypes were constructed for all family members from the online data. Phase was determined from grandparents, meaning that the parental haplotypes were known. Markers on either side of the crossover are typed and analyzed and determined to either be the same or different. Specifically, the two siblings are compared and if they have the same allelefromthat parent, then they are given an "s", if they have a different allelefromthat parent, then they are given a "d". Multiple microsatellite markers are typed on either side of the crossover until the point at which the crossover happens can be  139 determined. This is the point where a string of "s"s change to a string of "d"s. Refinement of the crossover location is dependent on marker availability and the informativeness of the parent at each marker. If a parent is homozygous at a marker, then it cannot be used to narrow down the crossover. We assume double crossovers are not missed as two crossovers are unlikely to occur over short distances, due to chiasma interference (Robinson and Lalande, 1995). Using these data, the amount of recombination in a given area was determined and an estimate for genetic distance in cM made. The region from 15ql l-ql3 was divided into 11 intervals and genetic distance was calculated within each interval (Figure 7-2). Intervals were determined based on previous estimates from published CEPH data, so that each interval was roughly 1Mb, and contained at least one informative marker. As well, the intervals were chosen to coincide with the six common deletion and duplication breakpoints described in Pujana et al. (2002). The number of crossovers in a region divided by the total number of informative meioses generates a %, which can be transformed into a cM value (1 cM = 1 % recombination). The number of informative meioses is determined by looking at the original haplotype data and counting the number of individuals in each family that are informative for that interval. Adding up all of the individuals for all of the families with data for that interval gives the total number of informative meioses. Occasionally some markers were uninformative but the haplotype data on either side of the marker suggested that the crossover was not in that region. The number of crossovers that occur in a given interval must be estimated, as some crossovers cannot be localized to a specific interval (e.g. when markers are uninformative). By calculating the probability that a crossover will occur in one interval rather than the neighbouring one, a genetic distance can be determined.  140  CM •<*  < o  co  in co in  oo 1  2  3  CO  in co m  co in  CO m  4  5  co co CO m  in  CM  CM  CO m  7  6  CM CM  co m  co m  CO  in  8  9  CB CO in  11  10  in CD  co m  CO  m a: m <  Figure 7-2: 15ql l-ql3 divided into 11 intervals in order to analyze recombination patterns.  Mothers of trisomy 15 and UPD 15:  The same intervals and microsatellite markers as used above were used to analyze recombination in meiosis I or II errors. Collection of data is essentially the same as for control data, except that occasionally only mothers' and patients' samples are used in the PCR and run on the gel. As these are all maternal meiotic errors, the paternal genotype is not necessary as we were only interested in recombination along the maternal chromosome. When analyzing the bands on the gel, the patients are assigned either an N or an R, meaning nonreduced and reduced respectively. In order to be informative, the mothers must be heterozygous for the marker typed. N is assigned if the patient is also heterozygous here, and R is assigned if the marker is reduced to homozygosity in the patient. If the mother is  141 homozygous then the marker is uninformative for this pair. Chromosomes that are " N " as close to the centromere as is possible to type are considered MI errors while an "R" near the centromere is considered an M i l error. Where there is a transition from N to R or vice versa, a crossover is inferred to have taken place. As only half of the recombination events will be observed when the proximal marker is N ; R to N transitions are divided by 2, as all of these will be observed while only half of the N to R transitions will be (Robinson et al., 1993). 7.2  RESULTS:  CEPH (control) genetic map of 15qll-ql3: Based on the present data, the genetic distance from 738CA to D15S144, which spans the 15ql l-ql4 rearrangement region, is estimated to be 32 cM in CEPH females and 27 cM in CEPH males (Table 7-1). The physical size for this portion of chromosome 15 is estimated to be less then 14 Mb. After comparing the observed amount of recombination across intervals 1-9(10 and 11 could not be included as there are too few total meioses to group with the others), there is a significantly increased amount of recombination in females (p=0.03; x ), but not in males (18.65 cM and 13.54 cM, respectively, compared to 7.3 cM 2  expected based on a physical estimate of 7.3 Mb). From the human May 2004 assembly (www.genome.ucsc.edu), D15S541/542 to D15S144 includes ~11Mb. Although the exact physical distance from the centromere to D15S541/542 is unknown, due to a gap in this region of the sequence, it is estimated cytogenetically to be much less than a chromosome band (<5 Mb) (Christian et al. 1995) and the UCSC genome browser shows a satellite sequence block at ~18.2 MB (with the region proximal to this being unmapped). We have used a marker (738CA), a partial duplication of the GABRA5 gene, that lies outside the large  142 region most frequently deleted in AS and PWS individuals (Ritchie et al, 1998). Although it's exact location in unknown, it lies between 18.6 Mb, which is the first gene on the UCSC map and 20.46 Mb, which is the first sequence clearly in the PWS/AS deletion region. Figure 7-3 a shows the crossover data for CEPH males and females, from which the genetic maps were estimated. The genetic and physical map data for this region of chromosome 15 is shown in Table 7-1. Both the male and female CEPH meioses appear to follow a similar pattern of recombination along the length of this region, which can be seen in Figure 7-4. As seen in Figure 7-4, there is a discrepancy between the sexes in intervals 1, 8 and 9, with interval 1 showing a significant increase in recombination in CEPH females compared to CEPH males (p=0.02; Fisher's exact). There was a female recombination hotspot within interval 8 and a male recombination hotspot within interval 9 reported previously that these data still support (Robinson and Lalande, 1995) (Figure 7-5). The other male hotspot reported previously within D15S541-D15S11 (intervals 2 and 3) appears to be a region of high homologous recombination in both sexes, although all confirmed crossovers in this region are distal to D15S543. A paternal crossover previously reported (Robinson and Lalande, 1995) was a genotyping error that was corrected in this data set.  143  Table 7-1: The most recent physical and genetic map data for 15ql l - q l 3 . The C E P H (control) male and female genetic maps of this region (738CA to D15S144) are estimated to be 27 c M and 32 c M respectively. STS  band Locus cent q10 738CA q11.2 AFM344TE5 D15S1035 q11.2 D15S541 D15S541 q11.2 D15S542 D15S542 q11.2 D15S18 D15S18 q11.2 D15S543 D15S543 q11.2 q11.2 D15S11 D15S11 ATA10C01 D15S646 q11.2 CHLC.GATA81F03 D15S817 q11.2 D15S673 D15S673 q11.2 AFMB344WC5 D15S1021 q11.2 AFM273YF9 D15S128 q11.2 AFM200WB4 D15S122 q12 AFM320vd9 D15S210 q12 AFMa284za9 D15S986 q12 AFMa312yg5 q12 D15S989 D15S113 q12 D15S113 Hs.302352 GABRB3 q12 D15S97 D15S97 q12 CHLC.GATA88H02 D15S822 q12 AFMA216ZC9 D15S975 q12 Mfd209 D15S219 q12 AFM214xg11 q12 D15S156 GATA8B06 D15S217 q12 AFMb336yf1 D15S1019 q13.1 AFMa081xb9 D15S1048 q13.1 AFM248vc5 D15S165 q13.3 AFM019tf6 D15S144 q13.3  Physical Map bp (x1,000,000) 17-18 18.6-20.46 20.46 20.459 20.46 20.477 21.183 21.637 21.9 22.151 22.534 22.567 22.682 23.231 23.3 23.569 23.573 23.762 24.343 24.379 24.973 25.059 25.508 25.58 25.677 27.38 27.585 29.048 31.388  Christian paper female (cM) NA  12 12  My data Genethon - Marshfield - My data female (cM) female (cM) male (cM) female (cM) NA NA NA NA 0 0 0 0 8.4 2.6 8.4 2.6  14.7 14.7  2.6 4.8  8.4 11.5  5.9  12.1  6.9 7.4 7.9  13.3 13.3 13.3  10.5 10.5  18.2 18.2  13.6  18.6  21.7 27.3  25.8 32.3  6.11 4.78 5.4 6.1 6.3 6.2 7.2 7.2 17 17  4.78 6.11 6.11 6.11 6.9 6.92  14.4  9.9 9.9 12.3 13.1 14.6 14.6  19 19.1 20.2 25.3  19.1 19.1 20.2 25.3  11.8  144  Maternal crossover event Paternal crossover event  y  4  4  •4  <  -»  Interval 1  Interval 2  Interval 3  Interval 4  Interval 9  Interval 10  Interval 11  N=181 (mat)  N=199 (mat)  N=199(mat)  N=222(ma ) N=222 (mat) N=223 (mat)  N=225 (mat) N=225 (mat)  N=247 (mat)  N=92 (mat)  N=92 (mat)  N=189 (pat)  N=205 (pat)  N=205 (pat)  N=210 (pat  N=223(pat)  N=246 (pat)  N=92 (pat)  N=92 (pat)  Interval 5  Interval 6  N=210 (pat) N=211 (pat)  Interval 7  Interval 8  N=223 (pat)  Figure 7-3a: Individuals from C E P H families were typed for microsatellite markers and recombinants were identified. Observed crossover events are indicated, where each coloured line represents the region within which the crossover was localized in an individual meiosis. Maternal crossovers (of N=246 total meioses) are indicated in pink and paternal crossovers (of N=247 total meioses) are indicated in blue.  145  — Ml error crossover event •—• Mil error crossover event  11 I Interval 1  Interval 2  i i iii I I • 1 11 i  Interval 3  Interval 4  Interval 6  Interval 7  Interval 8  Interval 9  Interval 10  N=124 (Ml){ N=125 (Ml)  N=126 (Ml)  N-=127 (Ml): N=127 (Ml)  N=127 (Ml)  N=127 (Ml)  N=127 (Ml)  N=128(MI)  N=132(MI)  N=132(MI)  N=22 (Mil)  N=25 (Mil)  N=25(MII)  N=25 (Mil)  N=25 (Mil)  N=25 (Mil)  N=25 (Mil)  N=25 (Mil)  N=25 (Mil)  N=23 (Mil)  Interval 5  N=25 (Mil)  Interval 11  Figure 7-3b: Individuals from UPD15 and T15 families as a result of an MI (of N=132 total meioses) or an Mil (of N=25 total meioses) error in meiosis were typed for microsatellite markers and recombinants were identified. Observed crossover events are indicated, where each line represents the region within which the crossover was localized in an individual meiosis.  146  14  S  — CffH males - « — CEPH females  •o o E  -A— Ml errors -X— Mil errors  Figure 7-4: The estimated c M values for C E P H males and females (controls) compared to MI and M i l error chromosomes. A significant difference is seen in interval 1 with females having more recombination than males (p=0.02; Fisher's exact). No other significant differences among groups were seen.  147  10 9 8  Intervals  Figure 7-5: Chromosome 15ql l-ql3 divided into 11 intervals from centromere on the left to microsatellite marker D15S144 on the right. CEPH control male and female cM/Mb data is shown. Note the sex-specific recombination hotspots, in particular, the two female hotspots of recombination in intervals 1 and 8 and the male recombination hotspot in interval 9. Intervals 3 and 5 appear to regions of high homologous recombination in both sexes. The genetic distance from 738CA to D15S541/542 (BP1 and interval 1) is estimated to be 8.4 cM in females and 2.6 cM in males. A l l crossovers that were observed in the region containing the proximal deletion breakpoints (intervals 1 and 2) appear to localize to BP1: no crossovers were localized to BP2 or interval 2 (see Figures 7-3 , 7-4 and 7-5). The physical distance from 738CA to D15S541/542 is estimated to be <2.2 Mb. Interval 1 is a hotspot of recombination for CEPH females as there is an increased amount of recombination although the increase is not significant (p=n.s.; x ), and recombination is also increased over that seen 2  in CEPH males (p=0.02; Fisher's exact). There is not an increased amount of recombination in males in this interval (p=n.s.; x ). 2  Interval 10 (D15S156 to D15S165) contains the most common distal breakpoint found in PWS and AS patients (BP3) as well as BP4, which is involved in larger inv dup(15)  148 marker chromosomes. This region is ~ 3.4 Mb in size, and the genetic map is estimated at 7.2 cM and 8.1 cM in CEPH females and males respectively. There is therefore, an increase of recombination in both females and males in this region, however, this increase is not significant (p=n.s.; x , for both females and males). 2  Interval 11 contains the distal breakpoint (BP5) that is often involved in large inv dup(15) marker chromosomes. The genetic map for this region is 6.5 cM and 5.7 cM in females and males respectively. This interval is only estimated to be -2.4 Mb, again showing an increased amount of recombination compared to the genomic average, however it is not significant (p=n.s.; x , for both females and males). 2  Genetic map for UPD15/T15 mothers:  Comparisons of recombination in 15ql l-ql3 between CEPH (control) females and females experiencing either meiosis I or meiosis II errors leading to UPD 15 or trisomy 15 reveal differences in the patterns observed. The genetic map for mothers having either a child with UPD 15 or T15 resulting from a meiosis I error (MI) from D15S541/542 to D15S144 is estimated to be 13.8 cM, while the same region was found to have a genetic map of 36 cM in mothers who had a meiosis II (Mil) error (Table 7-2). This region is estimated to be ~11.8Mb, and found to be 24cM in CEPH control females. There is not a significant decrease in recombination in this centromeric region in the MI error group of women (p= n.s.; x ) while there appears to be an increase in the amount of recombination in the M i l error 2  group (Figure 7-6) when comparing to the expected amount based on the genomic average. There is a significant decrease in recombination in the MI error group of women when comparing to the CEPH control female genetic map of the same region (13.8 cM compared to 24 cM, respectively - p<0.05; x ). There is also a significant increase in recombination in 2  149 the M i l group in this region when compared to CEPH female controls (36 cM compared to 24 cM, respectively - p<0.025; x ). Figure 7-3b shows the crossover data for the MI and M i l 2  error groups, from which the genetic maps were estimated.  Table 7-2: Genetic maps of CEPH male and female controls compared to the MI and M i l error groups of mothers. The total genetic map length for MI error mothers was estimated to be 13.8 cM while the genetic map for M i l error mothers was 36 cM. STS  Locus cent 738CA AFM344TE5 D15S1035 D15S541 D15S541 D15S542 D15S542 D15S18 D15S18 D15S543 D15S543 D15S11 D15S11 ATA10C01 D15S646 CHLC.GATA81 F03 D15S817 D15S673 D15S673 AFMB344WC5 D15S1021 AFM273YF9 D15S128 AFM200WB4 D15S122 AFM320vd9 D15S210 AFMa284za9 D15S986 AFMa312yg5 D15S989 D15S113 D15S113 Hs.302352 GABRB3 D15S97 D15S97 CHLC.GATA88H02 D15S822 AFMA216ZC9 D15S975 Mfd209 D15S219 AFM214xg11 D15S156 GATA8B06 D15S217 AFMb336yf1 D15S1019 AFMa081xb9 D15S1048 AFM248vc5 D15S165 AFM019tf6 D15S144  band q10 q11.2 q11.2 q11.2 q11.2 q11.2 q11.2 q11.2 q11.2 q11.2 q11.2 q11.2 q11.2 q12 q12 q12 q12 q12 q12 q12 q12 q12 q12 q12 q12 q13.1 q13.1 q13.3 q13.3  Mil error Ml error Physical Map bp My data - My data (x1,000,000) male (cM) female (cM) data (cM) data (cM) NA NA NA NA 17-18 0 18.6-20.46 0 20.46 8.4 2.6 20.459 8.4 20.46 2.6 20.477 0.34 8.4 0.43 21.183 2.6 6.34 11.5 0.85 21.637 4.8 21.9 22.151 7.5 0.85 22.534 5.9 12.1 22.567 8.66 0.85 13.3 22.682 6.9 13.56 0.87 7.4 13.3 23.231 14.45 0.89 7.9 13.3 23.3 23.569 23.573 23.762 15.34 18.2 0.91 10.5 24.343 15.34 18.2 0.91 24.379 10.5 24.973 25.059 25.508 16.04 5.65 18.6 25.58 13.6 25.677 27.38 27.585 22.68 6.77 21.7 25.8 29.048 35.98 13.79 32.3 31.388 27.3  150  14 12 10  s -A— Ml errors •a  -x— Mil errors  CS  - 0 — CFJH females  E  •a  10  11  Figure 7-6: Chromosome 15ql l-ql3 divided into 11 intervals from centromere on the left to microsatellite marker D15S144 on the right. MI error mothers and M i l error mothers cM/Mb data is shown compared to CEPH control female data. There is not a significant decrease in recombination in general in this centromeric region for the MI error mothers (p =n.s.; x ), while there is a significant increase in recombination for the M i l error mothers (p<0.001; x ). 2  2  7.3 DISCUSSION: Recombination and rearrangements:  There is increasing evidence for the involvement of repetitive DNA sequences as facilitators of some of the recurrent chromosomal rearrangements observed in humans. The high densities of repetitive elements, such as Alu elements, at some chromosomal translocation breakpoint regions has led to the suggestion that these sequences could provide hot spots for homologous recombination, and could mediate the translocation process (Kolomietz et al, 2002). Illegitimate recombination events between low-copy repetitive elements (LCR) have been implicated in the pathogenesis of various chromosomal rearrangements (Martinez-Garay et al, 2002).  151 It has been postulated that the high rate of chromosome rearrangements that occur in 15ql l-ql3 may be due to repetitive elements resulting in an excess of recombination in the region, leading to double strand breaks. The hypothesis that genomic instability in 15ql 1ql3 might be due to repeated DNA elements in this region was first proposed by Donlon et al. (1986). Low copy repeats (LCRs) flanking deletion breakpoints have been identified in other deletion syndromes, such as Smith-Magenis syndrome (SMS), Williams-Beuren syndrome (WBS), DiGeorge/velocardiofacial syndromes (DGS/VCFS) and neurofibromatosis type 1 (NF1) (Shaw et al., 2002). Such LCRs have been found to flank the common 15ql l-ql3 deletion (Christian et al, 1995; Amos-Landgraf et al, 1999; Pujana et al, 2002). SMS is consistent with a "reduced recombination/increased unequal crossing-over" hypothesis, as recombination is reduced in the SMS region. This hypothesis originally came fromfindingsin Charcot-Marie Tooth disease type 1A (CMT1 A), where male patients exhibit a lower recombination frequency in the CMT1A region than do female patients, and yet the duplication is found to occur 10 times more frequently in males (Shaw et al, 2002). The theory is that reduced recombination results in an extended region of unsynapsed chromosome segments in meiosis, allowing the chromosomes to slip and predispose to unequal crossovers. Unequal meiotic crossovers have also been found to be associated with the recurrent deletions at 17pl 1.2 in Smith-Magenis syndrome (Shaw et al, 2002) and at 7ql 1.23 in Williams-Beuren syndrome (Perez Jurado, 2003). However, the unequal recombination occurs within hotspots of recombination as I have observed for 15ql l-ql3 and the PWS/AS region. The genetic distance from D15S541 to D15S144, markers which flank the common  152 deletion, is ~24 cM in females and -24.8 cM in males for an ~11 Mb physical distance. This level of recombination is exceptionally high, particularly in light of the observation that recombination is generally suppressed at or near the centromere (Christian et al, 1995). In particular, recombination is highest at the common breakpoints. The region from 738CA to D15S541 (BP1) has a genetic distance of 8.4 cM in females and 2.6 cM in males for a <2.2 Mb physical distance and the regionfromD15S165 to D15S144 (BP5) has an even longer genetic map of 6.5 cM and 5.7 cM in CEPH females and males respectively for a region estimated to be -2.4 Mb. BP4, which is found between D15S156 and D15S165, also has a recombination rate that is double the genomic average of lcM/Mb, with genetic distances of 7 cM in females and 8 cM in males - an interval only 3.4 Mb long, based on physical data from May 2004. Christian et al. (1995) found unexpectedly high female recombination in the pericentromeric long arm of human chromosome 15 based on centromeric mapping in triploids. They observed a genetic distance of 12.0 cM from the centromere to D15S541/542 for a physical distance estimated cytogenetically to be much less than one metaphase band (<5 Mb). This is more than double the genomic average of lcM/IMb and represents an exception to the common assumption that recombination is decreased at human centromeric regions. However, other mapping data (Morton et al, 1999) suggested this rate might be much lower if the data is based on CEPH pedigrees using a centromeric a-satellite marker. Recombination in mothers of triploids may not reflect the normal situation. However, our data in CEPH female controls supports this high rate of meiotic recombination in females as the pericentromeric region of chromosome 15 was estimated to have a genetic distance of 8.4 Mb.  153 The data presented here indicate that there may be an association between the increased levels of homologous recombination seen in the 15ql l-ql3 region and the common chromosomal rearrangements found, particularly the common 4 Mb deletion that leads to either Prader-Willi syndrome or Angelman syndrome, depending on the parent-of-origin. The pericentromeric region including the two common proximal deletion breakpoints (BP1 and BP2) and the most common distal breakpoint (BP5) were all found to have high levels of recombination. However, although there is high recombination around BP1 (interval 1) there are no confirmed crossovers in BP2 (interval 2). The relationship to deletion formation is therefore not clear. This may be an artifact of crossover resolution, such that where an event occurs and where it resolves and is observed are two different things. Also, the physical distance between intervals 1 and 2 is fairly small. Recombination and nondisjunction: Previously, reduced recombination had been reported in the centromeric region of chromosome 15 for meiosis I error chromosomes (Robinson et al, 1998). The data presented here support this. Reduced recombination was found for MI errors (13.8 cM compared to 24 cM in CEPH controls), and recombination was increased in the centromeric region in meiosis II error chromosomes (36 cM compared to 24 cM in CEPH controls). Although the increased recombination found in this region in general in control females may predispose it to deletions and other rearrangements as mentioned above, it appears to be protective and ensures proper disjunction of chromosomes 15 at meiosis I and supports previous data. On the other hand, too much recombination appears to have a deleterious effect on the separation of chromosomes at meiosis II.  154 From my data it seems that there may be some specific recombination hotspots that are important for proper chromosome disjunction. There is a decrease in recombination in the MI error group in intervals 3, 5 and 8, while there is increased recombination in interval 9 when compared to CEPH control females (Figure 7-6). Interval 8 has been found to be a female hotspot for recombination, in this and previous studies, and it may be important for pairing of the chromosomes. This decrease found in MI error chromosomes may predispose them to segregation errors. There is an increase in recombination in the M i l error group in intervals 3, 6 and 7 with a decreased amount of recombination in interval 8 when compared to CEPH control females (Figure 7-6). Again it seems that interval 8 may be the important pairing site and this decrease may, once again, predispose M i l error chromosomes to missegregate. Although these decreases are not statistically significant, it would be interesting to look into this region further. It would also be useful to continue the analysis of recombination in this region and extend proximally the microsatellite marker typings in the MI and M i l error groups to 738CA, which had not been possible at the time of this project. As interval 1 is another female hotspot of recombination, I would like to know what the data shows for MI and M i l error chromosomes in this region. I would expect to see a significant decrease in recombination in both groups of women when compared to CEPH control females. From these studies, recombination appears to play an important role in the proper disjunction of chromosomes, such that a significant decrease or increase in the 15ql l-ql3 region is associated with nondisjunction at meiosis I or meiosis II, respectively. This is in agreement with previous studies of chromosome 15 nondisjunction and nondisjunction of chromosomes 16 and 21. As well, a decrease at a particular site - a potential pairing site in  155 interval 8 - is found to be important for in the missegregation of chromosomes in meiosis I and II.  156 Chapter 8: Discussion and Conclusion Nondisjunction (ND) of chromosomes and aneuploidy is afrequentcause of human genetic disease and increased understanding of the underlying mechanisms by which it occurs is important for reproductive health. Little is known about the mechanisms and causes of ND and how it can be predicted and prevented, even though there is extensive information about the association between the risk of trisomy in pregnancy and increasing maternal age. A number of researchers have hypothesized that the underlying mechanism involved in chromosome ND may be related to maternal aging. In order to test some of these hypotheses the studies discussed here were undertaken. Some features of human chromosomes are associated with aging. These include telomere length, replication timing at loci including centromeres, and somatic aneuploidy. A relationship between age and recombination has also been observed for chromosomes 15 and 21. In order to determine if any of these features could be used as clinical indicators of susceptibility to nondisjunction, I studied and compared them in two groups of women. I expected that the case group of women, having experienced at least one trisomic pregnancy, would have features associated with older women. Specifically, I expected there to be decreased telomere length, increased replication asynchrony, increased somatic aneuploidy and altered levels of recombination in these women when compared to age-matched controls. There was a trend in average telomere length toward being shorter in the case group of women compared to controls. As decreasing telomere length is associated with increasing age, this was what I had expected. When both the case and controls groups were subdivided based on pregnancy history and age some further trends appeared. The case group included women having only one trisomy and women having recurrent trisomy, with a wide range of  157 ages. The control group included women with no history of pregnancy loss and some had children and some did not. As well, some control women, referred to as "reproductively healthy (RH)" in Chapter 3 had children after the age of 37 as well as no history of pregnancy loss. When the case group was analyzed and subdivided based on age at first trisomy, there was a trend toward shorter telomeres in women who were < 35 years old at first trisomy. When the control group was subdivided based on reproductive history, the RH group, despite being the oldest group, had the longest telomeres. Although telomere length does not show a clear relationship to reproductive health, the trend is interesting. Replication timing data also showed some interesting trends when comparing the case and control groups. At both loci studied (15Z4 and 21q22) the younger women in the case group had an increased amount of replication asynchrony compared to control women of the same age, while the older case women had levels comparable to older control women. At 21q22, replication asynchrony was significantly increased in the case group when compared to controls, and this significant difference was also seen when the case group was subdivided based on age at first trisomy. The case women <35 years old had increased replication asynchrony when compared to control women <35 years old. It appears that abnormalities of the chromatin reflected in replication timing may play a role in trisomy occurring in younger women. Somatic aneuploidy has also been found to increase with age and I measured aneuploidy at 3 loci (15Z4, 21q22 and HPRT) in this study. There is once again, a trend towards an increased amount of aneuploidy in the case group of women compared to control women, although the samples were small, and no conclusions could be drawn. Recombination has long been the focus of study for many groups interested in its role in nondisjunction and chromosomal abnormalities. A reduced amount of recombination is  158 associated with nondisjunction at meiosis I while an increase in recombination has been found to be associated with nondisjunction at meiosis II for most chromosomes. I looked at recombination and its relationship to nondisjunction of chromosome 15 and its possible role in chromosome 15 rearrangements in 15ql l-ql3. As had been shown in previous studies, there was a decrease in recombination in 15ql l-ql3 on the meiosis I error chromosomes, while there was an increase in recombination along the meiosis II error chromosomes in this region, when compared to the CEPH female control map. Of particular interest was the finding that there may be important pairing sites at which recombination is most important. Decreased recombination was found in interval 8 for both MI and M i l chromosomes, indicating that it may be important for the proper disjunction of chromosome 15 at meiosis. In all cases in this study, when the case and control groups were subdivided based on age at first trisomy, the trend toward a significant difference was stronger in the younger women. Women presenting with a trisomic pregnancy at a young age, and even recurrent trisomy, are particularly interesting, as their chronological age would not predict that such events would be likely. However, their biological age may be different and reflected in some aspects of their chromosomes. Specifically, looking at younger women, that is women experiencing a trisomy under 35 years of age, and comparing them to age-matched control women should tell us something about their apparent predisposition to nondisjunction. It appearsfrommy data that trisomy that occurs in women under 35 years of age is likely due to some of these chromosomal features, while the mechanism in women having a trisomic pregnancy after the age of 35 may be different. I had the opportunity to study an interesting family that appeared to show an inherited predisposition to nondisjunction chromosome 21. As outlined in Chapter 5, there are four  159 cases of Down syndrome in this family all presenting to different women. The women were all under the age of 30 at the time of the trisomic births. This made them of particular interest and I studied all features of the chromosomes in these women. The trends were the same when comparing the four women in this group to controls, with the obvious problem of sample size. Telomere length was slightly shorter than normal, somatic aneuploidy was increased at one locus (15Z4) and the amount of recombination was altered. Surprisingly, recombination was found to be increased in this family compared to CEPH female controls. Although no obvious structural abnormality could be detected in these women cytogenetically, molecular analysis indicates that there is likely a duplicated region involving the chromosome 21 centromere in two women, which may explain their increased recombination and predisposition to nondisjoin chromosome 21. All of the features of chromosomes that I studied seem to show a trend toward what I predicted and what would be expected if women experiencing a trisomic pregnancy were prematurely aged. However, due to limited sample size, few of the findings and trends, although interesting, are conclusive or significant. Larger studies with larger samples should be carried out in order to further clarify whether or not these trends can be used as clinical indicators of a predisposition to nondisjunction in some women. Increasing the number of case women under the age of 35 at their first trisomy is going to increase our understanding of the mechanism leading to nondisjunction in this group. Changes in chromatin with age or in women experiencing a trisomy is of most interest for future studies, and larger samples are needed in order to better understand the mechanism involved. It would also be useful to measure the amount of methylation in the genome and at particular loci in women experiencing a trisomy in order to determine if  160 methylation or other such epigenetic modifications of the DNA may be involved in a predisposition to nondisjunction. As methylation and chromatin changes are possibly involved in replication timing, recombination and telomere length. The relationship between telomere length and epigenetic changes of the chromosomes was investigated in Chapter 6. I compared the telomere lengths of newborns conceived by ICSI with control newborns conceived naturally. There have been a number of reports implicating the ICSI procedure in imprinting disorders, as a number of ICSI newborns have had such syndromes. Telomere length has also been extensively studied in cloned animals and has been found to be altered - most often shorter than animals of the same age. Telomere length was not found to differ between ICSI and control newborns in my study, suggesting that the resetting of telomere length is normal in ICSI embryos. As decreased telomere length has been associated with cancers and premature aging syndromes, this was a concern for babies conceived through the ICSI procedure. I have always been interested in meiosis and chromosome behaviour, and the impact it has on human reproduction. The proper disjunction of chromosomes and the factors that are involved are so complex that I have wanted to be part of the research to help better understand the mechanism and why things go wrong. After completion of the studies discussed here, I still believe that, given the proper population of women to study, we can find some of the answers. Chromosome structure, and the importance of centromeres and telomeres to the process of meiosis still intrigue me. The sizes and stability of these structures are likely very important to the proper pairing and segregation of chromosomes at meiosis. This may not be related to aging necessarily, but may be another factor that contributes to risk in a multifactorial way. I also believe that methylation and chromatin  161 changes of the genes and chromosomes may play an important role in the proper segregation of chromosomes, and as mentioned above, more investigation is needed into this. The field of nondisjunction and aneuploidy in humans is a fascinating and important one. Nondisjunction results in abnormalities of chromosome number that often leads to pregnancy loss and the most recent data indicates that some women are predisposed to having a trisomy. If some factors or genetic mechanisms can be identified that can be used as clinical indicators of nondisjunction susceptibility, then they can be used in the genetic counseling of women experiencing pregnancy loss and recurrent trisomy. It would also help in our general understanding of the process of meiosis. Despite its importance to human reproduction and the number of groups involved in research in order to unravel the mystery, very little is known about the cause of nondisjunction. Although a lot of new and important research has come out of this field of nondisjunction and aneuploidy, we have yet to uncover any conclusive evidence of a direct cause of nondisjunction in humans.  162 References: Abruzzo, M A and TJ Hassold. Etiology of Nondisjunction in Humans. Env. Mol. Mut. 25 (suppl 26):38-47, 1995. Abruzzo, MA, DK Griffin, EA Millie, LA Sheean and TJ Hassold. The effect of Ychromosome alpha-satellite array length on the rate of sex chromosome disomy in human sperm. Hum. Genet. 97:819-823, 1996. Alberman, E, M Elliott, M Creasy and R Dhadial. Previous reproductive history in mothers presenting with spontaneous abortions. Br. J. Obstet. Gynaecol. 82(5):366-373, 1975. Alberman, E. The scope of perinatal statistics and the usefulness of international comparisons. Proc. Annu. Symp. Eugen. Soc. 17:57-73, 1981. Amiel, A, L Avivi, E Gaber and MD Fejgin. Asynchronous replication of allelic loci in Down syndrome. Eur J Hum Genet. 6:359-364,1998. Amiel, A, A Korenstein, E Gaber and L Avivi. Asynchronous replication of alleles in genomes carrying an extra autosome. Eur J Hum Genet. 7:223-230,1999. Amiel, A, O Reish, E Gaber, I Kedar, R Diukman and M Fejgin. Replication asynchrony increases in women at risk for aneuploidy offspring. Chromosome Research 8: 141-150, 2000. Amiel, A, O Reish, E Gaber, R Masterman, T Tohami and MD Fejgin. Asynchronous replication of alleles in genomes carrying a microdeletion. IMAJ. 4:702-705, 2002. Amiel, A, G Peretz, H Slor, G Weinstein and MD Fejgin. Molecular cytogenetic parameters in fibroblasts from patients and carriers of Xeroderma pigmentosum. Cancer Genet and Cytogenet. 149:154-160, 2004. Amos-Landgraf, JM, Y Ji, W Gottlieb, T Depinet, AE Wandstrat, SB Cassidy, DJ Driscoll, PK Rogan, S Schwartz and RD Nicholls. Chromosome Breakage in the Prader-Willi and Angelman Syndromes Involves Recombination between Large, Transcribed Repeats at Proximal and Distal Breakpoints. Am. J. Hum. Genet. 65:370-386, 1999. Anderson, LK, A Reeves, L M Webb and T Ashley. Distribution of crossing over on mouse synaptonemal complexes using immunofluorescent localization of MLH1 protein. Genetics 151(4):1569-1579, 1999. Angell, R. First-Meiotic-Division Nondisjunction in Human Oocytes. Am. J. Hum. Genet. 61:23-32,1997.  163 Arbour, L, B Christensen, T Delormier, R Piatt, B Gilfix, P Forbes, I Kovitch, J Morel and R Rozen. Spina Bifida, Folate Metabolism, and Dietary Folate Intake in a Northern Canadian Aboriginal Population. Int. J. Circumpolar Health. 61:341-351, 2002. Arnheim, N , P Calabrese and M Nordborg. Hot and Cold Spots of Recombination in the Human Genome: the Reason We Should Find Them and How, This Can Be Achieved. Am. J. Hum. Genet. 73:5-16,2003. Aviv, A, D Levy and M Mangel Growth, telomere dynamics and successful and unsuccessful human aging. Mech. Ageing Dev. 124:829-837, 2003. Avramopoulos, M Mikkelsen, D Vassilopoulos, M Grigoriadou and MB Petersen. Apolipoprotein E allele distribution in parents of Down's syndrome children. Lancet. 347(9005):862-5, 1996. Ayme, S and A, Lippman-Hand. Maternal-age effect in aneuploidy: does altered embryonic selection play a role? Am. J. Hum. Genet. 34(4):558-565, 1982. Bailey, JA, Z Gu, RA Clark, K Reinert, RV Samonte, S Schwartz, MD Adams, EW Myers, PW Li and EE Eichler. Recent Segmental Duplications in the Human Genome. Science 297:1003-1007,2002. v  Baker, SM, AW Plug, TA Prolla, CE Bronner, AC Harris, X Yao, D-M Christie, C Monell, N Arnheim, A Bradley, T Ashley and R M Liskay. Involvement of mouse Mlhl in DNA mismatch repair and meiotic crossing over. Nature Genetics. 13(3): 336-342, 1996. Baker, DJ, KB Jeganathan, JD Cameron, M Thompson, S Juneja, A Kopecka, R Kumar, RB Jenkins, PC deCtroen, P Roche and JM vanDeursen. BubRl insufficiency causes early onset of aging-associated phenotypes and infertility in mice. Nature Genetics 36(7): 744-749, 2004. Barbosa, ACC, PA Otto and A M Vianna-Morgante. Replication timing of homologous asatellite DNA in Roberts syndrome. Chrom. Res. 8:645-650, 2000. Barlow, A L and M A Hulten. Crossing over analysis at pachytene in man. Eur. J. Hum. Genet. 6(4):350-358, 1998. Beermann, F, I Bartels, U Franke and I Hansmann. Chromosome segregation at meiosis I in female T(2;4)lGo/+m mice: no evidence for a decreased crossover frequency with maternal age. Chromosoma 95(1): 1-7,1987. Bonduelle, M , I Ponjaert, A V Steirtegham, MP Derde, P Devroey and I Liebaers. Developmental outcome at 2 years of age for children born after ICSI compared with children born after IVF. Hum. Reprod. 18(2):342-350,2003.  164 Bosco, P, R M Gueant-Rodriguez, G Anello, C Barone, F Namour, F Caraci, A Romano, C Romano and JL Gueant. Methionine synthase (MTR) 2756 (a to G) polymorphism, double heterozygosity methionine synthase 2756 AG/methionine synthase reductase (MTRR) 66 AG, and elevated homocysteinemia are three risk factors for having a child with Down syndrome. Am. J. Hum. Genet. 121(3):219-224, 2003. Brandriff, B, L Gordon, L Ashworth, G Watchmaker, A Carrano and A Wyrobek. Chromosomal abnormalities in human sperm: Comparisons among four healthy men. Human Genetics 66: 193-201, 1984. Brook, JD, RG Gosden and AC Chandley. Maternal ageing and aneuploid embryos Evidence from the mouse that biological and not chronological age is the important influence. Hum. Genet. 66:41-45, 1984. Bruyere, H, R Rupps, BD Kuchinka, JM Friedman and WP Robinson. Recurrent Trisomy 21 in a Couple With a Child Presenting Trisomy 21 Mosaicism and Maternal Uniparental Disomy for Chromosome 21 in the Euploid Cell Line. Am. J. Med. Genet. 94:2000. Bugge, M , A Collins, MB Petersen, J Fisher, C Brandt, JM Hertz, L Tranebjaerg, C de Lozier-Blanchet, P Nicolaides, K Brondum-Nielsen, N Morton and M Mikkelsen. Nondisjunction of chromosome 18. Hum. Mol. Genet. 7(4):661-669, 1998. Butler, MG, J Tilburt, A DeVries, B Muralidhar, G Aue, L Hedges, J Atkinson and H Schwartz. Comparison of chromosome telomere integrity in multiple tissues from subjects at different ages. Cancer Genet. Cytogenet. 105(2):138-144, 1998. Carpenter, AT. Meiotic roles of crossing-over and of gene conversion. Cold Spring Harb Symp. Quant. Biol. 49:23-29, 1984. Carrozzo, R, E Rossi, SL Christian, K Kittikamron, C Livieri, A Corrias, L Pucci, A Fois, P Simi, L Bosio, L Beccaria, O Zuffardi and DH Ledbetter. Inter-and Intrachromosomal Rearrangements Are Both Involved in the Origin of 15ql l-ql3 Deletions in Prader-Willi Syndrome. Am. J. Hum. Genet. 61:228-231, 1997. Cawthon, RM, KR Smith, E O'Brien A Sivatchenko and RA Kerber. Association between telomere length in blood and mortality in people aged 60 years or older. The Lancet 361:393-195, 2003. Chadefaux-Vekemans, B, M Coude, F Muller, JF Oury, A Chali, J Jais and P Kamoun. Methylenetetrahydrofolate reductase polymorphism in the etiology of Down syndrome. Pediatr. Res. 51(6):766-767, 2002. Chartier-Harlin, M Parfitt, S Legrain, J Perez-Tur, T Brousseau, A Evans, C Berr, O Vidal, P Roques, V Gourlet et al. Apolipoprotein E, epsilon 4 allele as a major risk factor for sporadic early and late-onset forms of Alzheimer's disease: analysis of the 19ql3.2 chromosomal region. Hum Mol Genet. 3(4):569-74, 1994.  165 Christian, SL, WP Robinson, B Huang, A Mutirangura, MR Line, M Nakao, U Surti, A Chakravarti and DH Ledbetter. Molecular Characterization of Two Proximal Deletion Breakpoint Regions in Both Prader-Willi and Angelman Syndrome Patients. Am. J. Hum. Genet. 57:40-48, 1995. Christian, SL, JA Fantes, SK Mewborn, B Huang and DH Ledbetter. Large genomic duplicons map to sites of instability in the Prader-Willi/Angelman syndrome chromosome region (15qll0ql3). Hum. Mol. Genet. 8(6):1025-1037, 1999. Christianson, RE, SL Sherman and CP Torfs. Maternal meiosis II nondisjunction I trisomy 21 is associated with maternal low socioeconomic status. Genet. Med. 6(6):487-494, 2004. Cockwell, AE, PA Jacobs, SJ Beal and JA Crolla. A study of cryptic terminal chromosome rearrangements in recurrent miscarriage couples detects unsuspected acrocentric pericentromeric abnormalities. Hum. Genet. 112:298-302, 2003. Copeland, NG, NA Jenkins, DJ Gilbert, JT Eppig, LJ Maltais, JC Miller, WF Dietrich, A Weaver, SE Lincoln, RG Steen et al., A genetic linkage map of the mouse: current applications and future prospects. Science 262(5130):57-66, 1993. Cox, GF, J Burger, V Lip, UA Mau, K Sperling, B-L Wu and B Horsthemke. Intracytoplasmic Sperm Injection May Increase the Risk of Imprinting Defects. Am. J. Hum. Genet. 71:162-164, 2002. Czeizel, AE and E Csaba. Environmental Trichlorfon and Cluster of Congenital Abnormalities. Lancet 341:539-542, 1993. DeBaun, MR, EL Niemetz and AP Feinberg. Association of in vitro fertilization with Beckwith-Wiedemann syndrome and epigenetic alterations of LIT1 and HI9. Am. J. Hum. Genet. 72:150-160, 2003. DeLange, T. Telomere dynamics and genome instability. In Telomeres, (eds Blackburn, EH and CW Greider) 265-293 (Cold Spring Harbor Laboratory Press, Cold Spring Harbor, New York, 1995) DeRycke, M , I Liebaera and A Van Steirtegham. Epigenetic risks related to assisted reproductive technologies: risk analysis and epigenetic inheritance. Hum. Reprod. 17:24872494, 2002. Devroey, P and A Van Steirteghem. A review of ten years experience of ICSI. Hum. Reprod. Update. 10(l):19-28, 2004. Donis-Keller, H, P Green, C Helms, S Cartinhour, B Weiffenbach, K Stephens, TP Keith, DW Bowden, DR Smith, ES Lander, D Botstein, G Akots, KS Rediker, T Gravius, V A Brown, MB Rising, C Parker, JA Powers, DE Watt, ER Kauffman, A Bricker, P Phipps, H Muller-Kahle, TR Fulton, S Ng, JW Schumm, JC Braman, RG Knowlton, DF Barker, SM  166 Crooks, SE Lincoln, MJ Daly and J Abrahamson. A Genetic Linkage Map of the Human Genome. Cell 51: 319-337, 1987. Donlon, TA, M Lalande, A Wyman, G Bruns and SA Latt. Isolation of molecular probes associated with the chromosome 15 instability in the Prader-Willi syndrome. PNAS 83(18):6964,1986. Dorland, M , JM van Montfrans, RJ van Koolj, CB Lambalk and ER te Velde. Normal telomere lengths in young mothers of children with Down's syndrome. The Lancet 352:961962, 1998. Dunn, LC and D Bennett. Sex differences in recombination of linked genes in animals. Genet. Res. 9(2):211-220, 1967. Edelmann, L, RK Pandita and BE Morrow. Low-copy repeats mediate the common 3-Mb deletion in patients with velo-cardio-facial syndrome. Am. J. Hum. Genet. 64:1076-1086, 1999. Ensminger, AW and A Chess. Coordinated replication timing of monoallelically expressed genes along human autosomes. Hum. Mol. Genet. 13(6):651-658, 2004. Faddy, MJ and RG Gosden. A model conforming the decline in follicle numbers to the age of menopause in women. Hum. Reprod. 11(&):1484-1486, 1996. Fitzgerald, PH, AF Pickering, JM Mercer and PM Miethke. Premature centromere division: A mechanism of non-disjunction causing X chromosome aneuploidy in somatic cells of man. Ann. Hum. Genet. 38:417-428, 1975. Fitzgerald, PH and C M McEwan. Total Aneuploidy and Age-Related Sex Chromosome Aneuploidy in Cultured Lymphocytes of Normal Men and Women. Hum. Genet. 39:329337, 1977. Ford, JH. Spindle microtubular dysfunction in mothers of Down syndrome children. Hu. Genet. 68: 295-298, 1984. Freeman, SB, Q Yang, K Allran, LF Taft and SL Sherman. Women with a Reduced Ovarian Complement May Have an Increased Risk for a Child with Down Syndrome, Am. J. Hum. Genet. 66:1680-1683,2000. Gardner, RL and MF Lyon, X chromosome inactivation studied by injection of a single cell into the mouse blastocyst. Nature 231: 385-386, 1971. Gaulden, ME. Maternal age effect: The enigma of Down syndrome and other trisomic conditions. Mut. Res. 296:69-88, 1992.  167 Geller, LN and H Potter Chromosome Missegregation and Trisomy 21 Mosaicism in Alzheimer's Disease. Neurobiology of Disease 6: 167-179, 1999. Giglio, S, V Calvari, G Gregato, G Gimelli, S Camanimi, R Giorda, A Ragusa, S Guerneri, A Selicorni, M Stumm, H Tonnies, M Ventura, M Zollino, G Neri, J Barber, D Wieczorek, M Rocchi and O Zuffardi. Heterozygous submicroscopic inversions involving olfactory receptor-gene clusters mediate the recurrent t(4;8)(pl6;p23) translocation. Am. J. Hum. Genet. 71(2):276-285, 2002. Gimelli. G, M A Pujana, MG Patricelli, S Russo, D Giardino, L Larizza, J Cheung, L Armengol, A Schinzel, X Estivill and O Zuffardi. Genomic inversions of human chromosome 15ql l-ql3 in mothers of Angelman syndrome patients with class II (BP2/3) deletions. Hum. Mol. Genet. 12(8):849-858, 2003. Gosden, R, J Trasler, D Lucifero and M Faddy. Rare congenital disorders, imprinted genes, and assisted reproductive technology. Lancet 361:1975-1977, 2003. Gougeon, A, R Ecochard and JC Thalabard. Age-Related Changes of the Population of Human Ovarian Follicles: Increase in the Disappearance Rate of Non-Growing and EarlyGrowing Follicles in Aging Women. Biol. Reprod. 50:653-663, 1994. Griffin, DK. The Incidence, Origin, and Etiology of Aneuploidy. International Review of Cytology, 167: 263-296, 1996. Gueant, JL, R M Gueant-Rodriguez, G Anello, P Bosco, L Brunaud, C Romano, R Ferri, A Romano, M Candito and B Namour. Genetic determinants of folate and vitamin B12 metabolism: a common pathway in neural tube defect and Down syndrome? Clin. Chem. Lab. Med. 41(11):1473-1477, 2003. Haber, JE. Recombination: afrankview of exchanges and vice versa. Curr. Opin. Cell Biol. 12:286-292, 1999. Handel, MA, J Cobb and S Eaker. What are the spermatocyte's requirements for successful meiotic division? J Exp Zool 285(3): 243-250. Review, 1999. Hanish, JP, JL Yanowitz and T de Lange. Stringent sequence requirements for the formation of human telomeres. PNAS 91(19):8861-8865, 1994. Hassold, T and D Chiu. Maternal age-specific rates of numerical chromosome abnormalities with special reference to trisomy. Hum. Genet. 70:11-17, 1985. Hassold, TJ The origin of aneuploidy in humans. In: Vaughn-Dellarco et al. eds. Aneuploidy: etiology and mechanisms. NY: Plenum Press, 1985: 103-116, 1985. Hassold, T, D Pettay, A Robinson and I Uchida. Molecular studies of parental origin and mosaicism in 45, X conceptuses. Hum. Genet. 89(6):647-652, 1992.  168 Hassold, T, S Sherman and PA Hunt. The Origin of Trisomy in Humans. Etiology and Pathogenesis of Down Syndrome, pp 1-12, 1995. Hassold, T, M Merrill, K Adkins, S Freeman and S Sherman. Recombination and maternal age-dependent nondisjunction: molecular studies of trisomy 16. Am. J. Hum. Genet. 57:867874, 1995. Hassold, T, M Abruzzo, K Adkins, D Griffin, M Merrill, E Millie, D Saker, J Shen and M Zaragoza. Human Aneuploidy: Incidence, Origin, and Etiology. Env. Mol. Mut. 28:167175, 1996. Hassold, T and S Sherman. Down syndrome: genetic recombination and the origin of the extra chromosome 21. Clin. Genet. 57:95-100, 2000. Hassold, T, S Sherman and P Hunt. Counting cross-overs: characterizing meiotic recombination in mammals. Hum. Mol. Genet. 9(16):2409-2419, 2000. Hassold, T and P Hunt, To Err (Meiotically) is Human: The Genesis of Human Aneuploidy. Nature 2: 280-291, 2001 Hassold, TJ, LC Burrage, ER Chan, L M Judis, S Schwartz, SJ James, PA Jacobs and NS Thomas. Maternal Folate Polymorphisms and the Etiology of Human Nondisjunction. Am. J. Hum. Genet. 69: 434-439, 2001. Hassold, T, L Judis, ER Chan, S Schwartz, A Seftel and A Lynn. Cytological studies of meiotic recombination in human males. Cytogenet. Genome Res. 107: 249-255, 2004. Hawley, S Exchange and chromosome segregation in eukaryotes. In Genetic Recombination (eds. R. Kucherlapati and G. Smith) 497-525,1988. Hawley, RS, H Irick, AE Zitron, DA Haddox, A Lohe, C New, MD Whitley, T Arbel, J Jang, K McKim et al. There are two mechanisms of achiasmate segregation in Drosophila females, one of which requires heterochromatic homology. Dev. Genet. 13: 440-467,1993. Hecht, F, BK Hecht and CS Berger. Aneuploidy in recurrent spontaneous aborters: the tendency to parental nondisjunction. Clinical Genetics. 26:43-45, 1984. Henderson, SA and RG Edwards. Chiasmafrequencyand maternal age in mammals. Nature 218(136):22-28, 1968. Herrmann, FH, L Salazar-Sanchez, G Schuster, G Jimenez-Arce, R Grimm, X Gomez, M Chavez, K Wulff and W Schroder. Prevalence of Eight Molecular Markers Associated With Thrombotic Diseases in Six Amerindian Tribes and Two African Groups of Costa Rica. Am. J. Hum. Biol. 16:82-86, 2004.  169 Hobbs, CA, SL Sherman, P Y i , SE Hopkins, CP Torfs, RJ Hine, M Pogribna, R Rozen and SJ James. Polymorphisms in Genes Involved in Folate Metabolism as Maternal Risk Factors for Down Syndrome. Am. J. Hum. Genet. 67:623-630, 2000. Hodges, CA, A Hagan, D Jennings, R Keri, J Nilson and PA Hunt. Experimental evidence that changes in oocyte growth influence meiotic chromosome segregation. Hum. Reprod. 17(5):1171-1180, 2002. Hook, EB. Risk of Down syndrome in relation to maternal age. Lancet 2(7983):465, 1976. Hook, EB. Rates of chromosome abnormalities at different maternal ages. Obstet. Gynecol. 58(3):282-285, 1981. Hubert, R, M MacDonald, J Gusella and N Arnheim. High resolution localization of recombination hot spots using sperm typing. Nature Genetics 7:420-424, 1994. Hunt, PA and TJ Hassold. Sex Matters in Meiosis. Science 296:2181-2183, 2002. Hunt, PA, KE Koehler, M Susiarjo, CA Hodges, A Ilagan, RC Voigt, S Thomas, BF Thomas and TJ Hassold. Bisphenol a exposure causes meiotic aneuploidy in the female mouse. CurrBiol. 13(7):546-53, 2003. Jacobs, PA, M Brunton, W M Court Brown, R Doll and H Goldstein. Change of human chromosome count distribution with age: Evidence for a sex difference. Nature 197:10801081, 1961. Jacobs, PA, W M Court Brown, R Doll. Distribution of human chromosome counts in relation to age. Nature 16:1178-1180, 1963. Jacobs, PA The chromosome complement of human gametes. Oxf. Rev. Reprod. Biol. 14:47-72, 1992. James, SJ, M Pogribna, IP Pogribny, S Melnyk, RJ Hine, JB Gibson, P Y i , DL Tafoya, DH Swenson, V L Wilson and DW Gaylor. Abnormal folate metabolism and mutation in the methylenetetrahydrofolate reductase gene may be maternal risk factors for Down syndrome. Am. J. Clin. Nutr. 70:495-501,1999. Jamieson, JR Courts and JM Connor. The chromosome constitution of human preimplantation embryos fertilized in vitro. Hum Reprod 9: 709-715, 1994. Jeffreys, AJ, A Ritchie and R Neumann. High resolution analysis of haplotype diversity and meiotic crossover in the human TAP2 recombination hotspot. Human Molecular Genetics 9(5): 725-733, 2000.  170 Jeon, HY, SH Hyun, GS Lee, HS Kim, S Kim, YW Jeong, SK Kang, BC Lee, JY Han, C Ahn and WS Hwang. The Analysis of Telomere Length and Telomerase Activity in Cloned Pigs and Cows. Mol. Reprod. Dev. 71:315-320, 2005. Ji, Y , NA Rebert, JM Joslin, MJ Higgins, RA Schulz and RD Nicholls. Structure of the Highly Conserved HERC2 Gene and of Multiple Partially Duplicated Paralogs in Human. Genome Research 10:319-329, 2000. Jones, M , R Wagner and M Radman. Mismatch repair and recombination in E. Coli. Cell. 50(4):621-626, 1987. Jorde, LB, JC Carey, MJ Bamshad and RL White. Medical Genetics, 3 Edition, 2003. Juberg, RC, J Knops and PN Mowrey. Increased frequency of lymphocytic mitotic nondisjunction in recurrent spontaneous aborters. J. Med. Genet. 22(l):32-35, 1985. rd  Kalousek, DK and FJ Dill, Chromosomal mosaicism confined to the placenta in human conceptions. Science 221: 665-667, 1983. Kalousek, DK and M Vekemans. Confined placental mosaicism and genomic imprinting. Bailliere's Clin. Ob. Gyn. 14(4):723-730, 2000. Kanaar, R, JH Hoeijmakers and DC van Gent. Molecular mechanisms of DNA double strand break repair. Trends Cell Biol. 8(12):483-489, 1998. Kitsberg, D, S Selig, M Bradeis, I Simon, I Keshet, DJ Driscoll, RD Nicholls and H Cedar. Allele-specific replication timing of imprinted gene regions. Nature. 364:459-463, 1993. Kline, J, B Levin, Z Stein, D Warburton and R Hindin. Cigarette Smoking and Trisomy 21 at Amniocentesis. Genetic Epidemiology 10:35-42, 1993. Kline, J and B Levin. Trisomy and age at menopause: predicted associations given a link with rate of oocyte atresia. Paed. Perinat. Epid. 6:225-239, 1992. Kline, J, A Kinney, B Levin and D Warburton. Trisomic Pregnancy and Earlier Age at Menopause. Am. J. Hum. Genet. 67: 395-404, 2000. Kline, J, A Kinney, M L Reuss, A Kelly, B Levin, M Ferin and D Warburton. Trisomic pregnancy and the oocyte pool. Hum. Reprod. 19(7): 1633-1643, 2004. Koehler, KE, RS Hawley, S Sherman and T Hassold. Recombination and nondisjunction in humans and flies. Hum. Mol. Genet. 5:1495-1504, 1996. Koehler, KE and TJ Hassold. Human aneuploidy: lessons from achiasmate segregation in Drosophila melanogaster. Ann. Hum. Genet. 62:467-479, 1998.  171 Koehler, KE, EA Millie, JP Cherry, SE Schrump and TJ Hassold Meiotic Exchange and Segregation in Female Mice Heterozygous for Paracentric Inversions. Genetics 166: 11991214, 2004. Kolomietz, E, MS Meyn, A Pandita and JA Squire. The Role ofAlu Repeat Clusters as Mediators of Recurrent Chromosomal Aberrations in tumors. Genes, Chrom. And Cancer 35:91-112, 2002. Kong, A, J Barnard, DF Gudbjartsson, G Thorleifsson, G Jonsdottir, S Sigurdardottir, B Richardsson, J Jonsdottir, T Thorgeirsson, ML Frigge, NE Lamb, S Sherman, JR Gulcher and K Stefansson. Recombination rate and reproductive success in humans. Nature Genetics. 36(11): 1203-1206, 2004. Kotzot, D. Abnormal phenotypes in uniparental disomy (UPD): fundamental aspects and a critical review with bibliography of UPD other than 15. Am. J. Med. Genet. 82:265-274, 1999. Krejci, K and J Koch, Improved detection and comparative sizing of human chromosomal telomeres in situ. Chromosoma 107:198-203, 1998. Laitinen, J, J Samarut and E Holtta. A nontoxic and versatile protein salting-out method for isolation of DNA. Biotechniques 17(2):361,318,320-2, 1994. Lamb, NE, SB Freeman, A Savage-Austin, D Pettay, L Taft, J Hershey, Y Gu, J Shen, D Saker, K M May, D Avramopoulos, MB Petersen, A Hallberg, M Mikkelsen, TJ Hassold and SL Sherman. Susceptible chiasmate configurations of chromosome 21 predispose to nondisjunction in both maternal meiosis I and meiosis II. Nature Genetics 11:400-405, 1996. Lamb, NE, E Feingold and SL Sherman. Estimating Meiotic Exchange Patterns From Recombination Data: An Application to Humans. Genetics. 146:1011-1017, 1997. Lamb, NE, E Feingold, A Savage, D Avramopoulos, S Freeman, Y Gu, A Hallberg, J Hersey, G Karadima, D Pettay, D Saker, J Shen, L Taft, M Mikkelsen, MB Petersen, T Hassold and SL Sherman. Characterization of susceptible chiasma configurations that increase the risk for maternal nondisjunction of chromosome 21. Hum. Mol. Genet. 6:13911399, 1997. Lamb, NE, K Yu, J Shaffer, E Feingold and SL Sherman. Association between Maternal Age and Meiotic Recombination for Trisomy 21. Am. J. Hum. Genet. 76: 91-99, 2005. Lanza, RP, JB Cibelli, C Blackwell, VJ Cristofalo, M K Francis, G M Baerlocher, J Mak, M Schertzer, EA Chavez, N Sawyer, PM Lansdorp and MD West. Extension of Cell Life-Span and Telomere Length in Animals Cloned from Senescent Somatic Cells. Science 288:665669, 2000.  172 Laurent, A - M , M Li, S Sherman, G Roizes and J Buard. Recombination across the centromere of disjoined and non-disjoined chromosome 21. Hum. Mol. Genet. 12(17):22292239, 2003. Lawrie, N M , C Tease and M A Hulten. Chiasmafrequency,distribution and interference maps of mouse autosomes. Chromosoma 104(4):308-314, 1995. Ledbetter, DH, JM Zachary, JL Simpson, MS Golbus, E Pergament, L Jackson, MJ Mahoney et al. Cytogenetic results from the US Collaborative Study on CVS. Prenat. Diagn. 12:371345,1992. Lee, SH, S Lee, HS Jun, HJ Jeong, WT Cha, YS Chjo, JH Kim, SY Ku and K Y Cha. Expression of the mitochondrial ATPase6 gene and Tfam in Down syndrome. Mol. Cells. 15(2): 181-185, 2003. Lee, JY, A Hayashi-Hagihara and TL Orr-Weaver. Roles and regulations of the Drosophila centromere cohesion protein MEI-S332 family. PTRSLBBS 360(1455): 543-552, 2005. Lejeune, J, R Turpin, M Gautier. [Mongolism; a chromosomal disease (trisomy).] Bull. Acad. Natl. Med. 143(11-12):256-265,1959. LeMaire-Adkins, R, K Radke and PA Hunt. Lack of Checkpoint Control at the Metaphase/Anaphase Transition: A Mechanism of Meiotic Nondisjunction in Mammalian Females. The J. Cell Biol. 139(7):1611-1619,1997. Lercher, MJ and LD Hurst. Imprinted Chromosomal Regions of the Human Genome Have Unusually High Recombination Rates. Genetics 165:1629-1632, 2003. Li, J, M Xu, H Zhou, J Ma, and H Potter. Alzheimer presenilins in the nuclear membrane, interphase kinetochores, and centromeres suggest a role in chromosome segregation. Cell 90: 917-927, 1997. Litmanovitch, T, M M Altaras, A Dotan and L Avivi. Asynchronous replication of homologous a-satellite DNA loci in man is associated with nondisjunction. Cytogenet. Cell Genet. 81:26-35, 1998. Lo, AWI, GCC Liao, M Rocchi and KHA Choo. Extreme Reduction of ChromosomeSpecific a-Satellite Array is Unusually Common in Human Chromosome 21. Gen. Res. 9:895-908, 1999. Lynn, A C Kashuk, MB Petersen, JA Bailey, DR Cox, SE Antonarakis and A Chakravarti. Patterns of meiotic recombination on the long arm of human chromosome 21. Gen. Res. 10(9):1319-1332, 2000.  173 Mahtani, M M and HF Willard. Pulsed-Field Gel Analysis of a-Satellite DNA at the Human X Chromosome Centromere: High-Frequency Polymorphisms and Array Size Estimate. Genomics 7:607-613, 1990. Mailhes, JB, D Young and SN London. Postovulatory Ageing of Mouse Oocytes In Vivo and Premature Centromere Separation and Aneuploidy. Biol. Reprod. 58:1206-1210, 1998. Maratou, K, Y Siddique, A M Kessling and GE Davies. Variation in alphoid DNA size and trisomy 21: A possible cause of nondisjunction. Hum. Genet. 106:525-530, 2000. Martin, RH, E Spriggs and AW Rademaker. Multicolor Fluorescence In Situ Hybridization Analysis of Aneuploidy and Diploidy Frequencies in 225,846 Sperm from 10 Normal Men. Biology of Reproduction 54: 394-398, 1996. Martinez-Garay, I, S Jablonka, M Sutajova, P Steuemagel, A Gal and K Kutsche. A New Gene Family (FAM9) of Low-Copy Repeats in Xp22.3 Expressed Exclusively in Testis: Implications for Recombination in This Region. Genomics 80(3):259-267, 2002. Masumoto et al. Properties of CENP-B and its target sequence in a-satellite DNA. In Chromosome Segregation and Aneuploidy (BK. Vig, Ed) 31-43. Spriger-Verlag, Heidelberg, 1993. Mikkelsen, M and S Ayme. Chromosomal findings in chorionic villi: a collaborative study. In: Vogel FS and K Sperling (eds) Human genetics. Springer, Berlin, pp 596-606, 1987. Mikkelsen, M . The incidence of Down's syndrome and progress towards its reduction. Philos. Trans. R. Soc. Lond. B. Biol. Sci. 319(1194):315-324, 1988. Morrison, C and S Takeda. Genetic analysis of homologous DNA recombination in vertebrate somatic cells. Int. J. Biochem. Cell. Biol. 32(8):817-831, 2000. Morton, CC, SL Christian, TA Donlon, DJ Driscoll et al., Report of the fourth international workshop on human chromosome 15 mapping 1997. Cytogenet. Cell Genet. 84:11-21, 1999. Mostoslavsky, R. et al. Asynchronous replication and allelic exclusion in the immune system. Nature. 414:221-225, 2001. Murine, M Alikani, G Tomkin, J Grifo and J Cohen. Embryo morphology, developmental rates, and maternal age are correlated with chromosome abnormalities. Fertil Steril 64:382391, 1995. Neel, JV. Genetic studies of the Atomic Bomb Casualty Commission - Radiation Effects Research Foundation: 1946-1997. PNAS 95:5432-5436, 1998. Nowinski, DL VanDyke, BC Tilley, G Jacobsen, VR Babu, MJ Worsham, GN Wilson and L Weiss. Thefrequencyof aneuploidy in cultured lymphocytes is correlated with age and gender but not with reproductive history. AJHG 46: 1101-1111, 1990.  174 O'Leary, VB, A Parle-McDermott, A M Molloy, PN Kirke, Z Johnson, M Conley, JM Scott and JL Mills. MTRR and MTHFR polymorphism: link to Down syndrome. Am. J. Med. Genet. 107(2):151-155, 2002. Ohta, K, T Shibata and A Nicolas. Changes in chromatin structure at recombination initiation sites during yeast meiosis. EMBO J. 13(23):5754-5763, 1994. Paldi, A, G Gyapay and J Jami. Imprinted chromosomal regions of the human genome display sex-specific meiotic recombination frequencies. Current Biology 5(9): 1030-1035, 1995. Paliulis, L V and RB Nicklas. The Reduction of Chromosome Number in Meiosis Is Determined by Properties Built into the Chromosomes. The J. Cell Biol. 150(6): 1223-1231, 2000. Pangalos, CC Talbot, JG Lewis, PA Adelsberger, MB Petersen, JL Serre, MO Rethore, MC deBlois, P Parent, A A Schinzel et al. DNA polymorphism analysis in families with recurrence of free trisomy 21. AJHG 51: 1015-1027, 1992. Penrose, LS. Maternal Age in Familial Mongolism. The British Journal of Psychiatry. 97(409):73 8-747, 1951. Perez Jurado, AL. Williams-Beuren syndrome: a model of recurrent genomic mutation. Horm. Res. 59 (suppl 1):106-113, 2003. Petersen, MB, G Karadima, M Samaritaki, D Avramopoulos, D Vassilopoulos and M Mikkelsen. Association Between Presenilin-1 Polymorphism and Maternal Meiosis II Errors in Down Syndrome. American Journal of Medical Genetics 93: 366-372, 2000. Phillips, OP, S Cromwell, M Rivas, JL Simpson and S Elias. Trisomy 21 and maternal age of menopause: does reproductive age rather then chronological age influence risk of nondisjunction? Human Genetics 95: 117-118, 1995. Pittman, DL and JC Schimenti. Recombination in the Mammalian Germ Line. Curr. Top. Dev. Biol. 37:1-35, 1998. Poon, SSS, U M Martens, RK Ward and PM Lansdorp. Telomere Length Measurements Using Digital Fluorescence Microscopy. Cytometry. 36:267-278, 1999. Poon, SSS and PM Lansdorp Measurement of Telomere Length on Individual Chromosomes by Image Cytometry. Methods Cell Biol. 64:69-96, 2001. Potter, H and LN Geller Alzheimer's disease, Down's syndrome, and chromosome segregation. Lancet. 348(9019): 66, 1996.  175 Pujana, MA, M Nadal, M Guitart, L Armengol, M Gratacos and X Estivil. Human chromosome 15ql l-ql4 regions of rearrangements contain clusters of LCR15 duplicons. Eur. J. Hum. Genet. 10:26-35, 2002. Rancati, G, V Crispo, G Lucchini and S Piatti. Mad3/BubRl Phosphorylation During Spindle Checkpoint Activation Depends on both Polo and Aurora Kinases in Budding Yeast. Cell Cycle 4(7): 2005. Reish, O, R Gal, E Gaber, C Sher, T Bistritzer and A Amiel. Asynchronous replication of biallelically expressed loci: A new phenomenon in Turner syndrome. Genet in Med. 4(6):439-443, 2002. Reuss, ML, S Kolton and T Tharakan. Transvaginal ultrasonography in gynecologic office practice: assessment in 663 premenopausal women. Am. J. Obstet. Gynecol. 175(5): 11891194, 1996. Revenkova, E, M Eijpe, C Heyting, CA Hodges, PA Hunt, B Liebe, H Scherthan and R Jessberger. Cohesin SMCip is required for meiotic chromosome dynamics, sister chromatid cohesion and DNA recombination. Nature Cell Biology 6(6): 555-562, 2004. Risch, N , Z Stein, J Kline and D Warburton. The relationship between maternal age and chromosome size in autosomal trisomy. Am. J. Hum. Genet. 39(l):68-78,1986. Ritchie, RJ, M-G Mattei and M Lalande. A large polymorphic repeat in the pericentromeric region of human chromosome 15q contains three partial gene duplications. Hum. Mol. Genet. 7(8):1253-1260, 1998. Robinson, WP, A Bottani, X Yagang, J Balakrishman, F Binkert, M Machler, A Prader et al. Molecular, cytogenetic, and clinical investigations of Prader-Willi syndrome patients. Am. J. Hum. Genet. 49:1219-1234, 1991. Robinson, WP, R Spiegel and A A Schinzel. Deletion breakpoints associated with the PraderWilli and Angelman syndromes (15ql 1-ql 3) are not sites of high homologous recombination. Hum. Genet. 91(2):181-184, 1993. Robinson, WP and M Lalande, Sex-specific meiotic recombination in the PraderWilli/Angelman syndrome imprinted region. Human Molecular Genetics 4(5): 801-806, 1995. Robinson, WP, F Binkert, F Bernasconi, I Lorda-Sanchez, and EA Werder Molecular studies of chromosomal mosaicism: Relative frequency of chromosome gain or loss and possible role of cell selection. Am. J. Hum. Genet. 56: 444-451, 1995. Robinson, WP The Extent, Mechanism, and Consequences of Genetics Variation, for Recombination Rate Am. J. Hum. Genet. 59:1175-7783, 1996.  176 Robinson, WP, IJ Barrett, L Bernard, A Telenius, F Bernasconi, RD Wilson, RG Best, PN Howard-Peebles, S Langlois and DK Kalousek. Meiotic origin of trisomy in confined placental mosaicism is correlated with presence of fetal uniparental disomy, high levels trisomy in trophoblast, and increased risk of fetal intrauterine growth restriction. Am. J. Hum. Genet.60(4): 917-927, 1997. Robinson, WP, BD Kuchinka, F Bernasconi, MB Petersen, A Schulze, K Brondum-Nielsen, SL Christian, DH Ledbetter, A A Schinzel, B Horsthemke, S Schuffenhauer, RC Michaelis, S Langlois and TJ Hassold. Maternal meiosis I non-disjunction of chromosome 15: dependence of the maternal age effect on level of recombination. Human Molecular Genetics 7(6): 1011-1019, 1998. Robinson, WP, F Bernasconi, A Lau and DE McFadden. Frequency of Meiotic Trisomy Depends on Involved Chromosome and Mode of Ascertainment. American Journal of Medical Genetics 84: 34-42, 1999. Robinson, WP. Mechanisms leading to uniparental disomy and their clinical consequences. BioEssays. 22:452-459, 2000. Robinson, WP, SL Christian, BD Kuchinka, MS Penaherrera, S Das, S Schuffenhauer, S Malcolm, A A Schinzel, TJ Hassold and DH Ledbetter. Somatic segregation errors predominantly contribute to the gain or loss of a paternal chromosome leading to uniparental disomy for chromosome 15. Clin. Genet. 57:349-358, 2000. Robinson, WP, DE McFadden, and MD Stephenson. The Origin of Abnormalities in Recurrent Aneuploidy/Polyploidy. Am. J. Hum. Genet. 69(6): 1245-1254, 2001. Roeder, GS and JM Bailis. The pachytene checkpoint. Trends Genet. 16(():395-403, 2000. Rose A M , Baillie DL. A mutation in Caenorhabditis elegans which increases recombination frequency more than threefold. Nature 281:599-600 (1979). Rufer, N , W Dragowska, G Thornbury, E Roosnek and PM Lansdorp. Telomere length dynamics in human lymphocyte subpopulations measured by flow Cytometry. Nature Biotech. 16:743-747, 1998. Rufer, N , TH Brummendorf, S Kolvraa, C Bischoff, K Christensen, L Wadsworth, M Schulzer and PM Lansdorp. Telomere Fluorescence Measurements in Granulocytes and T Lymphocyte Subsets Point to a High Turnover of Hematopoietic Stem Cells and Memory T Cells in Early Childhood. J. Exp. Med. 190(2): 157-167, 1999. Rumble, B, R Retallack, C Hilbich, G Simms, G Multhaup, R Martins, A Hockey, P Montgomery, K Beyreuther and CL Masters. Amyloid A4 Protein and its precursor in Down's syndrome and Alzheimer's disease. N. Engl. J. Med. 320: 1446-1452, 1989.  177 Sandel, LL and V A Zakian. Loss of a Yeats Telomere: Arrest, Recovery, and Chromosome Loss. Cell 75:729-739, 1993. Saretzki, G and T Von Zglinicki Replicative Aging, Telomeres and Oxidative Stress. Ann. N.Y. Acad. Sci. 959:24-29, 2002. Schaetzlein, S, A Lucas-hahn, E Lemme, WA Kues, M Dorsch, MP Manns, H Niemann and K L Rudolph. Telomere length is reset during early mammalian embryogenesis. PNAS 101(21):8034-8038, 2004. Scheffer, GJ, FJ Broekmans, M Dorland, JD Habbema, CW Looman and ER te Velde. Antral follicle counts by transvaginal ultrasonography are related to age in women with proven natural fertility. Fertil. Steril. 72(5):845-851, 1999. Scherthan, H, S Weich, H Schwegler, C Heyting, M Harle and T Cremer. Centromere and Telomere Movements during Early Meiotic Prophase of Mouse and Man are Associated with the Onset of Chromosome Pairing. JCB 134(5): 1996: 1109-1125. Schon, EA, SH Kim, JC Ferreira, P Magalhaes, M Grace, D Warburton and SJ Gross. Chromosomal non-disjunction in human oocytes: is there a mitochondrial connection? Hum. Reprod. 15 (suppl. 2): 160-172, 2000. Schwartz, HS, GA Dahir and M G Butler. Telomere Reduction in Giant Cell Tumor of Bone and with Aging. Cancer Genet. Cytgenet. 71:132-138, 1993. Selig, S, K Okumura, DC Ward and H Cedar. Delineation of DNA replication time zones by fluorescence in situ hybridization. The EMBO Journal 11(3):1217-1225, 1992. Shaw, CJ, W Bi and JR Lupski. Genetic Proof of Unequal Croosovers in Reciprocal Deletion and Duplication of 17pl 1.2. Am. J. Hum. Genet. 71:1072-1081, 2002. Sherman, SL, MB Petersen, SB Freeman, J Hersey, D Pettay, L Taft, M Frantzen, M Mikkelsen and TJ Hassold. Non-disjunction of chromosome 21 in maternal meiosis I: evidence for a marernal age-dependent mechanism involving reduced recombiantion. Hum. Mol. Genet. 3:1529-1535, 1994. Shi, W, V Zakhartchenko and E Wolf. Epigenetic reprogramming in mammalian nuclear transfer. Differentiation 71(2):91-113, 2003. Shiels, PG, AJ Kind, K H Campbell, D Waddington, I Wilmut, A Colman and AE Schnike. Nature 399:316-317, 1999. Simchen, G and Y, Hugerat. What determines whether chromosomes segregate reductionally or equationally in meiosis? Bioessays 15(1): 1-8, 1993.  178 Singh, N, FAW Ebrahimi, AA Gimelbrant, AW Ensminger, MR Tackett, P Qi, J Gribnau and A Chess. Coordination of the random asynchronous replication of autosomal loci. Nature Genetics, 2003. Simon, I, T Tenzen, BE Reubinoff, D Hillman, JR McCarrey and H Cedar. Asynchronous replication of imprinted genes is established in the gametes and maintained during development. Nature. 401:929-932, 1999. Speed and Chandley Meiosis in the foetal mouse ovary. II. Oocyte development and agerelated aneuploidy. Does a production line exist? Chromosoma88:184-189, 1983. Sperling, K, J Pelz, R-D Wegner, A Dorries, A Gruters and M Mikkelsen. Significant increase in trisomy 21 in Berlin nine months after the accident: temporal correlation or causal relation? BMJ 309:158-161, 1994. Staessen, C, AM Maes, M Kirsch-Volders and C Susanne. Is there a predisposition for meiotic nondisjunction that may be detected by mitotic hyperploidy? Clinical Genetics. 24: 184-190, 1983. Stavropoulos, DJ, D Bick and DK Kalousek. Molecular cytogenetic detection of confined gonadal mosaicism in a conceptus with trisomy 16 placental mosaicism. Am. J. Hum. Gen. 63:1912-1914, 1998. Steinmetz, M, K Minard, S Horvath, J McNicholas, J Srelinger, C Wake, E Long, B Mach and L Hood. A molecular map of the immune response regionfromthe major histocompatibility complex of the mouse. Nature 300(5887):35-42, 1982. Stene, J. Detection of higher recurrence risk for age-dependent chromosome abnormalities with an application to trisomy GI. (Down's syndrome) Hum. Hered. 20(1):112-122, 1970. Stene, J, E Stene and M Mikkelsen. Risk for chromosome abnormality at amniocentesis following a child with a non-inherited chromosome aberration. A European Collaborative Study on Prenatal Diagnoses 1981. Prenat. Diagn. 4(SpecNo):81-95, 1984. Strachan, T and AP Read. Human Molecular Genetics 2. 2 Edition, 1999. nd  Strittmatter, WJ, AM Saunders, D Schmechel, M Pericak-Vance, J Enghild, GS Salvesen and AD Roses. Apolipoprotein E: high-avidity binding to P-amyloid and increased frequency of type 4 allele in late-onset familial Alzheimer disease. Proc. Natl. Acad. Sci. USA 90:19771981,1993. Subramanian, PS and AC Chinault. Replication timing properties of the human HPRT locus on active, inactive and reactivated X chromosomes. Som Cell Mol Genet. 23(2): 97-109, 1997.  179 Sugawara, S and K Mikamo. Maternal ageing and nondisjunction: a comparative study of two chromosomal techniques on the formation of univalents in first meiotic metaphase oocytes of the mouse. Chromosoma 93(4):321-325, 1986. Sullivan,BA, S Scwartz and HF Willard. Centromeres of Human Chromosomes. Env. Mol. Mut. 28:182-191, 1996. Sumner, A T . Scanning electron microscopy of mammalian chromosomes from prophase to telophase. Chromosoma 100(6):410-8, 1991. Sutcliffe, A G , K Saunders, R McLachlan, B Taylor, P Edwrads, G Grudzinskas, B Leiberman and S Thornton. A retrospective case-control study of developmental and other outcomes in a cohort of Australian children conceived by intracytoplasmic sperm injection compared with a similar group in the United Kingdom. Fertil. Steril. 79(3):512-516, 2003. Sym, M and GS Roeder. Crossover interference is abolished in the absence of a synaptonemal complex protein. Cell 79(2):283-292, 1994. Symington, LS, A Brown, S G Oliver, P Greenwell and T D Petes. Genetic analysis of a meiotic recombination hotspot on chromosome III of Saccharomyces cerevisiae. Genetics 128(4):717-727, 1991. Taneja, K L , E A Chavez, J Coull and P M Lansdorp. Multicolor Fluorescence In Situ Hybridization With Peptide Nucleic Acid Probes for Enumeration of Specific Chromosomes in Human Cells. Genes, Chroms. & Cancer 30:57-63, 2001. Tang, TT, SE Bickel, L M Yong and T L Orr-Weaver. Maintenance of sister-chromatid cohesion at the centromere by the Drosophila MEI-S332 protein. Genes Dev 12(24): 38433856, 1998. Tease, C and M A Hulten. Inter-sex variation in synaptonemal complex lengths largely determine the different recombination rates in male and female germ cells. Cytogenet. Genome Res. 107: 208-215, 2004. Thomas, NS, C E Brown, C Oley, S Healey and J A Crolla. Investigation of a cryptic interstitial duplication involving the Prader-Willi/Angleman syndrome critical region. Hum. Genet. 105(5):384-387, 1999. Thomas, NS, A R Collins, TJ Hassold and P A Hunt. A reinvestigation of non-disjunction resulting in 4 7 , X X Y males of paternal origin. Eur. J. Hum. Genet. 8(10):805-808, 2000. Tonami, Y , H Murakami, K Shirahige, M Nabanishi. A checkpoint control linking meiotic S phase and recombination initiation in fission yeast. PNAS 102(16): 5797-5801, 2005.  180 Torfs, CP and RE Christianson. The effect of maternal smoking and coffee consumption on the risk of having a recognized Down syndrome pregnancy. Am. J. Epidemiol. 152: 11851191,2000. Turnpenny, P and S Ellard. Emery's Elements of Medical Genetics. 12 Edition,2005. Van der Put, NM, F Gabreels, EMB Stevens, JAM Smeitink, FJM Trijbels, TKAB Eskes, LP van den Heuvel and HJ Blom. A Second Common Mutation in the Methylenetetrahydrofolate Reductase Gene: An Additional Risk Factor for Neural-Tube Defect? Am. J. Hum. Genet. 62: 1998. th  vanSteensel, B and T de Lange. Control of telomere length by the human telomeric protein TRF1. Nature 385(6618):740-743, 1997. Vaziri, H, F Schachter, I Uchida, L Wei, X Zhu, R Effros, D Cohen and CB Harley. Loss of Telomeric DNA during Aging of Normal and Trisomy 21 Human Lymphocytes. Am. J. Hum. Genet. 52:661-667, 1993. Vig, B and M Willcourt. Decondensation of pericentric heterochromatin alters the sequence of centromere separation in mouse cells. Chromosoma 107: 417-423, 1998. VonZglinicki, T, G Saretzki, W Docke and C Lotze. Mild Hyperoxia shortens telomeres and inhibits proliferation of fibroblasts: a model for senescence? Exp. Cell. Res. 220(1):186-193, 1995. Walker, MY and RS Hawley. Hanging on to your homolog: the roles of pairing, synapsis and recombination in the maintenance of homolog adhesion. Chromosoma 109:3-9, 2000. Wang, BB, CH Rubin and JD Williams. Mosaicism in chorionic villus sampling: an analysis of incidence and chromosomes involved in 2612 consecutive cases. Prenat Diagn 13: 179190, 1993. Warburton, D. Genetic Factors Influencing Aneuploidy Frequency. Basic Life Sciences 36:133-148, 1985. Warburton, D, J Kline, Z Stein, M Hutzler, A Chin and T Hassold. Does the Karyotype of a Spontaneous Abortion Predict the Karyotype of a Subsequent Abortion? - Evidence from 273 Women with Two Karyotyped Spontaneous Abortions. Am. J. Hum. Genet. 41:465-483, 1987. Warburton, D Chromosomal Causes of Fetal Death. Clin Obs Gyn 30 (2): 268-277, 1987. Warburton, D The effect of maternal age on thefrequencyof trisomy: change in meiosis or inutero selection? Prog. Clin. Biol. Res. 311; 165-181, 1989.  181 Warburton, D, L Dallaire, M Thangavelu, L Ross, B Levin and J Kline. Trisomy Recurrence: A Reconsideration Based on North American Data. Am. J. Hum. Genet. 75:2004 Warburton and Kinney (1996) Chromosomal differences in aneuploidy. Env mol mut 28: 37-247. Warren, AC, A Chakravarti, C Wong, SA Slaugenhaupt, SL Halloran, PC Watkins, C Metaxotou, SE Antonarakis. Evidence for Reduced Recombination on the Nondisjoined Chromosomes 21 in Down Syndrome. Science 237:652-655, 1987. Webb,T Inv dup(15) supernumerary marker chromosomes. J. Med. Genet. 31(8):585-594 (Review), 1994. Weissenbach, J, G Gyapay, C Dib, A Vignal, J Morissette, P Millasseau, G Vaysseix and M Lathrop. A second-generation linkage map of the human genome. Nature 359:794-801, 1992. Wolstenholme, J and RR Angell. Maternal age and trisomy - a unifying mechanism of formation. Chromosoma 109:435-438, 2000. Wu, TC and M Lichten. Meiosis-induced double-strand break sites determined by yeast chromatin structure. Science 263(5146):515-518, 1994. Yamamoto, A Endo and G Watanabe. Maternal age dependence of chromosome anomalies. Nature New Biol 241: 141-142, 1973. Yang, Q, SL Sherman, TJ Hassold et al. Risk factors for trisomy 21: maternal cigarette smoking and oral contraceptive use in a population-based case-control study. Genet. Med. 1: 80-88, 1999. Yeshaya, J, R Shalgi, M Shohat and L Avivi. Replication timing of the various FMR1 alleles detected by FISH: inferences regarding their transcriptional status. Hum. Genet. 102(1):614, 1998. Yin, H, E Baart, I Betzendahl, U Eichenlaub-Ritter. Diazepam induces meiotic delay, aneuploidy and predivision of homologues and chromatids in mammalian oocytes. Mutagenesis. 13(6):567-580, 1998. Young, LE, K Fernandes, TG McEvoy, SC Butterwith, CG Gutierrez, C Carolan, PJ Broadbent, JJ Robinson, I Wilmut and KD Sinclair. Epigenetic change in IGF2R is associated with fetal overgrowth after sheep embryo culture. Nature Genetics 27:153-154, 2001. Zaragoza, M, P Jacobs, P Rogan, S Sherman and T Hassold. Non-disjunction of human acrocentric chromosomes. Hum. Genet. 94:411-417, 1994.  182 Zetka M, Rose AM. Mutant rec-1 eliminates the meiotic pattern of crossing over in Caenorhabditis elegans. Genetics, 141:1339-1349, 1995. Zijlmans, JM, UM Martens, SSS Poon, AK Raap, HJ Tanke, RK Ward and PM Lansdorp. Telomeres in the mouse have large inter-chromosomal variations in the number of T2AG3 repeats. PNAS 94:7423-7428, 1997.  

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.831.1-0092280/manifest

Comment

Related Items