INSIGHTS INTO THE MATERNAL AGE EFFECT ON ANEUPLOIDY
By
ROSS ANTHONY ROWSEY
A dissertation submitted in partial fulfillment
of the requirements for the degree of
DOCTOR OF PHILOSOPHY
WASHINGTON STATE UNIVERSITY
School of Molecular Biosciences
MAY 2015
To the faculty of Washington State University:
The members and the Committee appointed to examine the dissertation of ROSS ANTHONY
ROWSEY find it satisfactory and recommend that it be accepted.
___________________________________
Terry Hassold, PhD, Chair
___________________________________
Patricia Hunt, PhD
___________________________________
Jon Oatley, PhD
___________________________________
William Davis, PhD
___________________________________
Wenfeng An, PhD
ii
ACKNOWLEDGEMENTS
The work in this document could not have been completed without the support of several
important individuals: Dr. Terry Hassold, for his invaluable mentorship and for sharing my enthusiasm for
the little things. Dr. Patricia Hunt, for being a second advisor, and for constantly challenging me to look at
the big picture, even when I was focused on the details. My committee, for guiding me along the path to
thinking like a scientist. My laboratory and my peers, for supporting me both in and out of the laboratory.
Lastly, my family, for believing in me unconditionally, and for pushing me to always strive for the best.
iii
INSIGHTS INTO THE MATERNAL AGE EFFECT ON ANEUPLOIDY
Abstract
by Ross Anthony Rowsey, Ph.D.
Washington State University
May 2015
Chair: Terry Hassold
Humans face extraordinary difficulties in reproduction, as over 10% of all pregnancies are
aneuploid. In most cases, aneuploidy causes complications in development; indeed, it is the leading
known cause of miscarriage and congenital birth defects in our species. In addition, high rates of
aneuploidy in pregnancy appear to be a human-specific condition, as levels of aneuploidy are orders of
magnitude lower in commonly studied model organisms. For decades, researchers have known that
increased maternal age goes hand in hand with increased risk of aneuploid pregnancy, a relationship
known as the maternal age effect. However, the cause of this age effect remains unclear, though
numerous hypotheses have been proposed. Our ability to analyze prophase stage oocytes has allowed
us to examine two of the most provocative of these models. In an examination of cells entering meiosis,
we did not observe aneuploidy, leading us to conclude that the errors that lead to aneuploidy occur at
some point in the meiotic process, and are not due to “predestination” events before meiosis begins. Our
focus then shifted to recombination in meiotic prophase, as abnormal recombination is the only known
molecular process linked to aneuploidy. We observed extraordinary variation in recombination rate
among individuals, but no apparent relationship with the timing of meiotic entry. Overall, our studies led
us to conclude that while recombination indeed plays a role in the genesis of the maternal age effect, it
remains only a part of the whole, and errors in later stages of meiosis also likely to contribute to the
maternal age effect.
iv
TABLE OF CONTENTS
Page
Acknowledgements ................................................................................................................................... iii
Abstract ..................................................................................................................................................... iv
List of Tables ............................................................................................................................................. vii
List of Figures ............................................................................................................................................ viii
CHAPTER I: Introduction ........................................................................................................................ 1
Meiosis: Segregating Chromosomes for the Next Generation .................................................... 3
Meiotic Prophase: Setting up Connections .................................................................................. 3
Pairing ............................................................................................................................. 4
Synapsis .......................................................................................................................... 4
Recombination ................................................................................................................ 5
Meiotic Divisions: Segregating the Chromosomes ...................................................................... 6
Male-Female Difference in Meiosis .............................................................................................. 6
Analyzing Meiosis ........................................................................................................................ 7
Cytological Approaches .................................................................................................. 7
Molecular analyses of gametes and early embryos ........................................................ 9
Genetic linkage analysis ................................................................................................. 10
When meiosis goes wrong: Aneuploidy ...................................................................................... 10
Aneuploidy in the Female: The Maternal Age Effect .................................................................. 12
Multiple Windows of Vulnerability ................................................................................................ 13
Pre-Meiosis: Off to a bad start? ..................................................................................... 13
Prophase Stage: Setting up for failure ............................................................................ 13
Arrest Stage: Waiting for something else to go wrong .................................................... 14
Meiotic Divisions: Where it all falls apart........................................................................ 15
Multiple Routes to Aneuploidy ........................................................................................ 15
Research Aims ............................................................................................................................. 16
References ................................................................................................................................... 17
v
CHAPTER II: Germline Mosaicism does not explain the Maternal Age Effect on Trisomy ............. 36
Abstract ........................................................................................................................................ 37
Introduction .................................................................................................................................. 38
Materials and Methods ................................................................................................................. 41
Study Population ............................................................................................................. 41
Processing ...................................................................................................................... 41
Immunofluorescence (IF) and Fluorescence In Situ Hybridization (FISH) ..................... 42
Scoring ............................................................................................................................ 43
Results and Discussion ................................................................................................................ 43
Rationale for the Analytic Approach ................................................................................ 43
A Pilot Study: FISH Studies of Oocytes in Trisomy 21 Fetuses .................................... 44
Analysis of Trisomy Mosaicism in Euploid Fetuses ........................................................ 45
Putting it into Perspective: Is Germ Cell Mosaicism an Important Source of Human
Aneuploidy ...................................................................................................................... 46
Acknowledgements ...................................................................................................................... 48
References ................................................................................................................................... 49
CHAPTER III: Examining Variation in Recombination Levels in the Human Female: A Test of
the Production-Line Hypothesis ............................................................................................... 60
Abstract ........................................................................................................................................ 61
Main Text ..................................................................................................................................... 62
Acknowledgements ...................................................................................................................... 66
References ................................................................................................................................... 67
CHAPTER IV: Summary and Future Directions ................................................................................... 79
Summary ...................................................................................................................................... 80
Future Directions .......................................................................................................................... 81
Prophase: Setting up oocytes for failure? ....................................................................... 81
Dictyate Arrest: Waiting to break down? ........................................................................ 85
Segregation: Where it all goes awry .............................................................................. 86
References ...................................................................................................................... 87
vi
LIST OF TABLES
CHAPTER I
Table I
Incidence of Aneuploidy at Various Stages of Development ................................... 22
Table II
Rate of Trisomy by Chromosome ............................................................................. 23
CHAPTER II
Table I
Detection of Chromosome 21 Signals by Stage of Prophase in Fetuses with
Trisomy 21 ................................................................................................................ 52
Table II
Number of FISH Signals per Chromosome in Leptotene Oocytes from Seven
Female Fetuses ........................................................................................................ 53
CHAPTER IV
Table I
Location of FISH Probes on Short Arm of Chromosome 16 .................................... 90
Table II
Average Length of Measured Regions ..................................................................... 91
vii
LIST OF FIGURES
CHAPTER I
Figure 1
Four Windows of Oocyte Development .................................................................... 24
Figure 2
Sub-Stages of Prophase and Synaptonemal Complex Appearance ....................... 26
Figure 3
The Synaptonemal Complex .................................................................................... 28
Figure 4
Resolution of Meiotic Double Strand Breaks ............................................................ 30
Figure 5
Cohesin Links Homologous Chromosome after Recombination.............................. 32
Figure 6
Abnormal Recombination Locations ......................................................................... 34
CHAPTER II
Figure 1
The expected number of FISH signals per chromosome is dependent on the
stage of meiotic prophase ........................................................................................ 54
Figure 2
Prophase stage oocytes from trisomy 21 fetuses seldom exhibit three
chromosome 21 signals ............................................................................................ 56
Figure 3
Putative trisomic cells may contain an artifactual FISH signal ................................. 58
CHAPTER III
Figure 1
Recombination in Human Oocytes ........................................................................... 69
Figure 2
Influence of Gestational Age on Genome-wide Recombination Levels ................... 71
Figure 3
Influence of Gestational Age on the Number of Crossovers on Individual
Chromosomes .......................................................................................................... 73
Figure 4
Influence of Maternal Age on Genome-wide Recombination Levels ....................... 75
Figure S1
Influence of Gestational Age on the Synaptonemal Complex Length of Individual
Chromosome ............................................................................................................ 77
CHAPTER IV
Figure 1
Use of BAC-FISH to mark specific genomic locations ............................................. 92
Figure 2
Local expansion of SC in regions containing crossover sites .................................. 94
viii
CHAPTER I
Introduction: The Maternal Age Effect on Aneuploidy
Introduction: The Maternal Age Effect on Aneuploidy
Humans face a unique challenge to reproduction: as a species, we have an extraordinary high
rate of miscarriage and birth defects, the leading cause of which is aneuploidy. Aneuploidy is the loss or
gain of a chromosome, with the most recognizable clinical example being trisomy 21 (Down
syndrome). These errors occur for all chromosomes at an astonishingly high frequency in humans, with
an estimated 10-30% of naturally occurring clinically recognized pregnancies being aneuploid [1]. The
high aneuploidy rate in humans appears to be a phenomenon unique to our species, as aneuploidy rates
in other eukaryotes are much lower. Mice, for example, have aneuploidy rates of around 1-2% of
pregnancies [2], and yeast have levels orders of magnitude below that, with aneuploidy occurring in fewer
than 0.0001% of tetrads [3]. Given the magnitude of the problem in our species and its devastating
clinical consequences, researchers have spent decades trying to understand the basis of aneuploidy in
humans.
One line of investigation has been to examine the origin of aneuploidy, and several key factors
have now been identified. First, aneuploidies primarily result from improper chromosome segregation
during the meiotic divisions [4]. Second, studies have demonstrated that aneuploidy originates in
oogenesis, with errors in maternal meiosis accounting for over 90% of all aneuploidies [1]. Third,
researchers have determined that the risk of having an aneuploid pregnancy increases with the age of the
mother (e.g. [1, 5, 6]). For example, in women under the age of 25 roughly 2% of clinically recognized
pregnancies are aneuploid, but in women over 40 years of age the risk increases by more than an order
of magnitude, to an estimated 35% of clinically recognized pregnancies [1].
Since most aneuploidies are maternally-derived, research has focused on developmental
windows in oogenesis that may make the oocyte vulnerable to errors that result in aneuploidy (Figure
1). Chronologically, the first of these windows is the pre-meiotic stage, at which time germ cells undergo
mitotic proliferation before beginning oogenesis. The next window is the meiotic prophase stage, during
which chromosomes undergo important processes such as synapsis and recombination. After prophase,
the oocyte enters a long arrest stage (from around birth until the time of ovulation), during which time the
cell remains relatively dormant. This arrest lasts potentially decades in the human, and as such is
another window in which errors may occur. Lastly, there is the possibility that errors occur as the
2
chromosomes divide at the first or second meiotic division. Overall, the production of oocytes in the
human female is a long and complicated process, with multiple steps along the way where processes
could go awry.
Given the clinical importance of human aneuploidy, researchers have spent decades trying to
elucidate the mechanisms by which these errors occur. Indeed, numerous hypotheses have been
proposed to explain the high rate of human aneuploidy, and most invoke events occurring in one of the
above vulnerable windows of oogenesis. The studies in this thesis examine models relating to the first
windows in which errors are hypothesized to occur, asking whether pre-meiotic divisions or early events
of prophase contribute to the age related increase in aneuploidy.
Meiosis: Segregating Chromosomes for the Next Generation
Meiosis is the process by which haploid gametes -- i.e. sperm and eggs -- are produced. To
achieve this, one round of DNA replication is followed by two chromosomal divisions, termed Meiosis I
and Meiosis II. Meiosis I involves the segregation of homologous chromosomes, and is termed the
reductional division because it reduces the number of chromosomes by half. Meiosis II segregates sister
chromatids and is an “equational”, mitotic-like division. Chromosome behavior during meiosis is highly
complex, with multiple processes that must occur for the chromosomes to segregate properly (Figure
1). Homologous chromosomes must properly pair together, synapse, and undergo recombination during
the prophase stage of Meiosis I, allowing for proper alignment and segregation during metaphase I and
anaphase I, respectively. Finally, the sisters again align and segregate during metaphase II and
anaphase II, leading to the production of haploid gametes.
Meiotic Prophase: Setting up Connections
After pre-meiotic replication, the chromosomes must undergo complex gymnastics as they
progress through meiotic prophase. Prophase is divided into five sub-stages (Figure 2) -- leptotene,
zygotene, pachytene, diplotene, and diakinesis -- during which time chromosomes pair, synapse, and
recombine. While each process serves a distinct function, they are highly interconnected, and a
disturbance in one typically disrupts the other two. All three of these processes must occur correctly to
ensure proper segregation of the chromosomes later in the meiotic divisions.
3
Pairing
To establish the connections between homologs to allow them to segregate properly, the first
step is to bring the homologous chromosomes into proximity with one another. This process is known as
pairing, and happens at the earliest stages of meiosis. The exact mechanism of pairing in mammalian
species is not fully understood, but a few key facts are known. First, homologous chromosomes are
brought into proximity by attaching telomeres to the nuclear membrane [7]. Second, initial interactions
between homologous chromosomes involve the establishment of an axis mediated by the cohesin
complexes that hold sister chromatids together. These interactions are stabilized via the formation of
double strand breaks (DSBs), which are important for both synapsis and recombination [8].
The pairing process begins with chromosome attachment to the nuclear membrane through the
SUN/KASH protein complex. SUN proteins are found within the inner nuclear membrane and associate
with telomeres [9-11], while KASH proteins are found on the outer membrane and are tethered to the
cytoskeleton [12]. Via this protein complex, the mechanical forces generated by the cytoskeleton are
transferred across the nuclear membrane to the chromosomes. These forces allow the chromosomes to
move around the nucleus to form a telomeric “bouquet”, thereby bringing homologous chromosomes into
close proximity [7].
Synapsis
While the pairing of chromosomes brings them into register, they must be intimately linked for
recombination to occur. The synaptonemal complex (SC) is the proteinaceous structure that fulfills this
role. The SC is comprised of three key components, two lateral/axial elements that line the homologous
chromosomes, and the central element that holds the two lateral/axial elements together (Figure 3). The
substages of prophase are defined by the appearance of the SC (Figure 2). In the earliest substage of
prophase, leptotene, the lateral/axial element of the SC, composed primarily of the structural protein
SYCP3, begins to attach to the homologous chromosomes [13]. As the cell enters zygotene, short
stretches of the lateral/axial elements pair and are fused together by the transverse filament of the central
element [14]. The transverse filament is composed of SYCP1 dimers, which attach to the lateral/axial
elements via their C-terminus [15]. Various other proteins bind to the N-terminus of SYCP1 to stabilize
4
the interactions between the homologs (e.g. [16, 17]). When all the chromosomes are fully synapsed, the
cell is said to be in pachytene.
Recombination
During prophase, the chromosomes exchange their genetic material through a process called
recombination. The process begins in leptotene with the formation of double-strand breaks (DSBs),
which are programmed breaks of the DNA catalyzed by the topoisomerase-like protein SPO11 (Figure
4A; [18, 19]). These DSBs trigger a DNA damage response, and the homologous repair pathway begins.
First, these DSBs are detected by the cell via local chromatin changes [20, 21], and the MRN complex
resects the 5’ strand, leaving 3’ overhangs of single stranded DNA (Figure 4B, [22]). This 3’ end then
forms a filament with proteins such as RAD51, DMC1, and a variety of accessory proteins, and this
filament is used to search for sequence homology [23, 24]. Specifically, DMC1 is the meiosis specific
molecule responsible for catalyzing invasion and joint molecule formation, while RAD51 serves as an
accessory binding protein [25]. Once that homologous site is found, this 3’ strand invades the
homologous sequence, forming a structure known as a D-loop (Figure 4C).
The 3’ end of the D-loop
structure is then used as an extension point, where DNA polymerase synthesizes DNA by using the
normal, intact strand as a template. The D-loop then proceeds down one of three pathways: either
continuing synthesis and eventually ligation, forming a double Holliday junction (Figure 4F; [26, 27];
undergoing D-loop collapse, resulting in the resolution of the DSB via synthesis dependent strand
annealing (SDSA) (Figure 4D; [28]); or being processed by MUS81, resulting in a so-called “noninterfering” crossover (Figure 4E; [29]). Very few sites are processed by MUS81, confirmed by knockout
studies in mice where removing MUS81 does not significantly disrupt recombination rates, as analyzed
via chiasma spreads [29]. Studies in yeast have demonstrated that SDSA accounts for around 26% of all
non-crossover sites [30], leaving the vast majority of D-loops to proceed through the first pathway, double
Holliday junction formation. Those D-loops that proceed to double Holliday junctions are stabilized by a
heterodimer consisting of two MutS mismatch repair proteins, MSH4 and MSH5, [26, 27], which can be
processed in different ways. That is, RNF212, a SUMO E3 ligase, is recruited to the MSH4/MSH5 sites to
stabilize them [31], and they can then be repaired by a second mismatch repair complex, including the
MutL homologs MLH1 and MLH3 (Figure 4H; [32]). However, other MSH4/MSH5 sites are processed by
5
HEI10, an ubiquitin ligase, which causes the removal of RNF212 and the resolution of the MSH4/MSH5
sites as non-crossover events (Figure 4G; [33, 34]).
While the recombination process is essential for genetic diversity, its primary function is to
physically link homologous chromosomes, which promotes proper segregation. This physical linkage is
not due to DNA, as the crossover site has already been fully repaired. However, the arms of sister
chromatids are held together by ring-like proteins called cohesins. Since a crossover event results in the
exchange of the arms of the chromosomes, this results in the arms of homologous chromosomes being
held together (Figure 5). This allows for the chromosomes to align properly as they prepare to undergo
the meiotic divisions.
Meiotic Divisions (Segregating the chromosomes)
To produce haploid gametes, meiotic cells undergo two rounds of chromosome segregation after
only a single round of replication. The first meiotic division is a specialized one known as the reductional
division, so-called because it reduces the number of chromosomes by half.
Spindle fibers form at both
poles of the cell, and attach to the chromosomes via the kinetochore, a protein complex that forms around
the centromere of the chromatids. In meiosis I, the kinetochore behaves uniquely, such that two sister
chromatids act as if they have a single kinetochore [35]. This allows for the homologous chromosome
pair to bi-orient along the metaphase plate and then segregate, sending one copy of each homologous
chromosome to opposite poles (Figure 1). This segregation is dependent on the cleavage of the physical
links between the homologous chromosomes that were formed when recombination occurred [36]. This
process produces two daughter cells, each with a single homolog composed of two sister chromatids. In
the second meiotic division, the kinetochores behave in a mitotic fashion, allowing for the separation of
sister chromatids. This results in four cells with haploid chromosome content.
Male-Female Differences in Meiosis
While the basic principles of meiosis are conserved throughout evolution, the specifics can be
drastically variable. There is no better example of this than the differences between male and female
meiosis in humans. Male-female differences can be summed up into three key components: end
products, timing, and fidelity. The first alteration, that of end products, is that both males and females
start with a single cell, but male meiosis results in four haploid gametes, which is exactly what one would
6
expect from one round of replication and two rounds of division. In contrast, the female undergoes
asymmetric divisions such that a single large haploid gamete, the egg, is produced, and the remaining
genetic material is discarded in the polar bodies.
Along with changes in how the cells undergo meiosis, the timing of meiosis between the sexes is
dissimilar. In the female, meiosis is a process that takes decades to complete. Before birth, all oocytes
begin meiosis, where they undergo prophase and arrest at diakinesis by the time of birth. This means
that all the oocytes that a female will ever have are formed before she is even born. Those oocytes then
remain arrested until they are ovulated, at least 10-15 years, and possibly as many as 45-50 years later.
Upon ovulation, the first meiotic division occurs and, if the egg is fertilized, it is followed by the second
division. This multi-year process to produce a single egg is in sharp contrast to male spermatogenesis,
since sperm are continuously produced after puberty begins. The time it takes to produce sperm is also
orders of magnitude faster than egg production; i.e., a matter of days rather than decades.
The final, and perhaps most striking, difference between male and female meiosis is the fidelity
with which gametes are produced. Males produce millions of gametes per day, whereas females only
produce one each cycle, so one might assume that the female has higher fidelity. However, this is not
the case, at least not in humans, as sperm have aneuploidy rates of approximately 1%, whereas eggs
have aneuploidy rates an order of magnitude higher, with some estimates as high as 50% [37]. These
observations highlight the fact that while meiosis is a conserved process, the sex-specific differences can
be remarkable.
Analyzing meiosis
There are three general approaches that have been used to examine meiosis in humans; i.e.,
cytogenetic analyses of oocytes or spermatocytes at various stages of meiosis or immediately following
meiosis; molecular analyses of oocytes or early pre-implantation embryos; and retrospective linkage
analyses of the products of conception, including analyses of embryos, fetuses or liveborn individuals.
Cytological approaches
Cytological methods are used at a variety of time points to assess the progress of meiosis. One
cytological method that can be used on the earliest stages of meiosis is surface spread preparations of
prophase stage oocytes or spermatocytes. With this method, cells are fixed onto a slide and then
7
immunofluorescence can be conducted on various meiotic proteins to assess meiotic
processes. Perhaps the most common set of proteins assayed are the synaptonemal complex proteins,
SYCP3 and SYCP1, as visualization of these allow for the staging of cells throughout meiosis. To assay
recombination, proteins such as RAD51, MSH4, and MLH1 allow analysis of double strand break
formation, repair intermediates, and fully formed recombination events, respectively. When combined
with fluorescence in situ hybridization (FISH), surface spread preparations can also provide chromosome
specific data on all of these processes, important because of the association of certain chromosomes
(e.g. 16, 21) with a high frequency of meiotic errors (Table 2). A major strength of this approach is the
large amount of data that can be collected from a single individual, as hundreds, if not thousands of
prophase stage cells can be obtained from fetal ovarian or testicular samples. Thus, this approach
provides a powerful technique to assess possible individual variation in meiotic progression and the
number of crossover events. . However, this technique has its limitations, most notably being the difficulty
in obtaining appropriate tissue samples. For studies of spermatogenesis, testicular material is required,
effectively limiting analyses to males attending infertility clinics. For female meiosis, the problem is even
more pronounced. Oocytes enter and complete prophase in utero, meaning that fetal ovarian tissue
samples are required to conduct the analyses. Our laboratory is in a fortunate situation to have
collaborations where we can obtain both types of material, enabling us to carry out the appropriate
analyses.
A second developmental timepoint at which oocytes and spermatocytes can be examined
cytologically is during diakinesis (e.g. [38, 39]), the last stage of meiotic prophase before the divisions. At
this time, the chromosomes have undergone recombination and are therefore held together by
chiasmata. If cells are spread on a slide using an air-dried technique, the exchanges between the
homologs can easily be visualized. Later stage oocytes can also be examined using conventional air
dried preparations. That is, examination of MII oocytes and accompanying polar bodies can be used to
assess chromosome segregation at the first meiotic division [40]. Again, however, these techniques
have their limitations. In addition to the difficulties in tissue acquisition, chromosome compaction at
diakinesis occurs just before ovulation in mammals and MII occurs at the time of fertilization, and
therefore collecting a large amount samples can be extremely difficult. Adding another layer of
8
complexity to this approach is that the analysis itself can be challenging. For example, at diakinesis
chromosomes are highly condensed and therefore the resolution is very low. With MII analysis, it is
usually possible to identify segregation errors, but the technique itself makes it relatively easy to “lose”
chromosomes, as they will spread too far from the cell. For this reason, only gains of chromosomes are
readily scored, as losses may be due to technical artifacts.
Finally, cytological analysis can be performed on mature sperm. Since sperm undergo numerous
post-meiotic modifications, analyzing the chromosome content requires specialized techniques. One of
the earliest accurate techniques for assessing the chromosome content of sperm was by the use of
hamster oocytes. Hamster oocytes are said to be “promiscuous”, allowing penetration by sperm of
various species; therefore human sperm can be activated, and then traditional chromosome spread
techniques can be used on the cells [41]. With advances in FISH technology, researchers then used
FISH on sperm nuclei to assess their chromosome content (e.g. [42]). Again, this technique has its
limitations, as only whole chromosome content can easily be detected, and not necessarily
rearrangements or deletions that may have clinical significance.
Molecular analyses of gametes and early embryos
In addition to cytological analyses, DNA based analyses can be used to observe the products of
meiosis. These techniques are used primarily after the meiotic divisions occur, either on polar bodies or
on cells removed from the blastocyst, usually in the preimplantation genetic diagnosis (PGD) setting (e.g.
[43]). The main techniques are array comparative genomic hybridization (aCGH), single nucleotide
polymorphism (SNP) arrays, and whole genome sequencing. aCGH is a method by which the DNA
content of the sample is hybridized to an array to detect variation (duplications or deletions) in copy
number of whole chromosomes or segments of chromosomes. SNP arrays also use hybridization to an
array, but with loci designed for specific polymorphisms. Similar to aCGH, this allows for copy number
assessment but in addition, it can also be used in conjunction with genotypic data from the parents to
specify the number and locations of recombination events. Finally, the decreasing cost of high
throughput sequencing is now making exon sequencing and even whole genome sequencing viable
options. By mapping the sequence reads to a reference genome, the gene or whole chromosome
9
content can be examined, and, similar to SNP arrays, the availability of parental samples can allow for the
determination of recombination sites.
Genetic linkage analysis
Last, and perhaps the most common method of observing meiosis, classical linkage analysis can
be used to examine the end products, the offspring of individuals (e.g. [44-47]). Specifically, by utilizing
linkage analysis to study the inheritance of alleles at polymorphic loci, it is possible to assess the number
and locations of recombination events that occurred in the oocytes and sperm that led to the offspring.
While a variety of polymorphic markers have been used over the years – e.g., blood group markers,
microsatellites and minisatellites – today the approach is largely restricted to the analysis of SNPs. By
assessing the amount of recombination between specific loci on the same chromosome, chromosomespecific and genome wide genetic maps can be constructed. While these studies are highly informative,
every offspring only provides information on a single meiotic event, meaning it is only possible to compare
multiple meioses in an individual if they have multiple offspring. Additionally, since recombination occurs
only between two of the four chromatids, only half of all crossovers are detectable, meaning we miss up
to 50% of all crossover events. Each technique has its merits and weaknesses, but our laboratory uses
cytological analysis of prophase stage oocytes and spermatocytes, since we are particularly interested in
the way in which homologs first interact with one another.
When Meiosis Goes Wrong: Aneuploidy
When the chromosomes fail to segregate properly during meiosis, aneuploidy or polyploidy
ensues, with aneuploidy being the more likely outcome. In newborns, 0.3% of all births are aneuploid, but
this truly represents the tip of the iceberg (Table 2; [48]). Stepping backward in development, roughly 4%
of all stillbirths are aneuploid, and over 35% of spontaneous abortions contain improper chromosome
number [48]. More recent technologies have allowed us to step even further back in gestation, and to
examine preimplantation embryos obtained in assisted reproduction technology (ART) settings. In a large
scale analysis of over 15,000 trophoectoderm biopsies, 40% of embryos were aneuploid, and almost half
of the aneuploid oocytes contained multiple chromosomal errors (2237 of 6168 aneuploid oocytes
contained aneuploidies for multiple chromosomes [43]). However, there is an obvious caveat to these
data; i.e., most preimplantation embryos involve infertile couples and include protocols in which the
10
gametes are manipulated, meaning that the data may not be truly representative of the “natural” human
aneuploidy rate.
Aneuploidy rates not only vary by the stage of pregnancy at which samples are ascertained, but
also by chromosome (Table 1; [48]). While nondisjunction for each human chromosome has been
reported, the distribution of chromosomes involved in trisomies does not fit a random distribution. For
example, trisomies 16, 18, 21, and 22 – four of the smaller chromosomes - account for over 50% of all
chromosomal abnormalities detected in spontaneous abortions [48]. Also varying is the relative viability
of each of these aneuploidies. Only four trisomies, (for chromosomes 13, 18, 21, and the sex
chromosomes) are compatible with live-birth, with +13 and +18 having very limited survival post-birth
[48]. Also interesting is the fact that trisomies appear much more commonly than monosomies, when a
priori, it might be expected that there would be an equal likelihood for either to occur [49]. Presumably,
this reflects differential survival between trisomies and monsomies; i.e., loss of a chromosome is likely
less well tolerated than is gain of a chromosome.
Theoretically, aneuploidy could result from errors in meiosis in either parent or from postfertilization mitotic errors, but two lines of evidence indicate that the vast majority of cases involve
maternal meiotic errors [1]. Firstly, in direct studies of mature gametes, there are drastically different
rates of aneuploidy. Karyotyping and FISH analysis on sperm display 1-4% aneuploidy [50, 51], while
similar studies on both naturally ovulated and stimulated oocytes have rates ranging from 20% to
upwards of 50% [51-53]. The second line of evidence supporting the maternal origin is DNA
polymorphism analyses on the origin of aneuploidies in preimplantation embryos, spontaneous abortions
or liveborn individuals. In preimplantation embryos that are aneuploid, upwards of 90% of the errors are
maternal in origin [54, 55]. However, there remains extreme chromosomal variation in the origins of these
aneuploidies. For example, Klinefelter syndrome (XXY), a male with an additional X chromosome, gets
the additional X chromosome from the sperm roughly 50% of the time, whereas trisomy 16 almost
exclusively originates in the oocyte [4]. These chromosome specific variations suggest that there may be
multiple routes that lead to aneuploidy, potentially explaining why different chromosomes have different
aneuploidy rates.
11
Aneuploidy in the Female: The Maternal Age Effect
While aneuploidy rates and the predominantly female meiotic origin are well documented, we
remain relatively ignorant of the underling mechanisms that give rise to aneuploidy. However, there are
two known correlates of the rates of aneuploidy: increasing maternal age, and abnormal recombination.
Abnormal recombination is the only molecular mechanism that has been linked to increases in
human aneuploidy. Crossovers are essential for maintaining the association between homologous
chromosomes until segregation, and studies have shown that trisomy is associated with one of three
types of abnormal recombination: crossovers placed too close to the centromere, crossovers placed too
close to the telomere, or failure to crossover (e.g. [1, 56-58]). Not every trisomy is linked with all three
types of abnormal recombination (Figure 6); e.g., trisomy 16 is primarily associated with distally placed
recombination sites [57], trisomy 18 with an absence of crossovers [59], and trisomy 21 with malpositioned crossovers or no crossovers at all [56, 58]. Nevertheless, each human trisomy that has been
appropriately studied has been associated with at least one category of abnormal recombination,
meaning that errors that occur during prophase are potentially setting up meiocytes to nondisjoin during
the meiotic divisions.
Since recombination is the only molecular mechanism associated with the origin of aneuploidy,
and there is extreme variability in aneuploidy rates between the sexes, recent studies in our laboratory
have focused on male-female differences in recombination. When examining sex-specific differences, it
becomes clear that female recombination is more variable than the male, and there are indeed more
errors that occur during female prophase [60]. When visualizing crossovers as they are forming at the
pachytene stage, oocytes have an exchange-less chromosome 2.6% of the time, over six times higher
than the 0.4% of sperm that fail to form an exchange on a chromosome [60]. This suggests that female
meiosis is inherently error prone, which contributes to the higher incidence of aneuploidy in the oocyte.
Along with abnormal recombination, the other known risk factor is that of advanced maternal
age. The increased risk of aneuploidy with increasing age, the maternal age effect, was first described by
Lionel Penrose, who observed increases in “mongolism” (now known as Down syndrome) with increased
maternal age [5]. Since the initial report by Penrose, multiple studies on clinically recognized pregnancies
have confirmed the drastic increase in trisomic pregnancies after the age of 35 [6, 61]. Women under
12
age 25 years have an approximate 2% rate of trisomic pregnancies, while in women 40 years of age and
older the percentage skyrockets to over 35% [1]. These figures understate the true impact of the problem
as they only account for clinically recognized pregnancies; while a large portion of chromosomally
abnormal conceptions are lost before clinical recognition.
Similar to abnormal recombination, advancing maternal age has different effects depending on
the chromosome involved. The larger chromosomes have an increase in trisomy rates after age 30, but
smaller chromosomes (13-22) have a nearly exponential increase in rates, a much more severe effect
than that seen on larger chromosomes [6, 61]. Perhaps the most unique chromosome-specific, agerelated aneuploidy rate is that of chromosome 16. Instead of remaining relatively low and spiking after
the 30’s, trisomy 16 displays a nearly linear increase [4, 57, 62]. These chromosome specific differences,
both in relation to maternal age and recombination, suggest that multiple mechanisms are involved in the
origin of human aneuploidy.
Multiple Windows of Vulnerability
While we do not know the specific mechanisms that lead to aneuploidy, it is interesting to
consider the developmental time points in the oocyte when errors might arise. Essentially, oocyte
development involves four windows in which errors could originate: Pre-Meiosis, Prophase, Meiotic
Arrest, and the Meiotic Divisions (Figure 1).
Pre-Meiosis: Off to a bad start?
In the production of gametes, the first time point at which errors could occur in the oocyte is
before meiosis begins, during the pre-meiotic stage. During this time, germ cells undergo migration and
mitotic proliferation, setting up a large pool of cells preparing to undergo meiosis. Very few theories have
addressed this window, as it is commonly thought that the errors leading to aneuploidy occur in meiosis.
However, a provocative hypothesis, known as the Oocyte Mosaicism Selection Model (OMSM), which will
be discussed further in Chapter II, suggests that errors in mitotic proliferation of the cells lead to the
maternal age effect [63, 64].
Prophase Stage: Setting up for failure
Events occurring at the earliest stages of meiosis are important for downstream chromosome
segregation. One of the most important events occurring during prophase is the establishment of
13
recombination sites, allowing the chromosomes to remain associated until the time of the meiotic
divisions. Because abnormal recombination is the only known molecular mechanism relating to
increased aneuploidy, there are numerous hypotheses that focus on events occurring at this
stage. Perhaps the most prominent of these hypotheses is known as the Production Line Hypothesis
[40], which will be discussed at more length in Chapter III. Simply put, the Production Line Hypothesis
suggests that the last oocytes produced have abnormally low recombination rates and are ovulated later
in life because they are produced later [40]. Other hypotheses, namely the Two-Hit Model, have
suggested that abnormal recombination at this stage may establish “at-risk” oocytes [58], which may be
perfectly normal if they are ovulated at an earlier age, but may undergo a second hit with advancing
maternal age, thereby resulting in increased aneuploidies at older maternal ages. In essence, the TwoHit Model links errors that occur in early prophase to errors that may occur later in meiosis, either during
the arrest stage or the divisions.
Arrest Stage: Waiting for something else to go wrong
The long dictyate arrest stage that oocytes go through has also been a focus of multiple
hypotheses on human female aneuploidy rates. Because the oocyte is dormant for multiple decades,
there is a long time span for potential errors to occur. One of the leading hypotheses relating to the
maternal age effect is that of protein degradation during this long arrest stage (e.g. [65]). The meiotic
cohesins are essential for allowing homologous chromosomes to remain associated until the first meiotic
division [66]. Cohesins are then cleaved along the chromosome arms, allowing homologous
chromosomes to separate, and then at MII are cleaved at the centromere, allowing sister chromatids to
separate. Since cohesins are laid down during pre-meiotic S-phase and are not cleaved until the
divisions, these proteins must function for decades in the human female. Studies in mice have shown
that these proteins do not turnover with time [67], and have also shown that cohesins laid down in premeiotic S-phase are sufficient for proper segregation during meiosis I [68]. Studies have also discovered
that decreased cohesin levels can increase the risk of aneuploidy and that cohesin levels appear to
decrease with time [69-71]. When combined, the findings from all of these studies suggest that loss of
cohesin with age may be an important mechanism of nondisjunction, linking maternal age and aneuploidy
rates. However, while these studies do provide a basic mechanism for the maternal age effect, they are
14
unlikely to be the only source of the maternal age effect, because they fail to address certain complexities
of the age curve, including the slight increase in aneuploidy in extremely young pregnancies, and the
chromosome-specific differences in rates of aneuploidy (e.g. chromosome 16 increases linearly,
chromosome 21 increases exponentially).
Meiotic Divisions: Where it all falls apart
While multiple hypotheses all provide mechanisms that partially explain events that may be
leading to the maternal age effect, the unifying factor of all of these hypotheses is that they lead to
nondisjunction during the meiotic divisions. Regardless if errors occurred earlier in the meiotic timeline,
aneuploidy simply cannot occur unless the chromosomes fail to segregate properly. Events before the
divisions may increase the risk for improper segregation, but the actual divisions separate the
chromosomes allowing aneuploidy to arise. Again, multiple hypotheses have been proposed as to what
could go wrong at this stage of meiosis, ranging from errors affected by hormone levels to errors in cell
cycle control (e.g. [72, 73]). One favored hypothesis is that of improper spindle assembly checkpoint
(SAC) function leading to improper segregation. The SAC monitors the formation of the spindle, the
complex that physically separates the chromosomes, and in most cells ensures that anaphase does not
occur until the chromosomes are ready to be segregated. However, studies in rodents have shown that
certain strains can activate the SAC before all chromosomes are attached to the spindle, effectively
triggering premature anaphase, greatly increasing the risk of aneuploidy [73].
Multiple Routes to Aneuploidy
While there are numerous hypotheses relating to the incidence of aneuploidy, each with merits
and flaws, to date there is no single hypothesis that can adequately explain the complexities of the
maternal age effect. Instead, it is likely that multiple hypotheses contribute to the incidence of
aneuploidy. For example, the link between abnormally placed distal recombination sites fits neatly with
the fact that cohesin proteins degrade over time. If the crossover site is extremely distal, there is a very
small portion of the chromosome arm that has cohesin holding together the homologous chromosomes,
so less cohesin needs to degrade before the homologs separate and segregate randomly at meiosis
I. Although there seems to be neat links between some of the hypotheses, there still remains a large
amount of mystery around the maternal age effect. Some hypotheses have not been properly vetted,
15
while others focus on observations in model organisms that do not have the same incidence of
aneuploidy that humans have. Because of these potentially incorrect hypotheses, the area around the
maternal age effect remains murky, as researchers are attempting to fit in puzzle pieces that do not
necessarily belong.
Research Aims
The studies in this thesis have focused on events occurring during or before meiotic
prophase. More specifically, these chapters focus on models of the maternal age effect. Experiments
were designed around our ability to examine human fetal oocytes, allowing us to directly observe
prophase stage oocytes.
Chapter II, published in the American Journal of Medical Genetics Part A (2013), focuses on a
hypothesis known as the Oocyte Mosaicism Selection Model (OMSM). The OMSM suggested that most
aneuploidy originated before meiosis, in stark contrast to most maternal age effect
hypotheses. Therefore, I analyzed chromosome content in early prophase stage oocytes to determine if
aneuploidies were already present, or if they were a product of events after meiosis began.
Chapter III, published in the American Journal of Human Genetics (2014), examined a long
standing hypothesis in the field known as the Production Line Model. The Production Line hypothesized
that aneuploidies are linked to the order in which oocytes undergo recombination, suggesting the first
formed oocytes are normal, while later formed oocytes have abnormal recombination and lead to the
increase in aneuploidy with age. To further examine this hypothesis, we have assayed human fetal
oocytes at various time-points to determine if recombination varies with timing of meiotic entry.
Chapter IV discusses the conclusions of the studies in this thesis, and proposes future
experiments to further understand the incidence of aneuploidy in the human female.
16
References
1.
Hassold, T. and P. Hunt, To err (meiotically) is human: the genesis of human aneuploidy. Nat Rev
Genet, 2001. 2(4): p. 280-91.
2.
Bond, D.J. and A.C. Chandley, Aneuploidy. Oxford monographs on medical genetics1983, Oxford
; New York: Oxford University Press. ix, 198 p., 8 p. of plates.
3.
Sora, S., G. Lucchini, and G.E. Magni, Meiotic Diploid Progeny and Meiotic Nondisjunction in
SACCHAROMYCES CEREVISIAE. Genetics, 1982. 101(1): p. 17-33.
4.
Hassold, T., et al., Human aneuploidy: incidence, origin, and etiology. Environ Mol Mutagen,
1996. 28(3): p. 167-75.
5.
Penrose, L.S., The relative effects of paternal and maternal age in mongolism. J Genet, 1933.
27(2): p. 219-223.
6.
Hassold, T., et al., Effect of maternal age on autosomal trisomies. Ann Hum Genet, 1980. 44(Pt
1): p. 29-36.
7.
Scherthan, H., A bouquet makes ends meet. Nat Rev Mol Cell Biol, 2001. 2(8): p. 621-7.
8.
Keeney, S. and M.J. Neale, Initiation of meiotic recombination by formation of DNA double-strand
breaks: mechanism and regulation. Biochem Soc Trans, 2006. 34(Pt 4): p. 523-5.
9.
Haque, F., et al., SUN1 interacts with nuclear lamin A and cytoplasmic nesprins to provide a
physical connection between the nuclear lamina and the cytoskeleton. Mol Cell Biol, 2006.
26(10): p. 3738-51.
10.
Malone, C.J., et al., UNC-84 localizes to the nuclear envelope and is required for nuclear
migration and anchoring during C. elegans development. Development, 1999. 126(14): p. 317181.
11.
Ding, X., et al., SUN1 is required for telomere attachment to nuclear envelope and
gametogenesis in mice. Dev Cell, 2007. 12(6): p. 863-72.
12.
Horn, H.F., et al., A mammalian KASH domain protein coupling meiotic chromosomes to the
cytoskeleton. J Cell Biol, 2013. 202(7): p. 1023-39.
13.
Yuan, L., et al., The murine SCP3 gene is required for synaptonemal complex assembly,
chromosome synapsis, and male fertility. Mol Cell, 2000. 5(1): p. 73-83.
14.
de Vries, F.A., et al., Mouse Sycp1 functions in synaptonemal complex assembly, meiotic
recombination, and XY body formation. Genes Dev, 2005. 19(11): p. 1376-89.
15.
Liu, J.G., et al., Localization of the N-terminus of SCP1 to the central element of the
synaptonemal complex and evidence for direct interactions between the N-termini of SCP1
molecules organized head-to-head. Exp Cell Res, 1996. 226(1): p. 11-9.
16.
Schramm, S., et al., A novel mouse synaptonemal complex protein is essential for loading of
central element proteins, recombination, and fertility. PLoS Genet, 2011. 7(5): p. e1002088.
17.
Costa, Y., et al., Two novel proteins recruited by synaptonemal complex protein 1 (SYCP1) are at
the centre of meiosis. J Cell Sci, 2005. 118(Pt 12): p. 2755-62.
17
18.
Baudat, F., et al., Chromosome synapsis defects and sexually dimorphic meiotic progression in
mice lacking Spo11. Mol Cell, 2000. 6(5): p. 989-98.
19.
Romanienko, P.J. and R.D. Camerini-Otero, The mouse Spo11 gene is required for meiotic
chromosome synapsis. Mol Cell, 2000. 6(5): p. 975-87.
20.
Fillingham, J., M.C. Keogh, and N.J. Krogan, GammaH2AX and its role in DNA double-strand
break repair. Biochem Cell Biol, 2006. 84(4): p. 568-77.
21.
Kinner, A., et al., Gamma-H2AX in recognition and signaling of DNA double-strand breaks in the
context of chromatin. Nucleic Acids Res, 2008. 36(17): p. 5678-94.
22.
Mimitou, E.P. and L.S. Symington, DNA end resection: many nucleases make light work. DNA
Repair (Amst), 2009. 8(9): p. 983-95.
23.
Benson, F.E., A. Stasiak, and S.C. West, Purification and characterization of the human Rad51
protein, an analogue of E. coli RecA. EMBO J, 1994. 13(23): p. 5764-71.
24.
Pezza, R.J., et al., Hop2/Mnd1 acts on two critical steps in Dmc1-promoted homologous pairing.
Genes Dev, 2007. 21(14): p. 1758-66.
25.
Cloud, V., et al., Rad51 is an accessory factor for Dmc1-mediated joint molecule formation during
meiosis. Science, 2012. 337(6099): p. 1222-5.
26.
Kneitz, B., et al., MutS homolog 4 localization to meiotic chromosomes is required for
chromosome pairing during meiosis in male and female mice. Genes Dev, 2000. 14(9): p. 108597.
27.
de Vries, S.S., et al., Mouse MutS-like protein Msh5 is required for proper chromosome synapsis
in male and female meiosis. Genes Dev, 1999. 13(5): p. 523-31.
28.
Paques, F. and J.E. Haber, Multiple pathways of recombination induced by double-strand breaks
in Saccharomyces cerevisiae. Microbiol Mol Biol Rev, 1999. 63(2): p. 349-404.
29.
Holloway, J.K., et al., MUS81 generates a subset of MLH1-MLH3-independent crossovers in
mammalian meiosis. PLoS Genet, 2008. 4(9): p. e1000186.
30.
McMahill, M.S., C.W. Sham, and D.K. Bishop, Synthesis-dependent strand annealing in meiosis.
PLoS Biol, 2007. 5(11): p. e299.
31.
Reynolds, A., et al., RNF212 is a dosage-sensitive regulator of crossing-over during mammalian
meiosis. Nat Genet, 2013. 45(3): p. 269-78.
32.
Baker, S.M., et al., Involvement of mouse Mlh1 in DNA mismatch repair and meiotic crossing
over. Nat Genet, 1996. 13(3): p. 336-42.
33.
Holloway, J.K., et al., Mammalian BLM helicase is critical for integrating multiple pathways of
meiotic recombination. J Cell Biol, 2010. 188(6): p. 779-89.
34.
Qiao, H., et al., Antagonistic roles of ubiquitin ligase HEI10 and SUMO ligase RNF212 regulate
meiotic recombination. Nat Genet, 2014. 46(2): p. 194-9.
35.
Toth, A., et al., Functional genomics identifies monopolin: a kinetochore protein required for
segregation of homologs during meiosis i. Cell, 2000. 103(7): p. 1155-68.
18
36.
Buonomo, S.B., et al., Disjunction of homologous chromosomes in meiosis I depends on
proteolytic cleavage of the meiotic cohesin Rec8 by separin. Cell, 2000. 103(3): p. 387-98.
37.
Nagaoka, S.I., T.J. Hassold, and P.A. Hunt, Human aneuploidy: mechanisms and new insights
into an age-old problem. Nat Rev Genet, 2012. 13(7): p. 493-504.
38.
Laurie, D.A. and M.A. Hulten, Further studies on chiasma distribution and interference in the
human male. Ann Hum Genet, 1985. 49(Pt 3): p. 203-14.
39.
Hulten, M., et al., The use and limitations of chiasma scoring with reference to human genetic
mapping. Cytogenet Cell Genet, 1978. 22(1-6): p. 37-58.
40.
Henderson, S.A. and R.G. Edwards, Chiasma frequency and maternal age in mammals. Nature,
1968. 218(5136): p. 22-8.
41.
Rudak, E., P.A. Jacobs, and R. Yanagimachi, Direct analysis of the chromosome constitution of
human spermatozoa. Nature, 1978. 274(5674): p. 911-3.
42.
Spriggs, E.L., A.W. Rademaker, and R.H. Martin, Aneuploidy in human sperm: the use of
multicolor FISH to test various theories of nondisjunction. Am J Hum Genet, 1996. 58(2): p. 35662.
43.
Franasiak, J.M., et al., The nature of aneuploidy with increasing age of the female partner: a
review of 15,169 consecutive trophectoderm biopsies evaluated with comprehensive
chromosomal screening. Fertil Steril, 2014. 101(3): p. 656-663 e1.
44.
Jorgenson, E., et al., Ethnicity and human genetic linkage maps. Am J Hum Genet, 2005. 76(2):
p. 276-90.
45.
Kong, A., et al., Recombination rate and reproductive success in humans. Nat Genet, 2004.
36(11): p. 1203-6.
46.
Kong, A., et al., A high-resolution recombination map of the human genome. Nat Genet, 2002.
31(3): p. 241-7.
47.
Broman, K.W., et al., Comprehensive human genetic maps: individual and sex-specific variation
in recombination. Am J Hum Genet, 1998. 63(3): p. 861-9.
48.
Hassold, T.J. and P.A. Jacobs, Trisomy in man. Annu Rev Genet, 1984. 18: p. 69-97.
49.
Jacobs, P.A., The chromosome complement of human gametes. Oxf Rev Reprod Biol, 1992. 14:
p. 47-72.
50.
Templado, C., F. Vidal, and A. Estop, Aneuploidy in human spermatozoa. Cytogenet Genome
Res, 2011. 133(2-4): p. 91-9.
51.
Martin, R.H., E. Ko, and A. Rademaker, Distribution of aneuploidy in human gametes:
comparison between human sperm and oocytes. Am J Med Genet, 1991. 39(3): p. 321-31.
52.
Kuliev, A., et al., Meiosis errors in over 20,000 oocytes studied in the practice of preimplantation
aneuploidy testing. Reprod Biomed Online, 2011. 22(1): p. 2-8.
53.
Pellestor, F., et al., The occurrence of aneuploidy in human: lessons from the cytogenetic studies
of human oocytes. Eur J Med Genet, 2006. 49(2): p. 103-16.
19
54.
Sills, E.S., et al., Determining parental origin of embryo aneuploidy: analysis of genetic error
observed in 305 embryos derived from anonymous donor oocyte IVF cycles. Mol Cytogenet,
2014. 7(1): p. 68.
55.
Rabinowitz, M., et al., Origins and rates of aneuploidy in human blastomeres. Fertil Steril, 2012.
97(2): p. 395-401.
56.
Lamb, N.E., et al., Susceptible chiasmate configurations of chromosome 21 predispose to nondisjunction in both maternal meiosis I and meiosis II. Nat Genet, 1996. 14(4): p. 400-5.
57.
Hassold, T., et al., Recombination and maternal age-dependent nondisjunction: molecular studies
of trisomy 16. Am J Hum Genet, 1995. 57(4): p. 867-74.
58.
Hassold, T. and S. Sherman, Down syndrome: genetic recombination and the origin of the extra
chromosome 21. Clin Genet, 2000. 57(2): p. 95-100.
59.
Fisher, J.M., et al., Molecular studies of trisomy 18. Am J Hum Genet, 1993. 52(6): p. 1139-44.
60.
Gruhn, J.R., et al., Cytological studies of human meiosis: sex-specific differences in
recombination originate at, or prior to, establishment of double-strand breaks. PLoS One, 2013.
8(12): p. e85075.
61.
Risch, N., et al., The relationship between maternal age and chromosome size in autosomal
trisomy. Am J Hum Genet, 1986. 39(1): p. 68-78.
62.
Hassold, T.J., et al., Molecular studies of non-disjunction in trisomy 16. J Med Genet, 1991. 28(3):
p. 159-62.
63.
Hulten, M.A., et al., On the origin of trisomy 21 Down syndrome. Mol Cytogenet, 2008. 1: p. 21.
64.
Hulten, M.A., et al., On the origin of the maternal age effect in trisomy 21 Down syndrome: the
Oocyte Mosaicism Selection model. Reproduction, 2010. 139(1): p. 1-9.
65.
Jessberger, R., Age-related aneuploidy through cohesion exhaustion. EMBO Rep, 2012. 13(6): p.
539-46.
66.
Uhlmann, F. and K. Nasmyth, Cohesion between sister chromatids must be established during
DNA replication. Curr Biol, 1998. 8(20): p. 1095-101.
67.
Tachibana-Konwalski, K., et al., Rec8-containing cohesin maintains bivalents without turnover
during the growing phase of mouse oocytes. Genes Dev, 2010. 24(22): p. 2505-16.
68.
Revenkova, E., et al., Oocyte cohesin expression restricted to predictyate stages provides full
fertility and prevents aneuploidy. Curr Biol, 2010. 20(17): p. 1529-33.
69.
Hodges, C.A., et al., SMC1beta-deficient female mice provide evidence that cohesins are a
missing link in age-related nondisjunction. Nat Genet, 2005. 37(12): p. 1351-5.
70.
Chiang, T., et al., Evidence that weakened centromere cohesion is a leading cause of age-related
aneuploidy in oocytes. Curr Biol, 2010. 20(17): p. 1522-8.
71.
Lister, L.M., et al., Age-related meiotic segregation errors in mammalian oocytes are preceded by
depletion of cohesin and Sgo2. Curr Biol, 2010. 20(17): p. 1511-21.
20
72.
Steuerwald, N., et al., Association between spindle assembly checkpoint expression and
maternal age in human oocytes. Mol Hum Reprod, 2001. 7(1): p. 49-55.
73.
Nagaoka, S.I., et al., Oocyte-specific differences in cell-cycle control create an innate
susceptibility to meiotic errors. Curr Biol, 2011. 21(8): p. 651-7.
21
Table 1: Incidence of Aneuploidy at Various Stages of Development
Stage of
Development
Aneuploidy Rate
Preimplantation
Embryos
~40% [43]
Spontaneous
Abortions
~25% [48]
22
Stillbirths
Liveborns
4% [48]
0.3% [48]
Table 2: Rate of Trisomy by Chromosome (data from [48])
Trisomic Chromosome
13
16
18
21
22
Sex Chromosomes
% of Clinically Recognized
Pregnancies
0.18%
1.13%
0.18%
0.45%
0.40%
0.17%
23
Figure 1 - Four Windows of Oocyte Development
Schematic of chromosome processes during meiosis. (A) Pre-meiotic stage, when chromosomes undergo
replication in preparation for meiosis. (B) Meiotic prophase, when chromosomes pair, synapse, and form
recombination sites. (C) Meiotic arrest stage, when oocyte remains dormant until the time of ovulation (D)
Meiotic divisions, when homologous chromosomes (Meiosis I) and sister chromatids (Meiosis II)
separate, resulting in the egg and associated polar bodies.
24
Figure 1
25
Figure 2 - Sub-stages of Prophase and Synaptonemal Complex Appearance
Meiotic prophase is divided into 5 substages. (A) In Leptotene, the SC (red) begins to form along the
chromosomes. (B) Zygotene, at which time synapsis begins; i.e., regions of the chromosomes are held
together by the SC, but the full complex is not yet fully formed. (C) In Pachytene, the SC is fully formed
between the homologs, and recombination occurs. (D) Diplotene begins the disassembly of the SC, with
the chromosomes remaining tethered at sites of recombination. (E) Finally, at Diakinesis the SC has fully
disassembled, and the recombination sites remain as the only links between the homologs.
26
Figure 2
27
Figure 3 - The Synaptonemal Complex
Schematic of the SC axial/lateral elements (in red), which line each homolog (in blue), and the central
element (in green), which links together the two axial/lateral elements.
28
Figure 3
29
Figure 4 - Resolution of Meiotic Double Strand Breaks
(A) The meiotic recombination pathway begins with SPO11 forming a DSB in one chromosome. (B) The
DSB is then resected by the MRN complex, leaving a 3’ overhang, and (C) the single stranded DNA forms
a filament with DMC1 and RAD51, invading the homologous site and forming a d-loop. (D) if the d-loop
collapses, it is repaired as a synthesis depended strand annealing event (SDSA), or (E) it may be
repaired via MUS81 into a crossover. (F) Most of the time, the d-loop with form a double Holliday
Junction, which then is resolved either (G) as a noncrossover event or (H) as a crossover event.
30
Figure 4
31
Figure 5 - Cohesin links homologous chromosomes after recombination
Cohesin forms ring-like structures (orange) that hold together sister chromatids. After recombination,
cohesin distal to the recombination sites links the homologous chromosomes together, as the arms have
been exchanged.
32
Figure 5
33
Figure 6 - Abnormal Recombination Locations
Recombination placement is deemed to be abnormal when it falls either (A) too close to the centromere,
(B) too close to the telomere, or (C) fails to form completely.
34
Figure 6
35
CHAPTER II
Germline Mosaicism Does Not Explain the Maternal Age Effect on Trisomy
Ross Rowsey, Anna Kashevarova, Brenda Murdoch, Carrie Dickenson, Tracey Woodruff, Edith Cheng,
Patricia Hunt, Terry Hassold
Rowsey R, Kashevarova A, Murdoch B, Dickenson C, Woodruff T, Cheng E, Hunt P, Hassold T. 2013.
Germline mosaicism does not explain the maternal age effect on trisomy. Am J Med Genet A
161(10):2495-2503
ABSTRACT
A variety of hypotheses have been proposed to explain the association between trisomy and increasing
maternal age in humans, virtually all of which assume that the underlying mechanisms involve meiotic
errors. However, recently Hultén and colleagues [1] proposed a provocative model – the Oocyte
Mosaicism Selection Model (OMSM) – that links age-dependent trisomy 21 to pre-meiotic errors in the
ovary. Specifically, they propose that nondisjunctional events occur in a proportion of germ cells as they
mitotically proliferate, resulting in mosaicism for trisomy 21. Assuming that the presence of an additional
chromosome 21 delays meiotic progression, these cells would be ovulated later in reproductive life,
resulting in an age-dependent increase in aneuploid eggs. Because this model has important clinical
implications, we initiated studies to test it. We first analyzed oocytes from two trisomy 21 fetuses,
combining immunostaining with FISH to determine the likelihood of detecting the additional chromosome
21 at different stages of meiosis. The detection of trisomy was enhanced during the earliest stage of
prophase (leptotene), before homologs synapsed. Accordingly, in subsequent studies we examined the
chromosome content of leptotene oocytes in seven second trimester female fetuses, analyzing three
chromosomes commonly associated with human trisomies (i.e. 13, 16, and 21). In contrast to the
prediction of the OMSM, we found no evidence of trisomy mosaicism for any chromosome. We conclude
that errors in pre-meiotic germ cells are not a major contributor to human aneuploidy and do not provide
an explanation for the age-related increase in trisomic conceptions.
37
INTRODUCTION
Aneuploidy (monosomy or trisomy) occurs with extraordinary frequency in humans, affecting 1030% of fertilized eggs and typically resulting in either very early pregnancy loss, miscarriage, or live-born
infants with developmental disabilities [2]. Since the vast majority of aneuploidies derive from errors in
oogenesis, considerable attention has been directed at identifying maternal risk factors. Although a
number of factors have been suggested, the most convincingly demonstrated is increasing age of the
mother [3]. The maternal age effect is remarkable, with women in their twenties having only a 2-3%
chance of a clinically recognized trisomic pregnancy, while women in their forties have a risk exceeding
30% [4]. Recent changes in reproductive patterns have heightened the importance of this relationship.
That is, in developed countries the mean maternal age at delivery has increased remarkably over the past
2-3 decades; indeed, in the United States the value in 2010 was 27.7 years, nearly three years higher
than the 25.0 mean for 1980 [5]. This would be expected to increase the incidence of aneuploidy among
clinically recognized pregnancies, and the increasing prevalence of trisomies 13, 18 and 21 in
pregnancies in Europe suggests that this is the case [6].
Although the relationship between age and aneuploidy is well documented, the underlying basis
of the effect remains obscure. Over the years, several different hypotheses have been proposed, most of
which can be placed into one of four broad categories based on the developmental stage of oogenesis at
which the precipitating event is postulated to occur.
The first category invokes events that occur in the fetal ovary when the oocyte enters meiotic
prophase. In humans, oocytes initiate meiosis at about 10-13 weeks gestation. Homologous
chromosomes pair, synapse, and undergo recombination and the cell then enters a period of dictyate
arrest that lasts from late in gestation until just prior to ovulation in the adult ovary [7]. Likely the most
famous model invoking events at this developmental stage is the production line model originally
proposed by Henderson and Edwards in 1968 [8]. In an analysis of female mice, they noted an increase
in the frequency of univalents in metaphase I oocytes from older females and suggested that age-related
decreases in recombination might be responsible for the elevated level of aneuploidy in older females.
However, since recombination occurs in the fetal ovary in mammals, they were forced to propose a two-
38
step model, i.e., that the first oocytes to enter meiosis are the first to be ovulated, and oocytes that enter
meiosis later form fewer chiasmata [8]. Consequently, oocytes ovulated at the end of the reproductive
lifespan would have the fewest chiasmata and be most susceptible to nondisjunction. Subsequent
evidence has supported the first tenant of this hypothesis, i.e., that the first oocytes to initiate meiosis are
ovulated first [9]. However, genetic linkage data from studies of humans provide no evidence of an agerelated decrease in recombination levels [10-12]; thus the relevance of the production line model to agerelated human aneuploidy is unclear.
A second category of hypotheses has focused on the long dictyate arrest that lasts from late in
gestation until just prior to ovulation. Some researchers have suggested that, because the maternal age
curve mimics that of an infectious process, a pathogenic agent may be involved in the genesis of
abnormal oocytes [13]. Others have suggested that the long period of meiotic arrest allows for the
accumulation of DNA damage, which has been shown to cause aneuploidy in model organisms [14].
However, perhaps the most widely supported of these “meiotic arrest” models proposes that the crucial
event is degradation of cohesin proteins [15-21]. In meiosis, the cohesion complex has dual functions:
ensuring cohesion between sister chromatids, and maintaining inter-homolog associations distal to the
sites of crossovers. Both activities are crucial in orchestrating the segregation of homologs at the first
meiotic division. It is thought that there is little or no turnover of the proteins of the cohesin complex, and
age-related degradation of cohesion established during fetal development has been postulated to lead to
the premature separation of homologs and/or sister chromatids, resulting in aneuploidy [16, 17, 20, 21].
A third group of hypotheses invoke events that occur at or around the time of the meiotic cell
divisions. For example, age related abnormalities may involve internal factors such as hormonal
variation, including changes in the levels of FSH [22] or LH [23] or the ability of the oocyte to respond to
those signals [15], or age-related changes in the microcirculation of the ovary that disrupt formation of the
meiotic spindle [24]. Other hypotheses relate to the ability of the oocyte to monitor errors. Cell cycle
checkpoint proteins such as MAD2 and BUB1 delay the cell cycle until bipolar attachment of homologs is
complete, and if protein levels decrease with age, normal functioning of the spindle assembly checkpoint
may be impaired [25]. Recent studies have shown that cell cycle control mechanisms differ between
39
mammalian oocytes and spermatocytes, operating in a less stringent manner in the former [26]. Thus, it
may be that errors occur at a similar frequency in male and female gametes but that the inability to sense
and eliminate them in eggs results in a higher frequency of maternally-derived aneuploidies.
Finally, a fourth category of hypotheses suggest that aneuploidy may be attributable to a
combination of “hits” that occur at different stages in meiosis. For example, Sherman and colleagues
have suggested that abnormalities in meiotic recombination in the fetal stages of oogenesis render
oocytes susceptible to nondisjunction, and that a second “hit” that occurs in the aging ovary increases the
likelihood that such oocytes will, indeed, nondisjoin [27]. Consistent with this idea, the incidence of cases
of trisomy 21 attributable to achiasmate chromosomes 21 appears to increase with maternal age [28, 29].
While the above models differ in the way that they explain the maternal age effect, they are united
by a common assumption, namely, that the age effect is attributable to events occurring during meiosis.
Recently, however, Hultén and colleagues proposed a new type of model – the Oocyte Mosaicism
Selection Model (OMSM) – that casts doubt on the assumption that the age effect is meiotic in origin [30].
In a study of surface spread human fetal oocytes from eight female fetuses, they performed FISH for two
subtelomeric loci on chromosome 21. They found that roughly 1 in 200 cells contained three
chromosome 21 FISH signals, suggesting that pre-meiotic mitotic nondisjunction generates a surprisingly
high proportion of primary oocytes that carry an extra chromosome 21. If these trisomic oocytes are
equally as likely to be ovulated as chromosomally normal oocytes, these observations suggest that the
vast majority of cases of trisomy 21 derive from mitotic, not meiotic, errors. Further, if as suggested by
Hultén and colleagues, the progression of trisomic oocytes through meiotic prophase is delayed, resulting
in their ovulation later in reproductive life [30], one does not have to invoke meiotic defects in the genesis
of the maternal age effect.
The implications of the OMSM are wide-ranging. Importantly, it suggests that further analyses of
meiosis and meiotic abnormalities are unlikely to provide important insights on the etiology of maternallyderived human aneuploidy. Further, it implies that therapeutic strategies to prevent the occurrence of
trisomy are doomed to failure, since intervention would have to occur in the first trimester of pregnancy.
40
Thus, if confirmed, the OMSM fundamentally changes our understanding of the origin of human
aneuploidy, and re-directs our research efforts.
Unfortunately, only one subsequent study has attempted to repeat the study. In an analysis of
eight samples and 51,146 cells, Morris and colleagues were unable to replicate the original observations
[31]. However, neither the original study nor the study of Morris and colleagues analyzed different stages
of meiotic prophase, and Hulten et al [30] did not analyze chromosomes other than 21. Thus, we decided
to re-examine the OMSM, analyzing levels of clinically relevant trisomies 13, 16 and 21 in oocytes at the
earliest stages of prophase. We found no evidence of trisomy mosaicism for any of the three
chromosomes, suggesting that factors other than pre-meiotic nondisjunction are responsible for the agerelated increase in human trisomies.
MATERIALS AND METHODS
Study Population
Material for this study consisted of prophase stage oocytes from ovarian samples of nine second
trimester female fetuses, two with trisomy 21 and the other seven presumed to be chromosomally normal.
Ovaries were collected either at the University of Washington Medical Center in Seattle, Washington or
the San Francisco General Hospital’s Women’s Options Center in San Francisco, California from
therapeutic or elective terminations of pregnancies. Immediately following the surgical procedures, whole
ovarian tissue samples were shipped by overnight delivery to Washington State University (WSU) for
processing. All procedures were approved by the University of California, San Francisco’s Human
Research Protection Program or the University of Washington Institutional Review Board, and by the
WSU Institutional Review Board, and informed consent was obtained from all study participants.
Processing
Upon receipt at WSU, ovaries were processed using a standard surface spreading technique
[32]. Briefly, ovaries were cleaned of excess tissue, placed in a watch glass, covered in hypo-extraction
buffer (600 mM TRIS, 500 mM sucrose, 500 mM EDTA, 500 mM DDT, 170 mM sodium citrate, and 100
mM PMSF), incubated at room temperature for about 45 minutes, suspended in 100 mM sucrose and
41
macerated with 26 gauge needles. Following maceration, 10 µL aliquots of cell suspension were spread
on slides coated with 2% PFA and placed in a humid chamber overnight. The following morning, slides
0
were air dried, washed in a 0.04% Photoflo™ solution for 2 minutes, and stored at 4 C prior to staining.
Immunofluorescence (IF) and Fluorescence In Situ Hybridization (FISH)
Slides were immunostained using similar methodology to that published previously [33]. Slides
were pre-incubated for 60 minutes in 1X antibody dilution buffer (ADB) composed of 10 ml normal donkey
serum (Jackson ImmunoResearch), 3 g BSA (Sigma-Aldrich), 50 µL Triton X-100 (Alfa Aesar), and 990
ml PBS. Sixty µL of 1X ADB containing kinetochore-associated CREST antisera (1:500 dilution, Fisher,
0
rabbit anti-human) was applied and the slides were incubated overnight at 37 C. Sixty µL of 1X ADB
containing an antibody against the synaptonemal complex protein SYCP3 (1:150 dilution; Novus, rabbit
0
anti-human) was then added to the slides and incubated for two hours at 37 C. Slides were then washed
twice in 1X ADB for 20 minutes, 60 µL of ADB containing ADAH (1:100 dilution, Jackson
0
ImmunoResearch) was added, and slides were incubated overnight at 37 C. The following morning 60
µL of 1X ADB containing RDAR (1:100 dilution, Jackson ImmunoResearch) was added and slides
0
incubated for 45 minutes at 37 C. Slides were washed twice in 1X PBS for 30 minutes, twice more in 1X
PBS for 1 minute, and coverslipped with 40 µL Prolong Gold Antifade Reagent (Invitrogen) and stored at
0
4 C. Images were captured on a Zeiss fluorescence microscope and coordinates recorded or slides were
scanned using the ASI Scanview Case Data Manager.
After initial examination of IF images, we used chromosome-specific FISH to identify individual
chromosomes. For the two fetuses with trisomy 21, we used a Vysis LS1 21 probe. For the other seven
fetuses, we used Cytocell Aquarius Satellite Enumeration point probes for chromosome 16 (Green) and a
probe detecting both chromosomes 13 and 21 (Red Dual Probe). Previously immunostained slides were
dehydrated in an ethanol series (75%, 90%, and 100%) at room temperature. Slides and probe mixtures
0
(6 µL each probe and 8 µL hybridization solution) were both pre-warmed at 37 C for 5 minutes and the
0
probe was applied to the slide and coverslipped. Preparations were denatured at 75 C for 5 minutes and
0
incubated overnight at 37 C in a humid chamber. The following morning, slides were washed in 0.4X
0
SSC at 72 C for 30 seconds, washed in 2X SSC/0.5% Tween-20 at room temperature for 2 minutes and
0
then coverslipped with 40 µL prolong Gold Antifade Reagent with DAPI (Invitrogen) and stored at 4 C
42
until analysis. Slides were evaluated in the same manner as after IF, on either the Zeiss fluorescence
microscope or scanned using the ASI Scanview Case Data Manager software. Cells from IF were
located, captured for FISH signals and scored.
Scoring
Cells were analyzed for the number of FISH signals displayed. To be classified as a signal, a
focus had to be punctuate, round to elliptical in shape and regular in appearance. For the two trisomy 21
fetuses, we simply counted the number of chromosome 21 FISH signals. For the other seven,
presumably euploid, fetuses, cells that contained an abnormal number of FISH signals were overlaid with
their corresponding IF images to determine whether the FISH signals co-localized with centromeres
(detected by CREST). FISH signals that localized to the centromere-specific CREST signal were
classified as bona fide, while those that did not localize to a CREST signal were classified as false
positive signals.
RESULTS AND DISCUSSION
Rationale for the analytic approach
Conceptually, the use of FISH to detect trisomy in mitotic cells is a simple exercise. A
chromosome-specific probe localizes to each copy of the target chromosome, thereby enabling detection
of the number of chromosomes per nucleus. The reliability of the approach depends on a number of
factors, including the quality of the cytological preparation, the type of cell being analyzed and the
hybridization efficiency; nevertheless under optimal conditions the analysis is straightforward. However,
the same does not apply to meiotic cells. Because prophase involves the pairing of homologous
chromosomes and, ultimately, their intimate association, the number of expected signals for a
chromosome pair depends on the developmental stage of the oocyte being examined. For example, in
the first stage of prophase (leptotene) homologs are largely unpaired; thus in a normal meiocyte two FISH
signals per cell should be evident, while the presence of three chromosome-specific signals would
indicate a trisomic cell (Figure 1A, D). However, as cells progress to zygotene, the homologous
chromosomes begin to synapse and, depending on whether the probe hybridizes to a synapsed or
unsynapsed region, a normal cell could have one or two signals and a trisomic cell could have one, two,
43
or three signals. (Figure 1B, E). Finally, because the subsequent pachytene stage is characterized by
complete synapsis of homologs, both normal and trisomic cells should exhibit a single large FISH signal,
although synaptic defects associated with trisomy may lead to the presence of two or three signals
(Figure 1C, F).
A pilot study: FISH studies of oocytes in trisomy 21 fetuses
The basic meiotic principles detailed above complicate assessments of the level of trisomy in
prophase stage oocytes. Importantly, the inclusion of cells at zygotene and pachytene should generate
underestimates of the real values of aneuploidy. Thus, in contrast to the original analyses of Hultén et al
[30]– in which no attempt was made to discriminate by stage of prophase – we reasoned that careful
staging of oocytes and a focus on the analysis of leptotene stage cells would generate the most accurate
determinations of trisomy mosaicism. To test this, we conducted a pilot study of prophase stage oocytes
in trisomy 21 fetuses, since this situation provides the maximal likelihood of identifying three FISH signals.
We examined 37 prophase cells from two fetal ovarian samples, using an antibody to SYCP3 to identify
the synaptonemal complex and a distal chromosome 21 FISH probe to determine the number of
chromosomes 21 per cell (Figure 2). We visualized three chromosome 21 signals in both of two
leptotene cells examined, but rarely identified three signals in either zygotene or pachytene cells (Figure
2; Table I). These observations fit our expectations vis a vis the likelihood of detecting three chromosome
21 signals in trisomic cells and are consistent with previous cytological studies of meiocytes from trisomy
21 cases (Table I). Indeed, taken together, the results from four recent studies of trisomy 21 female
fetuses (including the present report) indicate that the majority of leptotene stage cells but only 29-39% of
zygotene and 0-7% of pachytene cells exhibit three chromosome 21 signals [34-36].
The implication of these observations is straightforward: accurate estimates of the frequency of
trisomic oocytes cannot be obtained by pooling data from cells at different stages of prophase. The
concern about accuracy would be minimized if leptotene cells were the predominant cell type identified in
fetal oocytes. However, this is not the case, as previous analyses of late first and second trimester
female fetuses indicate that, at least among conceptuses of 15 weeks gestation or more, the majority of
oocytes have completed leptotene [33, 37] and progressed to zygotene, pachytene or beyond.
44
Analysis of trisomy mosaicism in euploid fetuses
We concluded that an accurate assessment of trisomy mosaicism would require analysis of cells
at leptotene, as this is the only stage at which a trisomic oocyte would reliably display three signals.
Accordingly, we initiated studies of trisomy mosaicism in leptotene stage cells of seven female fetuses,
with gestational ages ranging from 16-23 weeks. We did not karyotype the samples. However, given the
low rate of chromosome abnormality in elective pregnancy terminations [e.g. 38], it seems reasonable to
assume that all were karyotypically normal.
For each sample we first immunostained the preparations, using SYCP3 to detect synaptonemal
complexes and CREST to identify centromeric regions. Subsequently, the slides were denatured and a
FISH probe recognizing the pericentromeric region of chromosome 16 and one detecting the
pericentromeric regions of chromosomes 13 and 21 were applied. In total, we examined 1405 leptotene
stage oocytes from the seven samples; of these 1034 were informative for the chromosome 16 probe and
1206 for the probe detecting both chromosomes 13 and 21 (Table II). For chromosome 16,
approximately two-thirds of the cells (686/1034 = 66.3%) contained two signals, with almost all of the
remainder (339/1034 = 32.8%) exhibiting a single signal. An additional nine cells (9/1034 = 0.9%) had
three signals, consistent with trisomy 16. However, when we merged the immunostained and FISH
images, it was clear that, in each of these nine cells, one of the three FISH signals did not co-localize with
a CREST signal (Figure 3). Thus, these are clearly non-specific signals, and when removed from the
data set, the results provide no evidence for any trisomy 16 cells.
Similar results were observed for chromosomes 13 and 21, for which the presence of four signals
would be consistent with euploidy for these chromosomes, five signals with trisomy for one of the two
chromosomes and six signals with trisomy for both chromosomes. In the majority of cells (758/1206 =
62.9%) we identified four signals, but in five cells a fifth signal was observed. However, as with
chromosome 16, in each of these cells one of the three signals did not co-localize with a CREST signal,
indicating a false positive result. Thus, taken together, our analyses provide no evidence for trisomy
mosaicism for either chromosomes 13, 16 or 21 in fetal ovarian preparations.
45
Putting it into perspective: is germ cell mosaicism an important source of human aneuploidy?
There is a wealth of data linking mosaicism for trisomy 21 to the etiology of Down syndrome. For
example, a number of case studies of recurrent Down syndrome have identified mosaicism for trisomy 21
in gonadal, skin or blood samples of one of the parents [39-45]. Further, in studies of 374 singleton Down
syndrome patients, Uchida and Freeman identified parental mosaicism in nearly 3% of the families [41],
and in a review of 221 cases, Harris et al. reported parental mosaicism in 1.6% of the families [39]. Thus,
the importance of parental mosaicism to a small, but non-trivial, proportion of cases of Down syndrome is
well established.
However, the OMSM model proposed by Hultén and colleagues suggests a much more important
role for mosaicism in the genesis of human aneuploidy [30]. Specifically, it suggests that trisomy
mosaicism is a major source of trisomy 21 and the cause of the maternal age effect on trisomy 21, as well
as a likely contributor to other human trisomies [1, 46]. Several aspects of the OMSM are appealing.
First, it provides a unifying model for the genesis of trisomy 21, and unlike many recent reports, requires
only a single “hit” (i.e., pre-meiotic mitotic nondisjunction) to explain the origin of aneuploidy. Second, by
assuming that trisomic oocytes are developmentally delayed and consequently among the last to be
ovulated, the model provides a mechanism for generating a maternal age-dependent increase in the
incidence of disomic oocytes. Finally, by invoking different segregation modes for the various meiotic
configurations that three chromosomes can adopt in prophase, the OMSM is able to account for the wellknown – if complicated – association between meiotic recombination errors and trisomy 21 [28].
However, the OMSM also has at least two important shortcomings. First, Hultén and colleagues
suggest that it provides a straightforward alternative to more complex models of maternal age-dependent
trisomy that invoke a variety of biological and environmental factors [1]. Unfortunately, as appealing as it
would be to have only one major contributor to age-related aneuploidy, mounting evidence indicates that
the origin of the age effect is, indeed, complicated [2]. Studies of model organisms have reported a
number of routes to age-related aneuploidy, including abnormalities in formation of the synaptonemal
complex [14, 47], in maintenance of the sister chromatid cohesion complex [16, 17, 20] and in the
function of cell cycle checkpoint controls [18, 26]. Further, analyses of humans also provide evidence of
multiple, chromosome-specific mechanisms of nondisjunction and, presumably, of maternal age-related
46
trisomy. For example, it is now clear that the parent and meiotic stage of origin varies depending on the
chromosome involved [3, 48-52], that there are different maternal age curves for trisomies involving
different chromosomes [4] (and even for different meiotic origins involving the same chromosome [53]),
and that the effects of abnormal meiotic recombination depend on maternal age and on the chromosome
involved [29, 54]. Further, recent direct analyses of prophase stage fetal oocytes provide compelling
evidence that abnormalities in pairing or recombination – and not the presence of three chromosomes –
are common in the human female, and that they vary depending on the specific chromosome [33]. Thus,
there is now ample evidence that no one mechanism is likely to explain all, or even a substantial
proportion of, cases of human aneuploidy.
Second, and more important, the OMSM has yet to be replicated and, indeed, the available data
contradict the basic tenet of the model; i.e., the occurrence of trisomy 21 mosaicism in fetal oocytes. For
example, the results of the present study provide no evidence for the presence of trisomic oocytes for any
of the three chromosomes we analyzed. For chromosome 21 our results are highly significantly different
2
from the 0.66% level of trisomy 21 mosaicism reported by Hultén et al. [30] for prophase cells (χ = 7.95;
2
p<0.01) as well as for the 0.54% value identified for all ovarian cells (χ = 6.51; p=0.01). Further, this
almost certainly underestimates the real difference between the observations of Hultén et al. [30] and
ours, since their analysis included cells (i.e., zygotene and pachytene) with a low likelihood of trisomy
detection.
Similarly, other analyses have failed to identify appreciable levels of trisomy 21 mosaicism. For
example, Cheng and colleagues analyzed 3008 prophase stage cells (1195 at leptotene, 511 at zygotene
and 1302 at pachytene) from eight euploid fetuses and were unable to identify a single cell with three
chromosome 21 signals [34]. More recently, Robles and colleagues examined seven euploid fetuses and
identified trisomy 21 in 1/425 (0.2%) leptotene cells but in none of 985 zygotene or pachytene cells [36].
Finally, in the only study conducted after that of Hultén and colleagues [30], Morris and colleagues
analyzed both fetal ovarian and skin samples, and observed extremely low levels of trisomy 21 mosaicism
in each of the two tissue types (2/8365 = 0.02% in ovarian cells and 5/8245 = 0.06% in skin cells) [31].
Thus, taken together, the results of four different analyses suggest that – for reasons that are unclear –
the original report of Hultén et al. [30] markedly overestimated the incidence of trisomy 21 in fetal oocytes.
47
In summary, we and others have failed to replicate the observations of Hultén et al. [30];
accordingly, we conclude that pre-meiotic mitotic nondisjunction is unlikely to cause the maternal age
effect and probably only contributes to a small subset of human aneuploidies. This interpretation is
consistent with the growing body of data from studies of aneuploidy in model organisms and direct
analyses of meiosis in human oocytes, which indicate the existence of multiple sources of human
aneuploidy and multiple causes of the maternal age effect. Thus, we suggest that future investigations of
aneuploidy focus on events occurring in meiosis, and not those that occur in germ cells before they enter
the meiotic pathway.
ACKNOWLEDGEMENTS
This work was supported by NIH grants R01 HD21341 (to T.H.) and R01 ES013527 (to P.H.).
We thank the staff and faculty at San Francisco General Hospital Women’s Options Center for assistance
in the collection of tissues. We also thank Katie Stephenson, Dylan Atchley and Cynthia Megloza for their
assistance in recruitment and data collection.
48
REFERENCES
1.
Hulten, M.A., et al., On the origin of the maternal age effect in trisomy 21 Down syndrome: the
Oocyte Mosaicism Selection model. Reproduction, 2010. 139(1): p. 1-9.
2.
Nagaoka, S.I., T.J. Hassold, and P.A. Hunt, Human aneuploidy: mechanisms and new insights
into an age-old problem. Nat Rev Genet, 2012. 13(7): p. 493-504.
3.
Hassold, T. and P. Hunt, To err (meiotically) is human: the genesis of human aneuploidy. Nat Rev
Genet, 2001. 2(4): p. 280-91.
4.
Hassold, T. and D. Chiu, Maternal age-specific rates of numerical chromosome abnormalities
with special reference to trisomy. Hum Genet, 1985. 70(1): p. 11-7.
5.
Martin, J.A., et al., Births: final data for 2010. Natl Vital Stat Rep, 2012. 61(1): p. 1-71.
6.
Loane, M., et al., Twenty-year trends in the prevalence of Down syndrome and other trisomies in
Europe: impact of maternal age and prenatal screening. Eur J Hum Genet, 2013. 21(1): p. 27-33.
7.
Johnson, M.H., Essential reproduction. 7th ed2013, Chichester, West Sussex: John Wiley &
Sons. 392.
8.
Henderson, S.A. and R.G. Edwards, Chiasma frequency and maternal age in mammals. Nature,
1968. 218(5136): p. 22-8.
9.
Polani, P.E. and J.A. Crolla, A test of the production line hypothesis of mammalian oogenesis.
Hum Genet, 1991. 88(1): p. 64-70.
10.
Hussin, J., et al., Age-dependent recombination rates in human pedigrees. PLoS Genet, 2011.
7(9): p. e1002251.
11.
Coop, G., et al., High-resolution mapping of crossovers reveals extensive variation in fine-scale
recombination patterns among humans. Science, 2008. 319(5868): p. 1395-8.
12.
Kong, A., et al., Recombination rate and reproductive success in humans. Nat Genet, 2004.
36(11): p. 1203-6.
13.
Hook, E.B., Issues pertaining to the impact and etiology of trisomy 21 and other aneuploidy in
humans; a consideration of evolutionary implications, maternal age mechanisms, and other
matters. Prog Clin Biol Res, 1989. 311: p. 1-27.
14.
Wang, H. and C. Hoog, Structural damage to meiotic chromosomes impairs DNA recombination
and checkpoint control in mammalian oocytes. J Cell Biol, 2006. 173(4): p. 485-95.
15.
Hodges, C.A., et al., SMC1beta-deficient female mice provide evidence that cohesins are a
missing link in age-related nondisjunction. Nat Genet, 2005. 37(12): p. 1351-5.
16.
Lister, L.M., et al., Age-related meiotic segregation errors in mammalian oocytes are preceded by
depletion of cohesin and Sgo2. Curr Biol, 2010. 20(17): p. 1511-21.
17.
Chiang, T., et al., Evidence that weakened centromere cohesion is a leading cause of age-related
aneuploidy in oocytes. Curr Biol, 2010. 20(17): p. 1522-8.
18.
Vogt, E., et al., Spindle formation, chromosome segregation and the spindle checkpoint in
mammalian oocytes and susceptibility to meiotic error. Mutat Res, 2008. 651(1-2): p. 14-29.
49
19.
Liu, L. and D.L. Keefe, Defective cohesin is associated with age-dependent misaligned
chromosomes in oocytes. Reprod Biomed Online, 2008. 16(1): p. 103-12.
20.
Jessberger, R., Age-related aneuploidy through cohesion exhaustion. EMBO Rep, 2012. 13(6): p.
539-46.
21.
Tachibana-Konwalski, K., et al., Rec8-containing cohesin maintains bivalents without turnover
during the growing phase of mouse oocytes. Genes Dev, 2010. 24(22): p. 2505-16.
22.
Roberts, R., et al., Follicle-stimulating hormone affects metaphase I chromosome alignment and
increases aneuploidy in mouse oocytes matured in vitro. Biol Reprod, 2005. 72(1): p. 107-18.
23.
Crowley, P.H., et al., A chiasma-hormonal hypothesis relating Down's syndrome and maternal
age. Nature, 1979. 280(5721): p. 417-8.
24.
Gaulden, M.E., Maternal age effect: the enigma of Down syndrome and other trisomic conditions.
Mutat Res, 1992. 296(1-2): p. 69-88.
25.
Steuerwald, N., et al., Association between spindle assembly checkpoint expression and
maternal age in human oocytes. Mol Hum Reprod, 2001. 7(1): p. 49-55.
26.
Nagaoka, S.I., et al., Oocyte-specific differences in cell-cycle control create an innate
susceptibility to meiotic errors. Curr Biol, 2011. 21(8): p. 651-7.
27.
Hassold, T. and S. Sherman, Down syndrome: genetic recombination and the origin of the extra
chromosome 21. Clin Genet, 2000. 57(2): p. 95-100.
28.
Oliver, T.R., et al., New insights into human nondisjunction of chromosome 21 in oocytes. PLoS
Genet, 2008. 4(3): p. e1000033.
29.
Lamb, N.E., et al., Association between maternal age and meiotic recombination for trisomy 21.
Am J Hum Genet, 2005. 76(1): p. 91-9.
30.
Hulten, M.A., et al., On the origin of trisomy 21 Down syndrome. Mol Cytogenet, 2008. 1: p. 21.
31.
Morris, C.R., et al., Origin of trisomy: no evidence to support the ovarian mosaicism theory.
Prenat Diagn, 2012. 32(7): p. 668-73.
32.
Peters, A.H., et al., A drying-down technique for the spreading of mammalian meiocytes from the
male and female germline. Chromosome Res, 1997. 5(1): p. 66-8.
33.
Cheng, E.Y., et al., Meiotic recombination in human oocytes. PLoS Genet, 2009. 5(9): p.
e1000661.
34.
Cheng, E.Y., et al., An analysis of meiotic pairing in trisomy 21 oocytes using fluorescent in situ
hybridization. Cytogenet Cell Genet, 1998. 80(1-4): p. 48-53.
35.
Barlow, A.L., C. Tease, and M.A. Hulten, Meiotic chromosome pairing in fetal oocytes of trisomy
21 human females. Cytogenet Genome Res, 2002. 96(1-4): p. 45-51.
36.
Robles, P., et al., Pairing and synapsis in oocytes from female fetuses with euploid and aneuploid
chromosome complements. Reproduction, 2007. 133(5): p. 899-907.
37.
Roig, I., et al., Evolution of the meiotic prophase and of the chromosome pairing process during
human fetal ovarian development. Hum Reprod, 2005. 20(9): p. 2463-9.
50
38.
Burgoyne, P.S., K. Holland, and R. Stephens, Incidence of numerical chromosome anomalies in
human pregnancy estimation from induced and spontaneous abortion data. Hum Reprod, 1991.
6(4): p. 555-65.
39.
Harris, D.J., et al., Parental trisomy 21 mosaicism. Am J Hum Genet, 1982. 34(1): p. 125-33.
40.
Tseng, L.H., et al., Recurrent Down's syndrome due to maternal ovarian trisomy 21 mosaicism.
Arch Gynecol Obstet, 1994. 255(4): p. 213-6.
41.
Uchida, I.A. and V.C. Freeman, Trisomy 21 Down syndrome. Parental mosaicism. Hum Genet,
1985. 70(3): p. 246-8.
42.
Sachs, E.S., et al., Trisomy 21 mosaicism in gonads with unexpectedly high recurrence risks. Am
J Med Genet Suppl, 1990. 7: p. 186-8.
43.
Frias, S., et al., Detection of mosaicism in lymphocytes of parents of free trisomy 21 offspring.
Mutat Res, 2002. 520(1-2): p. 25-37.
44.
Pangalos, C.G., et al., DNA polymorphism analysis in families with recurrence of free trisomy 21.
Am J Hum Genet, 1992. 51(5): p. 1015-27.
45.
Bruyere, H., et al., 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,
2000. 94(1): p. 35-41.
46.
Hulten, M.A., et al., Germinal and Somatic Trisomy 21 Mosaicism: How Common is it, What are
the Implications for Individual Carriers and How Does it Come About? Curr Genomics, 2010.
11(6): p. 409-19.
47.
Yuan, L., et al., Female germ cell aneuploidy and embryo death in mice lacking the meiosisspecific protein SCP3. Science, 2002. 296(5570): p. 1115-8.
48.
Hassold, T.J., et al., Molecular studies of non-disjunction in trisomy 16. J Med Genet, 1991. 28(3):
p. 159-62.
49.
Bugge, M., et al., Non-disjunction of chromosome 13. Hum Mol Genet, 2007. 16(16): p. 2004-10.
50.
Bugge, M., et al., Non-disjunction of chromosome 18. Hum Mol Genet, 1998. 7(4): p. 661-9.
51.
Hall, H.E., et al., The origin of trisomy 13. Am J Med Genet A, 2007. 143A(19): p. 2242-8.
52.
Hall, H.E., et al., The origin of trisomy 22: evidence for acrocentric chromosome-specific patterns
of nondisjunction. Am J Med Genet A, 2007. 143A(19): p. 2249-55.
53.
Yoon, P.W., et al., Advanced maternal age and the risk of Down syndrome characterized by the
meiotic stage of chromosomal error: a population-based study. Am J Hum Genet, 1996. 58(3): p.
628-33.
54.
Sherman, S.L., N.E. Lamb, and E. Feingold, Relationship of recombination patterns and maternal
age among non-disjoined chromosomes 21. Biochem Soc Trans, 2006. 34(Pt 4): p. 578-80.
51
TABLE I – Detection of chromosome 21 signals by stage of prophase in fetuses with trisomy 21.
- - -No. Signals- - -
Present Study
Cheng et al., 1998
Robles et al., 2007*
Barlow et al., 2002*
Stage
No. Cells
1
2
3
Leptotene
2
Zygotene
21
Pachytene
14
Leptotene
3,069
Zygotene
662
Pachytene
2,067
Leptotene
3
Zygotene
27
Pachytene
84
Pachytene
88
0
(0%)
14
(67%)
8
(57%)
430
(14%)
316
(48%)
1,351
(65%)
0
(0%)
11
(41%)
55
(65%)
37
(42%)
0
(0%)
6
(29%)
5
(36%)
966
(31%)
257
(39%)
616
(30%)
0
(0%)
9
(33%)
28
(33%)
51
(58%)
2
(100%)
1
(5%)
1
(7%)
1,673
(55%)
89
(13%)
100
(5%)
3
(100%)
7
(26%)
1
(1%)
0
(0%)
* Original scoring was based on pairing configurations and has been converted assuming that cells with 3
univalents = 3 signals; cells with bivalent + univalent = 2 signals; and cells with trivalent/partial trivalent =
1 signal.
52
TABLE II – Number of FISH signals per chromosome in leptotene oocytes from seven female fetuses
- - - - - - - - - - - - - -No. signals- - - - - - - - - - - - - Chromosome*
No. Cells
1
2
3
4
5
No. false
positives^
16
13 / 21
1034
1206
339
1
686
222
0
220
0
758
0
0
9
5
* For chromosome 16, the presence of three signals would indicate a trisomic cell. The probe use for
chromosomes 13 and 21 detects both chromosomes; thus the presence of five signals would indicate a
trisomy for either chromosome 13 or 21.
^ Cells in which the additional FISH signal failed to co-localize with a CREST signal.
53
Figure 1 – The expected number of FISH signals per chromosome is dependent on the stage of meiotic
prophase.
In euploid oocytes (A-C), homologs may be completely separated at leptotene, allowing for detection of
both chromosomes. At zygotene, homologs begin to synapse, so that either one or two FISH signals may
be apparent. At pachytene, homologs are completely synapsed; thus, typically one signal should be
evident. In trisomic oocytes (D-F), three signals should be readily apparent at leptotene but, depending
on the extent of synapsis at zygotene and pachytene, one, two or three FISH signals may be observed.
54
Figure 1
55
Figure 2 – Prophase stage oocytes from trisomy 21 fetuses seldom exhibit three chromosome 21 signals.
Representative immunostaining (A) and FISH (B) images of a late zygotene oocyte from a trisomy 21
fetus. Due to partial synapsis between the three homologs, only two chromosome 21 signals are evident.
56
Figure 2
57
Figure 3 – Putative trisomic cells may contain an artifactual FISH signal.
We identified 14 leptotene cells with an additional FISH signal, but in each instance one of the three
signals failed to co-localize with a CREST signal, indicating a false positive result. Shown here are FISH
(A) and immunostained (B) images of the same cell, in which we identified three chromosome 16 FISH
signals but on immunostaining, CREST signals overlapped the FISH signals in only two of the three
instances (i.e., the green arrows indicate where FISH signals overlap CREST signals, while the white
arrow indicates where an artifactual FISH signal is present).
58
Figure 3
59
CHAPTER III
Examining Variation in Recombination Levels in the Human Female: A Test of the Production Line
Hypothesis
Ross Rowsey, Jennifer Gruhn, Karl W. Broman, Patricia A. Hunt, Terry Hassold
Rowsey R, Gruhn J, Broman K, Hunt P, Hassold T. 2014. Examining Variation in Recombination Levels in
the Human Female: A Test of the Production Line Hypothesis. Am J Hum Genet 95(1): 108-112
ABSTRACT
The most important risk factor for human aneuploidy is increasing maternal age, but the basis of
this association remains unknown. Indeed, one of the earliest models of the maternal age effect – the
“Production Line Model” proposed by Henderson and Edwards in 1968 – remains one of the most-cited
explanations. The model has two key components: that the first oocytes to enter meiosis are the first
ovulated, and that the first to enter meiosis have more recombination events (crossovers) than those that
enter meiosis later in fetal life. Studies in rodents demonstrate that the first oocytes to enter meiosis are,
indeed, the first to be ovulated, but the association between timing of meiotic entry and recombination
levels has not been tested. We recently initiated molecular cytogenetic studies of second trimester human
fetal ovaries, allowing us to directly examine the number and distribution of crossover-associated proteins
in prophase stage oocytes. Our observations on over 8,000 oocytes from 191 ovarian samples
demonstrate extraordinary variation in recombination within and among individuals, but provide no
evidence of a difference in recombination levels between oocytes entering meiosis early or late in fetal
life. Thus, our data provide a direct test of the second tenet of the Production Line model and suggest
that it does not provide a plausible explanation for the human maternal age effect, meaning that – 45
years after its introduction – we can finally conclude that the Production Line Model is not the basis for the
maternal age effect on trisomy.
61
MAIN TEXT
Aneuploidy is the leading known cause of pregnancy wastage and congenital birth defects in our
species, occurring in as many as 35% of all human pregnancies (reviewed in [1]). The vast majority of
aneuploidy is maternal in origin, suggesting that egg production is inherently error-prone in humans [2].
Further, this risk increases dramatically with the age of the woman: for women in their twenties the
likelihood of having a clinically recognized trisomic pregnancy is about 2-3%, but this value increases to
over 30% for women in their forties [3].
The basis of the effect of maternal age on aneuploidy remains a mystery, but a number of
potential mechanisms have been proposed. One of the earliest and most enduring hypotheses is the
Production Line Model, initially proposed in 1968 [4] on the basis of apparent age-dependent changes in
chiasma frequency in mouse oocytes. It ascribes the age effect to differences in recombination and has
two key components. First, it assumes that there is a direct relationship between the timing of meiotic
entry in the fetal ovary and the timing of ovulation in the adult; i.e., oocytes that are the first to enter
meiosis are the first ovulated, and oocytes entering meiosis last will be ovulated at the end of the
reproductive lifespan. Second, it assumes that meiotic recombination rates vary with gestational age,
with the first oocytes entering meiosis having higher recombination levels than oocytes that enter later.
According to this model, the last oocytes ovulated will have the lowest numbers of crossovers and may, in
fact, have an increased frequency of “crossover-less” homologs, greatly increasing the risk of
nondisjunction.
In the intervening 45 years, two lines of experimental evidence that are consistent with the tenets
of the Production Line Model have been produced. First, radiolabelling studies in mice and rats suggest
that there is, indeed a production line; i.e., the first oocytes to enter meiosis appear to be the first to be
ovulated [5, 6]. Second, studies in a variety of organisms have demonstrated the importance of
recombination abnormalities to the occurrence of meiotic nondisjunction (e.g. [7-9]). The evidence from
humans is especially strong, since abnormal levels or positioning of recombination events have been
detected in all trisomies that have been appropriately studied (e.g. [2, 10, 11]). However, the predicted
relationship between recombination levels and maternal age – i.e., that oocytes from older women have
62
lower levels of recombination than oocytes from younger women – has not been demonstrated. Some
genetic linkage studies have shown a reduction in recombination in pregnancies involving older
women[12, 13], but others have either found no effect or have reported an increase in recombination
levels with increasing maternal age [14]. Thus, the data from linkage analyses are equivocal and,
moreover, this approach may not be appropriate for a key reason. That is, virtually all linkage studies are
based on liveborn individuals, but the vast majority of aneuploid conceptuses terminate in utero.
Consequently, traditional linkage analysis of liveborn populations has limited power to assess the
relationship between recombination and maternal-age related aneuploidy.
Accordingly, we decided to directly test the second tenet of the Production Line Model by
examining meiotic recombination in human fetal oocytes. We reasoned that recombination differences in
oocytes that initiate meiosis at different times would be evident in a population of fetal ovarian samples as
a change in recombination levels with gestational age. Thus, we took advantage of our ongoing analyses
of fetal ovarian samples from elective terminations of pregnancy at the San Francisco General Hospital
Women’s Option Center in San Francisco, California (e.g.[15-17]), as well as data from our previous
studies at the University of Washington Medical Center in Seattle, Washington [16], asking whether the
levels of recombination are affected by the gestational age of the fetus. For these analyses, we collected
fetal ovarian samples from elective terminations of pregnancy, as previously described [15-17]. Our
studies were conducted according to the principles expressed in the Declaration of Helsinki, and were
approved by the University of California-San Francisco, University of Washington, and Washington State
University Institutional Review Boards, and informed consent was obtained from all study participants.
We utilized immunofluorescence to examine crossover-associated proteins in prophase stage
oocytes from these samples. Specifically, we analyzed the number and distribution of foci for the DNA
mismatch repair protein MLH1, thought to localize to approximately 90% of crossovers in pachytene
stage cells of mammalian species[18]. Since MLH1 foci occur in the context of the synaptonemal
complex (SC), we also visualized the SC using antibodies to the axial element protein SYCP3 (Figure
1a). In total, we analyzed 8518 cells from 191 fetal samples with gestational ages ranging from 14 and
26 weeks, typically examining between 25-65 cells per case.
63
Extensive individual variation was evident in the 191 cases. That is, mean MLH1 values (+/S.E.) ranged from 51.1 +/- 1.3 to 92.3 +/- 2.1, meaning that the level of recombination for the cases with
the lowest MLH1 values was only 55% of that for the cases with the highest values (Figure 1b). Pooling
the data from all cases, the overall mean number of MLH1 foci per cell was 66.3+/- 0.6. Assuming that
one MLH1 focus = one crossover = 50 cM, this yields a genome-wide female map length of approximately
3315 cM. This estimate is consistent with previous cytological studies of recombination in human
females, in which the inferred genome-wide estimates have ranged from about 3000-4000 cM [16, 19-21];
Figure 1c). However, these values are consistently lower than those derived from linkage analysis, where
estimates range from about 4000-4500 cM ([13, 22-27]; Figure 1c). Previously, we [16, 17] have
suggested two possible reasons for this discrepancy. First, all cytological analyses have involved
immunostaining assays of MLH1, and have not assessed the less than 10% of crossovers that result from
alternative recombination pathways; e.g. non-interfering crossovers associated with the endonuclease
MUS81[28, 29]. Second, localization of MLH1 on SCs appears to be asynchronous in human oocytes,
and not all foci are visible in pachytene stage oocytes [16, 21]. These caveats aside, the general
conclusions from our and other cytological studies of recombination are in good agreement with data from
linkage studies and provide evidence for surprising variability in genome-wide recombination levels in
human females by comparison with human males [16, 17, 19, 21, 24].
In subsequent analyses, we tested the effects of gestational age on meiosis. Initially, we
examined the mean number of MLH1 foci per cell for each case, and sorted cases by gestational age
(Figure 2). As is evident from Figure 2, there was no apparent effect of gestational age on genome-wide
mean MLH1 values. Figure 2 also shows no change in the range of MLH1 values within individual cases
across multiple gestational ages. Subsequently, we analyzed the distribution and mean number of MLH1
foci on four chromosomes known to be nondisjunction-prone or to contribute to clinical disorders (i.e.,
chromosomes 16, 18, 21 and 22). We found no association between gestational age and the placement
(data not shown) or number (Figure 3) of MLH1 foci on these chromosomes. Importantly, in contrast to
the prediction of the Production Line Model, there was no increase in “crossover-less” chromosomes with
increasing gestational age (Figure 3). Finally, we examined the length of the synaptonemal complex, a
variable known to be directly correlated with genome-wide MLH1 values [19, 30]. We analyzed the SC
64
lengths for chromosomes 16, 18, 21 and 22, but found no effect of gestational age on SC length for any
of the chromosomes (Figure S1). Thus, taken together, our analyses failed to detect any genome-wide or
chromosome specific recombination-associated changes attributable to the gestational age of the sample.
In a final set of studies, we asked whether the age of the mother might influence recombination
levels in the oocytes of her female fetuses (i.e., a potential grand-maternal age effect). However, as is
evident from Figure 4, we found no indication of such an effect.
Two important conclusions derive from our analyses. First, we found no evidence that the
gestational age of the fetus influences the level or positioning of crossover events. Thus, the suggestion
by Henderson and Edwards [4] that a “gradient” in the fetal ovary causes the first-formed oocytes to have
more chiasmata than those formed last appears to be incorrect. Accordingly, we conclude that the
Production Line Model as initially proposed is not the cause of the maternal age effect on human
aneuploidy. Nevertheless, the observations that led to the model – i.e., declining numbers of chiasmata
with increasing maternal age -- can easily be reconciled with our data. That is, recent studies in rodents
indicate an age related loss of cohesin in oocytes [31, 32]. In addition to tethering sister chromatids,
cohesin serves to link homologous chromosomes together during the first meiotic division. Thus, loss of
cohesion with increasing maternal age may cause homologs tethered by single distal exchanges to slip
apart from one another and, on cytological examination of diakinesis preparations, would yield an
apparent increase in the number of univalents.
Second, and equally important, our observations suggest extraordinary variation in genome-wide
crossover levels among different fetal samples. Clearly, individual variation in recombination rates has
been documented previously; e.g., in an analysis of different CEPH families, Broman et al reported
female maps as low as 3300 centimorgans and as high as 4700 centimorgans [23]. Our observations
suggest even greater variability, with genome-wide maps ranging from approximately 2500 cM to over
4600 cM among the different samples. Intriguingly, this level of variation is not evident in the human male
[17, 23], suggesting that recombination is less tightly controlled in human oogenesis than in
spermatogenesis. While the basis of this sex-specific difference remains unclear, it seems unlikely that it
is can be explained by the asynchrony of the process in females. However, by combining SNP analyses
65
of recombination-associated loci (e.g., see ref [33]) with direct studies of recombination levels, it may be
possible to illuminate the genetic underpinnings of this surprising difference in variation among human
males and females.
ACKNOWLEDGEMENTS
We thank the staff and faculty at San Francisco General Hospital’s Women’s Options Center, San
Francisco, CA, and at the University of Washington Medical Center, Seattle, WA for assistance in the
collection of tissues. We also thank Tracey Woodruff, Carrie Dickenson, Katie Stevenson, Dylan Atchley,
Mei-Lani Bixby, Cynthia Megloza, Jody Steinauer, Edith Cheng, Theresa Naluai-Cecchini, Terah Hansen,
Elizabeth Pascucci, Changqing Zhou, Chris Small and Heather Hagen for their assistance in recruitment
and data collection. This work was supported by NIH grants HD21341 (to TH) and ES013527 (to PH),
and by an NIH training grant awarded to the School of Molecular Biosciences at Washington State
University (T32 GM083864 to RR and JG).
66
REFERENCES
1.
Nagaoka, S.I., T.J. Hassold, and P.A. Hunt, Human aneuploidy: mechanisms and new insights
into an age-old problem. Nat Rev Genet, 2012. 13(7): p. 493-504.
2.
Hassold, T. and P. Hunt, To err (meiotically) is human: the genesis of human aneuploidy. Nat Rev
Genet, 2001. 2(4): p. 280-91.
3.
Hassold, T. and D. Chiu, Maternal age-specific rates of numerical chromosome abnormalities
with special reference to trisomy. Hum Genet, 1985. 70(1): p. 11-7.
4.
Henderson, S.A. and R.G. Edwards, Chiasma frequency and maternal age in mammals. Nature,
1968. 218(5136): p. 22-8.
5.
Polani, P.E. and J.A. Crolla, A test of the production line hypothesis of mammalian oogenesis.
Hum Genet, 1991. 88(1): p. 64-70.
6.
Hirshfield, A.N., Heterogeneity of cell populations that contribute to the formation of primordial
follicles in rats. Biol Reprod, 1992. 47(3): p. 466-72.
7.
Ward, J.O., et al., Mutation in mouse hei10, an e3 ubiquitin ligase, disrupts meiotic crossing over.
PLoS Genet, 2007. 3(8): p. e139.
8.
Koehler, K.E., et al., Recombination and nondisjunction in humans and flies. Hum Mol Genet,
1996. 5 Spec No: p. 1495-504.
9.
Ross, L.O., R. Maxfield, and D. Dawson, Exchanges are not equally able to enhance meiotic
chromosome segregation in yeast. Proc Natl Acad Sci U S A, 1996. 93(10): p. 4979-83.
10.
Lamb, N.E., et al., Association between maternal age and meiotic recombination for trisomy 21.
Am J Hum Genet, 2005. 76(1): p. 91-9.
11.
Hassold, T., H. Hall, and P. Hunt, The origin of human aneuploidy: where we have been, where
we are going. Hum Mol Genet, 2007. 16 Spec No. 2: p. R203-8.
12.
Hussin, J., et al., Age-dependent recombination rates in human pedigrees. PLoS Genet, 2011.
7(9): p. e1002251.
13.
Bleazard, T., et al., Fine-scale mapping of meiotic recombination in Asians. BMC Genet, 2013.
14: p. 19.
14.
Kong, A., et al., Recombination rate and reproductive success in humans. Nat Genet, 2004.
36(11): p. 1203-6.
15.
Rowsey, R., et al., Germline mosaicism does not explain the maternal age effect on trisomy. Am
J Med Genet A, 2013. 161(10): p. 2495-503.
16.
Cheng, E.Y., et al., Meiotic recombination in human oocytes. PLoS Genet, 2009. 5(9): p.
e1000661.
17.
Gruhn, J.R., et al., Cytological studies of human meiosis: sex-specific differences in
recombination originate at, or prior to, establishment of double-strand breaks. PLoS One, 2013.
8(12): p. e85075.
67
18.
Guillon, H., et al., Crossover and noncrossover pathways in mouse meiosis. Mol Cell, 2005.
20(4): p. 563-73.
19.
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, 2004. 107(3-4): p. 208-15.
20.
Tease, C., G.M. Hartshorne, and M.A. Hulten, Patterns of meiotic recombination in human fetal
oocytes. Am J Hum Genet, 2002. 70(6): p. 1469-79.
21.
Lenzi, M.L., et al., Extreme heterogeneity in the molecular events leading to the establishment of
chiasmata during meiosis i in human oocytes. Am J Hum Genet, 2005. 76(1): p. 112-27.
22.
Jorgenson, E., et al., Ethnicity and human genetic linkage maps. Am J Hum Genet, 2005. 76(2):
p. 276-90.
23.
Broman, K.W., et al., Comprehensive human genetic maps: individual and sex-specific variation
in recombination. Am J Hum Genet, 1998. 63(3): p. 861-9.
24.
Kong, A., et al., A high-resolution recombination map of the human genome. Nat Genet, 2002.
31(3): p. 241-7.
25.
Matise, T.C., et al., A second-generation combined linkage physical map of the human genome.
Genome Res, 2007. 17(12): p. 1783-6.
26.
Ju, Y.S., et al., A genome-wide Asian genetic map and ethnic comparison: the GENDISCAN
study. BMC Genomics, 2008. 9: p. 554.
27.
Coop, G., et al., High-resolution mapping of crossovers reveals extensive variation in fine-scale
recombination patterns among humans. Science, 2008. 319(5868): p. 1395-8.
28.
de los Santos, T., et al., The Mus81/Mms4 endonuclease acts independently of double-Holliday
junction resolution to promote a distinct subset of crossovers during meiosis in budding yeast.
Genetics, 2003. 164(1): p. 81-94.
29.
Holloway, J.K., et al., MUS81 generates a subset of MLH1-MLH3-independent crossovers in
mammalian meiosis. PLoS Genet, 2008. 4(9): p. e1000186.
30.
Lynn, A., et al., Covariation of synaptonemal complex length and mammalian meiotic exchange
rates. Science, 2002. 296(5576): p. 2222-5.
31.
Chiang, T., et al., Evidence that weakened centromere cohesion is a leading cause of age-related
aneuploidy in oocytes. Curr Biol, 2010. 20(17): p. 1522-8.
32.
Lister, L.M., et al., Age-related meiotic segregation errors in mammalian oocytes are preceded by
depletion of cohesin and Sgo2. Curr Biol, 2010. 20(17): p. 1511-21.
33.
Kong, A., et al., Common and low-frequency variants associated with genome-wide
recombination rate. Nat Genet, 2014. 46(1): p. 11-6.
68
Figure 1 – Recombination in Human Oocytes.
(A) Representative image from a pachytene stage human fetal oocyte. Antibodies against SYCP3
(representing the axial-lateral elements of the synaptonemal complex) are visualized in red and against
the crossover-associated DNA mismatch repair protein MLH1 in green, and CREST antiserum-positive
signals (recognizing centromeric regions) are visualized in blue. (B) Distribution of mean MLH1 values
per cell in 191 fetal ovarian samples. (C) Estimates of female genetic map lengths from genetic linkage
studies (left, in blue) and cytological studies of pachytene oocytes (right, in red). References are
indicated beneath each estimate.
69
Figure 1
70
Figure 2 – Influence of Gestational Age on Genome-Wide Recombination Levels.
For each of the 191 cases, the mean number of MLH1 foci per case is represented by red diamonds, with
the values for individual cells represented by blue diamonds. No obvious effect of gestational age on
recombination levels was observed.
71
Figure 2
72
Figure 3 – Influence of Gestational Age on the Number of Crossovers on Individual Chromosomes.
For a subset of cases, we analyzed the number of MLH1 foci on individual chromosomes; i.e., (A)
chromosome 16; no. of cases = 9 (B) chromosome 18, no. of cases = 7 (C) chromosome 21; no. of cases
= 11 and (D) chromosome 22; no. of cases = 11. There was no obvious effect of gestational age on the
number of MLH1 foci per chromosome; in particular, the number of chromosomes lacking an MLH1 focus
was not affected by gestational age.
73
Figure 3
74
Figure 4 – Influence of Maternal Age on Genome-Wide Recombination Levels.
For each of the 119 cases, the mean number of MLH1 foci per case is represented by red diamonds, with
the values for individual cells represented by blue diamonds. No obvious effect of maternal age on
recombination levels was observed.
75
Figure 4
76
Figure S1 – Influence of Gestational Age on the Synaptonemal Complex Length of Individual
Chromosomes.
For a subset of cases, we analyzed the length of the SC on individual chromosomes: i.e., (A)
chromosome 16; no. of cases = 9 (B) chromosome 18; no. of cases = 7 (C) chromosome 21; no. of cases
= 11 and (D) chromosome 22; no. of cases = 11. There was no obvious effect of gestational age on the
length of the SC on any chromosome. Data points represent means of each analyzed case, and error
bars represent SE (points with no error bars represent cases with only one measured SC).
77
Supplemental Figure 1
78
Chapter IV
SUMMARY AND FUTURE DIRECTIONS
Summary:
With increasing numbers of couples waiting until later in life to procreate, the impact of the
maternal age effect cannot be overstated. Indeed, it is arguably the most important etiological agent for
any human genetic disorder. The studies in this thesis have focused on examining hypotheses that
suggest a fetal origin to the maternal age effect. Accordingly, the studies presented in this thesis rely on
studies of human fetal oocytes.
Chapter II examines a “predestination” model of the maternal age effect, the Oocyte Mosaicism
Selection Model (OMSM). The OMSM hypothesized that the maternal age effect was due to pre-meiotic,
mitotic segregation errors, and that all females had similar rates of these errors prior to entering meiosis
[1]. Utilizing immunofluorescence and FISH, we analyzed the chromosome content of meiotic prophase
stage cells. In contrast to the original report by Hulten and colleagues [2], we observed no evidence for
pre-meiotic aneuploidy, the basis of the OMSM. Instead, we found that a small number of apparently
trisomic oocytes were false-positive results which, when accurately vetted, were shown to be artifactual.
Instead of supporting the OMSM, this study demonstrated the necessity of proper controls and showed
that the maternal age effect is not due to errors occurring during pre-meiotic mitotic segregation.
In Chapter III we examined the possible connection between abnormal meiotic recombination and
the timing of meiotic entry and maternal age-dependent nondisjunction. A long standing model of the age
effect, the Production Line Model [3], proposed that the first oocytes to enter meiosis in the fetal ovary are
genetically the “best” (i.e., have normal levels of recombination), and are consequently the most likely to
generate chromosomally normal gametes. The Hassold-Hunt laboratory’s access to prophase oocytes
allowed us to correlate the recombination rates of oocytes with the time at which they entered meiosis.
We found no evidence for a relationship between the timing of meiotic entry and recombination rate.
Therefore, our studies show no support for the Production Line Model. Instead, the determination of
recombination rates across a broad range of individuals demonstrated the extreme variability in human
female recombination, regardless of time of entry into meiosis. This variability leads to abnormal
recombination levels in some oocytes, one of many factors that presumably contribute to the maternal
age effect.
80
In conclusion, these studies examined and disproved previous, popular models of the maternal
age effect. They also provided general observations at a genome wide level across numerous
individuals. Our observations of human prophase stage oocytes show that recombination occurs in a
diverse fashion in the human female, and is likely the first of many hurdles in the formation of a
chromosomally normal oocyte.
Future Directions:
Though aneuploidy has been the subject of intense scrutiny, we still do not fully understand its
complexities. Advanced technologies have increased our understanding of the incidence of aneuploidy,
but two key questions still remain: (1) why are human oocytes so susceptible to nondisjunction and (2)
what is the mechanism of the maternal age effect? The abnormally high risk of human oocyte
nondisjunction is likely due to an unknown evolutionary factor, but increasing evidence suggests there is
no single path that leads to the maternal age effect. Instead, studies show that there are likely multiple
routes to aneuploidy, acting at various time points in oocyte development, and further studies must be
performed at each of these stages to fully understand the age effect.
Prophase: Setting up oocytes for failure?
Prophase is essential for meiosis, because it brings together and links the homologous
chromosomes. Multiple studies have shown that disruptions in prophase can lead to meiotic failure,
whether it be by disrupting pairing and synapsis (e.g. [4-7]) or by disrupting recombination (e.g. [810]). Recent studies performed in the Hassold-Hunt laboratory have shown that female meiosis is much
more variable than the male counterpart [11]. Indeed, studies in Chapter III show the extraordinary range
of recombination rate in the human female. This broad range means that some oocytes have abnormally
low recombination rates, leading to a higher chance of a chromosome without an exchange, a defect that
greatly increases the risk of nondisjunction. These data beg the question: if the consequences of
abnormal recombination are so dire, what mechanism regulates crossover number and placement, and
why is it so “dysregulated” in human oocytes?
81
There have been numerous studies on the regulation of various steps in the recombination
pathway, ranging from changes in formation of DSBs, to changes in how those breaks are resolved [12,
13]. The process by which varying numbers of DSBs are repaired into similar numbers of crossovers at
the expense of non-crossovers is called “crossover homeostasis”, but the pathway behind this
mechanism remains unclear. Observations in the mouse have shown that changing the amount of DSBs
on the same strain via inhibiting or overexpressing Spo11 still results in similar crossover numbers [12],
but studies in mouse and human show that different recombination rates between strains correspond with
higher initial DSB numbers, suggesting that processes at or before DSB formation also regulate
recombination rate [11, 13]. This apparent contradiction implies that multiple factors influence
recombination, wherein higher DSB numbers increase the potential for higher crossover formation, while
changes in homeostasis cause significant variation in final crossover number. Both of these changes
may be related to chromosome compaction, since human oocytes undergo meiosis in an
undermethylated and, presumably, relatively loose chromatin state.
Recent studies in C. elegans have observed SC length differences around crossover sites,
suggesting altered chromatin compaction, evidenced by the increase in SC length on spreads, and
potentially altered crossover homeostasis [14]. If changes in SC length link chromatin compaction and
recombination, it brings about an “order of events” question; i.e., does an altered compaction state
designate crossover sites, or do crossover sites locally alter compaction?
We have initiated preliminary studies to address these questions, conducting experiments similar
to those in the initial C. elegans report [14]. In their study, Libuda et al. [14] utilized IF to visualize the SC,
crossover sites, and the HIM-8 locus, a fixed genomic point [14]. They measured the length of the SC
between the HIM-8 locus and the end of the chromosome, and when they compared regions with and
without crossovers, they observed a local increase in SC length in regions housing exchanges [14]. To
extend these studies, we have started a similar experiment, asking if this relationship also exists in
human female meiosis. We immunostained meiotic prophase preparations for the SC-associated
proteins SYCP1 and SYCP3 and for the crossover-associated protein MLH1, allowing us to
simultaneously monitor SC morphology and crossing-over. Subsequently, we used FISH to examine
specific genomic regions on the short arm of chromosome 16, chosen because it is the chromosome with
82
the highest aneuploidy rate [15]. Our approach was straightforward: to investigate the relationship
between recombination and chromosome compaction, we combined FISH to BACs spaced evenly along
16p (Table 1) with our IF images to examine SC length in microns and MLH1 marked recombination sites
and pachytene stage oocytes. Preliminary studies were conducted on three human fetal ovarian
samples, which, when compared to other human samples previously studied, had relatively “average”
levels of synaptic defects and genome-wide recombination. Following IF and FISH, we were able to
overlay the two images, and therefore determine the location of specific genomic loci along the
synaptonemal complex (Figure 1). Initially, we examined four regions on the short arm of chromosome
16 (i.e., 0-5 Mb from the p arm telomere, 5-15 Mb, 15-25 Mb, and 25-35 Mb; see Table 2). Perhaps not
surprisingly, the relationship between physical distance and protein complex distance was not a linear
relationship; it varied by region and also varied among individuals. For example, the 15-25 Mb and 25-35
Mb regions would be roughly the same physical distance, about 10 Mb. If there were indeed a simple
linear relationship between SC length and physical distance, one would expect these regions to have a
similar length. However, we found that in two of the three cases (313 and 331) the 25-35 Mb region was
significantly longer than the 15-25 Mb region (3.43 microns vs 2.35 microns in case 313; 2.85 microns vs
2.14 microns in case 331; Table 2; p<0.0001 for each).
We then asked whether, as in C. elegans, recombination sites locally alter SC length. In each of
the three individuals, we examined the four genomic regions. Due to large variability between cells,
measurements were normalized as a percentage of the p-arm length. The most distal region (i.e., 0-5
Mb) was uninformative, since crossovers were rarely identified there. However, for each of the other
three regions, and in each of the three fetal ovarian samples, we observed increases in SC length in
regions containing an MLH1 focus (Figure 2). Further studies are underway to increase sample size, and
to confirm or refute these intriguing preliminary observations.
After characterizing the relationship between MLH1 and compaction, the next steps would be to
examine recombination proteins acting immediately upstream of crossovers, e.g., RNF212 or MSH4. It is
possible that compaction designates crossover sites, leading one to expect local expansion around a
subset of MSH4 sites, with those destined to become crossovers having local expansion and those
destined to become noncrossovers being more compact. Understanding the role of RNF212 would be
83
more complicated, as RNF212 initially co-localizes with all MSH4 sites but is progressively removed by
HEI10 from future noncrossover sites [16, 17]. Therefore, one would expect to observe a similar RNF212
localization pattern as observed with MSH4 sites, but late RNF212 sites would exhibit local expansion
observed around MLH1 sites.
In contrast, if crossover sites alter local chromatin compaction, one would expect completely
different observations at the sites of upstream recombination proteins. For example, if the recombination
machinery prevents compaction, all MSH4 sites would likely display similar compaction levels. Because
subsets of those sites were repaired by the recombination machinery as non-crossovers, either
compaction around those sites would occur, or expansion might occur around crossover sites. Due to the
significant decrease in SC length between the zygotene and pachytene stages, it is likely that compaction
occurs during this transition, and that the recombination machinery prevents local compaction around
sites designated as crossovers.
The Hassold-Hunt laboratory’s studies on chromosome compaction also raise interesting
questions on the spacing of crossover sites, i.e., crossover interference (reviewed in [18]). While
interference has been characterized in numerous species (e.g. [19-22]) and mathematical models of
interference abound [23-25], the underling mechanisms remain obscure. However, several properties of
interference are clear: it functions across physical distances (microns of SC length) rather than genomic
distance (megabases) [26, 27]; it is disrupted if the SC does not fully form [21], and it appears to behave
similarly in numerous organisms [23]. The existence of chromatin modifications that occur around
crossover sites would provide a potential physical signal that may cause interference. Also, since these
chromatin modifications alter the SC and the SC is required for proper interference, it suggests that the
signals may be propagated along the SC, giving rise to the physical distance that interference acts over.
Several interesting questions would result if physical stress along the SC causes interference.
First, if local chromatin expansion causes interference, does disrupting chromatin compaction alter
interference? For example, it is thought that compaction of chromosomes in meiosis is related to the
condensin proteins [28], though little is known about them in mammalian systems. If condensin is
responsible for compaction, can altering the functionality of condensin result in decreased meiotic
chromosome condensation? The mouse kleisin beta condensin subunit knockout would be a useful tool
84
for examining condensation and interference [29]. The mouse line is fertile, but the meiotic process has
not been examined. If compaction is disrupted, it likely disrupts interference. Condensin knockouts in C.
elegans lead to increased crossover number and increased SC length [28]. However, assessing
interference in C. elegans is more difficult, as wild-type individuals display a single crossover per
chromosome, or total interference. The fact that the condensin knockout causes multiple crossovers per
chromosome suggests disruption of interference, though it is unclear if similar results would be observed
in mammalian models. Disruption of compaction would likely cause an increase in SC length and
recombination rate, but whether a simultaneous decrease in interference would be observed remains
unclear. Abnormally low chromatin compaction would likely decrease interference, (e.g. increase in
recombination rate is larger than expected increase from change in SC length). However, since microns
of SC length is the measure by which interference occurs, it would also be possible to observe
proportional increases in SC length and recombination rate, and therefore no change in interference.
Studying interference more stringently will illuminate how crossover sites are determined, which will lead
to insight into the association of abnormal crossover placement and aneuploidy.
Dictyate Arrest: Waiting to break down?
After recombination occurs, the oocyte enters an extended arrest period. In this period, the
meiotic chromosomes remain dormant, held together by cohesin proteins, and resume meiosis only upon
ovulation. This window can last decades in the human female, and therefore provides an extended
window in which errors could occur. Studies of knockout mice and observational studies in the human
have shown that cohesin proteins can degrade during this arrest stage, and might not be replaced (e.g.
[30-33]). Cohesin also provides a link between distal recombination events and aneuploidy, as distal
recombination sites would have a smaller region of the homologous pair held together. Degradation of
cohesin has increasingly become a favored explanation for the maternal age effect, but is likely only a
part of the whole picture. For example, it fails to explain the spike in aneuploidy in pregnancies involving
extremely young women, as well as the fact that there is variation in maternal age-related incidence rates
in different trisomies (e.g. trisomy 16 displays a linear increase, while most others exhibit an exponential
increase; [15]).
85
Recent studies in Drosophila have discovered the presence of a cohesin rejuvenation system
[34], providing an interesting target of research in higher organisms. To date, no rejuvenation of mouse
cohesins has been observed. Providing replenishment or preventing degradation of the cohesin proteins
are promising candidates for development of a therapeutic, which is the ultimate goal of maternal age
effect research. While this is unlikely to be achieved in the near future, it may not be that far off. Perhaps
the Drosophila rejuvenation system could be reengineered in a mouse model, or possibly vertebrates
have that system, but it remains undiscovered as of yet. One could envision a vertebrate rejuvenation
system that breaks down with age, giving rise to the observed cohesin loss. Regardless, future studies
will likely implicate cohesin as a part of the complex puzzle that is the maternal age effect.
Segregation: Where it all goes awry
The unifying factor of aneuploidy is that regardless of the origin of the predisposing risk factors
the actual chromosome errors all happen at the same time, when the chromosomes divide. This does not
mean that the root cause of all errors is in the meiotic divisions; rather chromosomes would separate
improperly during the meiotic divisions, resulting in aneuploidy. It is abundantly clear that female
gametes are much more error-prone, leading researchers to focus on the difference between male and
female meiotic divisions. It is clear that male gametes have a much more stringent anaphase checkpoint
than females, triggering apoptosis in cells that would slip through in the female. This sex difference in
checkpoint stringency could be due to lack of checkpoint signaling due to degradation of checkpoint
proteins in the long female meiotic arrest stage [35], or possibly because females lack certain genes that
may be essential to the process [36-38]. The fact that error containing oocytes survive elicits an
interesting questions for the segregation of chromosomes: do the segregation mechanisms get worse, or
are the oocytes that divide later of a worse quality due to multiple factors across earlier timepoints,
causing the checkpoint segregation system to be compromised. Overall, there are numerous avenues of
future research on the maternal age effect, and plenty of work to be done to understand this complex
puzzle.
86
REFERENCES
1.
Hulten, M.A., et al., On the origin of the maternal age effect in trisomy 21 Down syndrome: the
Oocyte Mosaicism Selection model. Reproduction, 2010. 139(1): p. 1-9.
2.
Hulten, M.A., et al., On the origin of trisomy 21 Down syndrome. Mol Cytogenet, 2008. 1: p. 21.
3.
Henderson, S.A. and R.G. Edwards, Chiasma frequency and maternal age in mammals. Nature,
1968. 218(5136): p. 22-8.
4.
Yuan, L., et al., The murine SCP3 gene is required for synaptonemal complex assembly,
chromosome synapsis, and male fertility. Mol Cell, 2000. 5(1): p. 73-83.
5.
Yuan, L., et al., Female germ cell aneuploidy and embryo death in mice lacking the meiosisspecific protein SCP3. Science, 2002. 296(5570): p. 1115-8.
6.
Yang, F., et al., Mouse SYCP2 is required for synaptonemal complex assembly and chromosomal
synapsis during male meiosis. J Cell Biol, 2006. 173(4): p. 497-507.
7.
Hamer, G., et al., Disruption of pairing and synapsis of chromosomes causes stage-specific
apoptosis of male meiotic cells. Theriogenology, 2008. 69(3): p. 333-9.
8.
Baudat, F., et al., Chromosome synapsis defects and sexually dimorphic meiotic progression in
mice lacking Spo11. Mol Cell, 2000. 6(5): p. 989-98.
9.
Romanienko, P.J. and R.D. Camerini-Otero, The mouse Spo11 gene is required for meiotic
chromosome synapsis. Mol Cell, 2000. 6(5): p. 975-87.
10.
Edelmann, W. and R. Kucherlapati, Role of recombination enzymes in mammalian cell survival.
Proc Natl Acad Sci U S A, 1996. 93(13): p. 6225-7.
11.
Gruhn, J.R., et al., Cytological studies of human meiosis: sex-specific differences in recombination
originate at, or prior to, establishment of double-strand breaks. PLoS One, 2013. 8(12): p.
e85075.
12.
Cole, F., et al., Homeostatic control of recombination is implemented progressively in mouse
meiosis. Nat Cell Biol, 2012. 14(4): p. 424-30.
13.
Baier, B., et al., Variation in genome-wide levels of meiotic recombination is established at the
onset of prophase in mammalian males. PLoS Genet, 2014. 10(1): p. e1004125.
14.
Libuda, D.E., et al., Meiotic chromosome structures constrain and respond to designation of
crossover sites. Nature, 2013. 502(7473): p. 703-6.
15.
Hassold, T.J. and P.A. Jacobs, Trisomy in man. Annu Rev Genet, 1984. 18: p. 69-97.
16.
Reynolds, A., et al., RNF212 is a dosage-sensitive regulator of crossing-over during mammalian
meiosis. Nat Genet, 2013. 45(3): p. 269-78.
87
17.
Qiao, H., et al., Antagonistic roles of ubiquitin ligase HEI10 and SUMO ligase RNF212 regulate
meiotic recombination. Nat Genet, 2014. 46(2): p. 194-9.
18.
Berchowitz, L.E. and G.P. Copenhaver, Genetic interference: don't stand so close to me. Curr
Genomics, 2010. 11(2): p. 91-102.
19.
Sturtevant, A.H., A Third Group of Linked Genes in Drosophila Ampelophila. Science, 1913.
37(965): p. 990-2.
20.
Petkov, P.M., et al., Crossover interference underlies sex differences in recombination rates.
Trends Genet, 2007. 23(11): p. 539-42.
21.
Hillers, K.J. and A.M. Villeneuve, Chromosome-wide control of meiotic crossing over in C.
elegans. Curr Biol, 2003. 13(18): p. 1641-7.
22.
Zhang, L., et al., Interference-mediated synaptonemal complex formation with embedded
crossover designation. Proc Natl Acad Sci U S A, 2014. 111(47): p. E5059-68.
23.
Zhang, L., et al., Crossover patterning by the beam-film model: analysis and implications. PLoS
Genet, 2014. 10(1): p. e1004042.
24.
King, J.S. and R.K. Mortimer, A polymerization model of chiasma interference and corresponding
computer simulation. Genetics, 1990. 126(4): p. 1127-38.
25.
Lande, R. and F.W. Stahl, Chiasma interference and the distribution of exchanges in Drosophila
melanogaster. Cold Spring Harb Symp Quant Biol, 1993. 58: p. 543-52.
26.
Drouaud, J., et al., Sex-specific crossover distributions and variations in interference level along
Arabidopsis thaliana chromosome 4. PLoS Genet, 2007. 3(6): p. e106.
27.
Billings, T., et al., Patterns of recombination activity on mouse chromosome 11 revealed by high
resolution mapping. PLoS One, 2010. 5(12): p. e15340.
28.
Mets, D.G. and B.J. Meyer, Condensins regulate meiotic DNA break distribution, thus crossover
frequency, by controlling chromosome structure. Cell, 2009. 139(1): p. 73-86.
29.
Gosling, K.M., et al., A mutation in a chromosome condensin II subunit, kleisin beta, specifically
disrupts T cell development. Proc Natl Acad Sci U S A, 2007. 104(30): p. 12445-50.
30.
Hodges, C.A., et al., SMC1beta-deficient female mice provide evidence that cohesins are a
missing link in age-related nondisjunction. Nat Genet, 2005. 37(12): p. 1351-5.
31.
Lister, L.M., et al., Age-related meiotic segregation errors in mammalian oocytes are preceded by
depletion of cohesin and Sgo2. Curr Biol, 2010. 20(17): p. 1511-21.
32.
Tachibana-Konwalski, K., et al., Rec8-containing cohesin maintains bivalents without turnover
during the growing phase of mouse oocytes. Genes Dev, 2010. 24(22): p. 2505-16.
88
33.
Chiang, T., et al., Evidence that weakened centromere cohesion is a leading cause of age-related
aneuploidy in oocytes. Curr Biol, 2010. 20(17): p. 1522-8.
34.
Weng, K.A., C.A. Jeffreys, and S.E. Bickel, Rejuvenation of meiotic cohesion in oocytes during
prophase I is required for chiasma maintenance and accurate chromosome segregation. PLoS
Genet, 2014. 10(9): p. e1004607.
35.
Steuerwald, N., et al., Association between spindle assembly checkpoint expression and maternal
age in human oocytes. Mol Hum Reprod, 2001. 7(1): p. 49-55.
36.
Vernet, N., et al., The expression of Y-linked Zfy2 in XY mouse oocytes leads to frequent meiosis 2
defects, a high incidence of subsequent early cleavage stage arrest and infertility. Development,
2014. 141(4): p. 855-66.
37.
Vernet, N., et al., Mouse Y-linked Zfy1 and Zfy2 are expressed during the male-specific interphase
between meiosis I and meiosis II and promote the 2nd meiotic division. PLoS Genet, 2014. 10(6):
p. e1004444.
38.
Vernet, N., et al., Spermatid development in XO male mice with varying Y chromosome shortarm gene content: evidence for a Y gene controlling the initiation of sperm morphogenesis.
Reproduction, 2012. 144(4): p. 433-45.
89
Table I - Location of FISH Probes on Short Arm of Chromosome 16
Probe Location
BAC ID No.
5Mb
RP11-469N12
Genomic Location (in bp; ptelomere=0)
4,446847-4,633,469
15Mb
RP11-81L19
15,519,957-15,681,446
25Mb
RP11-185O21
25,660,176-25,795,340
35Mb
Control 16 (centromere probe)
~34,500,000
90
Table II - Average Length of Measured Regions (in Microns)
Case
0-5 Mb
5-15 Mb
15-25 Mb
25-35 Mb
313 (n=25)
0.99 +/- .38
2.33 +/- .41
2.35 +/- .54
3.44 +/- .78
331 (n=17)
1.07 +/- .23
2.19 +/- .54
2.14 +/- .45
2.85 +/- .62
333 (n=14)
1.25 +/- .34
2.58 +/- .38
2.89 +/- .61
2.94 +/- .80
91
Figure 1 - Use of BAC-FISH to mark specific genomic locations
(A) IF for SYCP3 (blue), SYCP1 (red) and MLH1 (green), and (B) subsequent BAC-FISH for our probes
(in green) along chromosome 16 (5Mb, 15Mb, 25Mb, and 35Mb) identified based on their relative
distance from the p-terminus (p-term, white arrow) of the chromosome.
92
Figure 1
93
Figure 2 - Local expansion of SC in regions containing crossover sites
In each of our regions (25-35, 15-25, 5-15), regions were group as either having an MLH1 focus (green)
or lacking a focus (red). For all cases and all regions, we see an increase in % of p-arm length when an
MLH1 focus is present, suggesting local chromatin expansion in regions around MLH1 sites.
94
Figure 2
95