Molecular Human Reproduction, Vol.15, No.12 pp. 805–819, 2009
Advanced Access publication on July 7, 2009 doi:10.1093/molehr/gap055
NEW RESEARCH HORIZON Review
Oogenesis and cell death in human
prenatal ovaries: what are the criteria
for oocyte selection?
G.M. Hartshorne1,6, S. Lyrakou2, H. Hamoda 3,
E. Oloto 4, and F. Ghafari 5
1
Clinical Sciences Research Institute, Warwick Medical School, University of Warwick, Coventry CV2 2DX, UK 2EuroGene Genetics Lab Ltd,
21 –23 Gounari Street, 18531 Piraeus, Greece 3Queen Charlotte’s Hospital, Du Cane Road, Greater London W12 0HS, UK 4British
Pregnancy Advisory Service, 20 Timothy’s Bridge Road, Stratford-upon-Avon, Warwickshire CV37 9BF, UK 5Luton Institute for Research in
Applied Natural Sciences, 250 Butterfield, Great Marlings, Luton LU2 8DL, UK
6
Correspondence address. E-mail: [email protected]
abstract: Prenatal oogenesis produces hundreds of thousands of oocytes, most of which are discarded through apoptosis before birth.
Despite this large-scale selection, the survivors do not constitute a perfect population, and the factors at the cellular level that result in apoptosis or survival of any individual oocyte are largely unknown. What then are the selection criteria that determine the size and quality of the
ovarian reserve in women? This review focuses on new data at the cellular level, on human prenatal oogenesis, offering clues about the
importance of the timing of entry to meiotic prophase I by linking the stages and progress through MPI with the presence or absence of
apoptotic markers. The characteristics and responsiveness of cultured human fetal ovarian tissue at different gestational ages to growth
factor supplementation and the impact of meiotic abnormalities upon apoptotic markers are discussed. Future work will require the use
of a tissue culture model of prenatal oogenesis in order to investigate the fate of individual live oocytes at different stages of development.
Key words: oogenesis / fetal / oocyte / apoptosis / meiosis
Introduction
Hypothesis
The quality of mature oocytes is frequently suboptimal in humans.
Despite only a small proportion of a large prenatal population
being selected to continue development, many surviving oocytes
carry anomalies that may impair their ultimate function as gametes
(Hunt and Hassold, 2008). There have been many attempts to
explain the biological rationale for such apparently wasteful cellular
attrition in the germline; for example, sub-lethal anomalies could contribute to the potential for variation and evolutionary success, and
non-surviving cells could donate crucial components to be recycled
by survivors (Baum et al., 2005). However, none of the ideas proposed can explain the massive scale and apparent lack of efficiency
of the selection process. A large majority of oocytes are lost
before birth, and most of the rest before puberty. The remaining
oocytes are not a perfectly viable population and so the criteria
upon which they have been selected remain unclear. In this review,
we focus on what is known in this area and add some additional
insights from our ongoing research.
The purpose of our work is to understand the bases upon which
oocytes are selected over the prolonged period from their prenatal
formation, when oogonia enter meiosis, to their eventual demise, or
exceptionally, their ovulation during reproductive adulthood. Our
current hypothesis is that prenatal loss of oocytes reflects selection
at the cellular level, for or against cell death, on the basis of certain
measurable criteria. Accordingly, we hypothesize that events at the
cellular level must be coordinated to result in the quantitatively reliable
changes that eventually result in a relatively consistent population of
oocytes in the ovarian reserve. At present, we are studying how prenatal oocytes are selected for further development or apoptosis at an
individual oocyte level.
Our concept of selection also infers a deliberate process rather than
random chance. Our idea is that a cellular decision is made, to survive
or die, in association with or in response to the prevailing circumstances (a bottom-up approach). The alternative approach might be
that a set of controlling measures originates from a higher level, for
& The Author 2009. Published by Oxford University Press on behalf of the European Society of Human Reproduction and Embryology. All rights reserved.
For Permissions, please email: [email protected]
806
example, a requirement for an ovarian reserve of a particular size (a
top-down approach). Biomarkers that reflect the ovarian reserve,
such as the concentration of anti-Mullerian hormone, or response
to gonadotrophin-releasing hormone challenge, are sometimes
measured clinically in order to estimate the ovarian reserve of a
woman at risk of infertility (Coccia and Rizzello, 2008). Such
methods are particularly relevant where there is a known likelihood
of reduced numbers of oocytes, for example, owing to advanced
age, or chemotherapy. However, definitions by which to diagnose
‘normal’ or ‘abnormal’ on the basis of these markers have not been
achieved and no clinically feasible means of slowing the age-related
decline of the ovarian reserve has yet emerged. We have taken the
contrasting bottom-up approach because the size of the ovarian
reserve seems remarkably resistant to alteration owing to adverse
external circumstances arising from the maternal environment, such
as undernutrition (Lumey and Stein, 1997), which are capable of
affecting many other developmental processes (Barker and Clark,
1997). Therefore, we are looking for factors local to the fetal ovary,
as a first step towards understanding oocyte selection for both
quality and quantity.
Overview of oocyte selection
Selection for survival or cell death within the ovary can occur at several
stages of oocyte development. Selection during germ cell migration to
the ovary is not considered here. This article mainly addresses the first
two stages listed below, when selected oogonia enter meiosis and
progress through meiotic prophase I (MPI) to arrest at the diplotene
stage. However, an overview is provided to place these in context.
This article refers to humans throughout, unless otherwise stated.
We have focused on humans since important differences from
model animals have been noted. A timeline for human prenatal oogenesis is presented in Fig. 1 showing the progressive but often overlapping stages. The following are important stages in oogenesis:
(i)
(ii)
(iii)
(iv)
(v)
oogonia to oocyte transition (entry to MPI);
MPI;
breakdown of oocyte nests;
enclosure by somatic cells to form primordial follicles;
timing of initiation of early follicle growth (or failure of timely
arrest of follicle growth at the primordial stage);
(vi) follicular growth and competition leading towards ovulation in
adulthood.
(i) Cell death is quantitatively a major activity of prenatal oocytes,
particularly in the mid-trimester for humans (Vaskivuo et al., 2001)
and from around 13 dpc in mice (Lobascio et al., 2007). This contrasts
with the preceding stage of active expansion of the germline, where
primordial germ cells (PGCs) migrate to the ovary and replicate to
form diploid oogonia, which continue to undergo mitosis (Kurilo,
1981). Apoptotic cell death does occur in PGCs and oogonia,
notably via mechanisms that are responsive to the presence of survival
factors such as stem cell factor (SCF) (also known as kit ligand)
(Krysko et al., 2008) and possibly oxidative stress (Murdoch et al.,
2003, sheep), but the overall result in vivo at this time is the enlargement of the precursor cell population. This stage has been reviewed
previously (De Felici et al., 2005; Oktem and Oktay, 2008).
Hartshorne et al.
(ii) Oogonia enter MPI and are now termed ‘oocytes’. Oocytes progress through MPI, via the leptotene, zygotene and pachytene stages
towards the diplotene stage, to enter meiotic arrest at that point.
Extensive apoptosis occurs as oocytes progress through MPI, leading
to the loss of many of them. In humans, meiosis is not synchronous
(Gondos, 1987), and even in species such as the mouse, where
more synchrony is evident, there is still variation between neighbouring oocytes in terms of meiotic progression. The local signals controlling the survival of oogonia and oocytes in early MPI are not clear,
although much progress has been made towards understanding the
female germline in mice (e.g. Farini et al., 2005).
(iii) Oocytes entering meiosis are normally found in so-called
‘nests’. These clustered oocytes, generally lacking closely associated
pre-granulosa cells, are believed to arise from a group of daughter
cells derived from the same dividing oogonium. The oocytes within
a nest communicate via intercellular bridges and also tend to maintain
more synchrony within, compared with between, nests (Gondos,
1987; Pepling, 2006, mouse). However, inevitably, not all of the
oocytes of a nest survive to form follicles, and it has been proposed
that some of the oocytes may die altruistically to ensure that one or
two from a particular nest can survive (Pepling and Spradling, 2001,
mouse). The factors that determine which oocytes from a nest are
selected for survival are unclear, but may represent one level of
coordination that could eventually result in a suitably sized ovarian
reserve. Estrogen may be a mediator in this process (Zachos et al.,
2002).
(iv) Oocytes continuing through meiosis to complete the diplotene
stage (or perhaps earlier in MPI, Ohno et al., 1962; Yang and Fortune,
2008, cattle) can recruit somatic pre-granulosa cells to surround them,
resulting in primordial follicles. The formation of primordial follicles
can be influenced by genetic manipulation, for example, knockout
mice illustrate that several genes influence oocyte survival or apoptosis, or nest breakdown, thereby affecting the numbers of follicles
formed (e.g. bcl-2, Ratts et al., 1995, mice; caspase 2, Bergeron
et al., 1998, mice; bax, Perez et al., 1999, mice). Bax and bcl-2 are
active in human fetal ovaries (Albamonte et al., 2008), but many of
the potential effectors identified in mice have not yet been studied
in humans. Primordial follicles have been reported from as early as
13 weeks of gestation in humans (Forabosco and Sforza, 2007), but
are generally considered to become abundant from around 16 –20
weeks of gestation (Rabinovici and Jafe, 1990; Albamonte et al.,
2008). Changes in gene expression in human fetal ovaries around
the time of primordial follicle assembly are now providing clues
about the pathways controlling early folliculogenesis (Fowler et al.,
2009). For recent reviews of the timing and control of primordial follicle assembly, see Skinner (2005), or Maheshwari and Fowler (2008).
(v) If the resulting primordial follicles are to participate in reproductive life, they must remain in the primordial stage for several years, at
least until puberty. In this state, the oocytes remain arrested in the
diplotene stage of MPI, and the surrounding granulosa cells support
the oocyte’s metabolic and signalling requirements (for example, see
Matzuk et al., 2002; Gilchrist et al., 2008). However, a proportion
of oocytes and follicles either do not become, or do not stay, arrested
(probably by escaping local inhibitors, Wandji et al., 1997, baboon;
Yang and Fortune, 2008, cattle), or they initiate growth during prenatal
or prepubertal life. These growing follicles are a consistent feature of
ovaries during childhood (Peters et al., 1976). They always become
807
Oogenesis and cell death in prenatal ovaries
Figure 1 Timeline for human prenatal oogenesis. The figure is a diagrammatic representation of the timings of major prenatal events in oogenesis,
including: primordial germ cell (PGC) migration and oogonial proliferation, entry to meiosis and the stages of meiotic prophase 1 (MP1), nest formation
and breakdown, folliculogenesis, germ cell apoptosis and the numbers of germ cells at certain stages. The nuclear patterns shown at each stage of MPI
are symbolic of synaptonemal complex (SC) conformations. For ease of illustration, nuclei are not always shown. Shading indicates cell death. The
diagram draws upon the published literature; however, where data are conflicting or unavailable, informed estimates have been made. Particular reference has been made to the following publications: Lintern Moore et al. (1974); Peters et al., (1978); Motta and Makabe (1986); Carlson (2004);
Cummings and Kavlock (2004); Bendsen et al., (2006); Maheshwari and Fowler (2008).
atretic due to the prevailing immaturity of the hypothalamo-pituitary
axis (Djahanbakhch et al., 2007). Their function is unclear; however,
a significant consequence is to reduce the ovarian reserve from
around 700 000 at birth to 300 000 at puberty (Faddy et al., 1999).
In mice, follicle growth begins only after birth (in contrast to
humans). In mice, specific deletion of Pten in oocytes caused the premature activation of the primordial follicle pool, first evident at postnatal day 8 and resulting in complete elimination of all follicles by about
3 months (Reddy et al., 2008). The possible role of Pten in human fetal
ovaries has not yet been studied.
Also in mice, there is an interesting shift in the predominant mechanism of oocyte loss from apoptosis before folliculogenesis, to multiple
pathways including non-apoptotic mechanisms, causing loss of primordial follicles (Rodrigues et al., 2009; Tingen et al., 2009). The critical
role of the somatic cell fraction is barely understood at this early
stage, but must have major impacts, most obviously as pre-granulosa
cells engage intimately with the oocyte during folliculogenesis, their
cytoplasmic extensions forming junctions with the oolemma.
(vi) Follicles that initiate growth during adulthood may be capable of
yielding a mature fertilizable oocyte, depending on the endocrine
environment at the time of their growth. The process of selection
of the ovulating follicle is competitive, and apoptotic processes in
the adult ovary have been studied extensively (Hussein, 2005).
This article will highlight some emerging concepts on oocyte selection during prenatal oogenesis, shedding light on the interactions
between apoptosis and timing of meiosis, using data acquired from
human tissues, where possible.
Oogonia to oocyte transition:
timing of entry into MPI
This transition is marked by entry into meiosis of diploid oogonia
derived from the PGCs. In humans, estimates of the proportion of
oogonia that enter meiosis vary widely (50%, Baker, 1963; 90%,
Kurilo, 1981). Using data from McClellan et al. (2003) in CD-1
808
mice, we estimate that at most 80% of oogonia make this transition.
Genetic mutant mice that lack bax, a pro-apoptotic gene, have excessive numbers of oocytes entering meiosis, suggesting this gene’s involvement in oogonial selection (Alton and Taketo, 2007). Oogonia that
do not enter meiosis are removed either by apoptosis (Albamonte
et al., 2008) or by arrest in a metaphase-like conformation associated
with an unusual form of cell death (Wartenberg et al., 2001). Retinoic
acid (RA) may be the critical factor whose presence induces meiosis in
females, but whose relative absence protects against premature
meiosis in males (Bowles and Koopman, 2007, mice). Oogonia
located towards the medulla in humans are the first to enter MPI
(Bendsen et al., 2006) and the most advanced during fetal folliculogenesis, initiating follicle formation and growth first (Byskov and Lintern
Moore, 1973, mice). In mice and other species studied, meiosis
begins in the part of the ovary that is in closest proximity to the mesonephros or the rete ovarii, derived from mesonephros cells migrating
into the gonad (Bowles and Koopman, 2007). It is likely that the timing
and geographical organization of entry into meiosis reflect the location
and origins of RA, which originates from the mesonephros or rete
ovarii, and which is a small diffusible molecule whose gradient may
have a vital role in the organization and timing of the onset of MPI at
the ovarian level (Bowles and Koopman, 2007, mice). RA has not yet
been studied in human fetal ovaries, where a similar role may be anticipated; however, additional local control will be necessary to account
for the variability in entry to meiosis among adjacent oocytes, which
would not be expected if RA gradient were the sole mediator.
A possible influence of the timing of meiotic entry upon later oocyte
quality has long been suspected (Jones, 2008). The timing of entry to
meiosis has been proposed as one possible mechanism by which the
maternal age-associated increase in chromosomal abnormalities in
children could arise. The so-called ‘first-in, first-out’ hypothesis links
the sequence of oocyte formation with utilization, proposing that
those formed later are utilized later in life, resulting in increasing
abnormalities as the age of the mother increases (Henderson and
Edwards, 1968; Polani and Crolla, 1991). Other hypotheses are
based upon ideas on accumulated oxidative damage (Harman,
1956), telomere function (Keefe et al., 2006) and, most recently,
ovarian mosaicism (Hultén et al., 2008); however, the origin of the
maternal age effect is an enduring enigma that may only be elucidated
by a greater understanding of prenatal events in oogenesis.
We are interested in how human oogonia may be selected to enter
MPI. We have therefore investigated the timing of entry to MPI, with a
view to understanding whether the timing is important for subsequent
utilization or apoptosis of oocytes.
Entry into meiosis requires the availability of precursor cells. We
checked this in primary explants of fetal human ovaries by alkaline phosphatase (AP) staining of cells manually spread on glass slides. AP is present
in PGCs and oogonia, but is reduced and eventually lost as oocytes enter
meiosis. In humans, strongly AP-positive precursor cells were evident at
least as late as 19 weeks of gestation (Fig. 2), long after MPI has begun
in many cells. The arrest of the remaining oogonia in a metaphase-like
state, which probably originates as an aborted entry into meiotic prophase, appears to occur in humans at around 23–26 weeks (Kurilo,
1981), so some of the persistent oogonia may be capable of making a
late but successful transition into MPI. Using a meiosis-specific marker
for synaptonemal complex (SC) proteins, we found that zygotene-stage
oocytes were evident from 14 weeks, and pachytene-stage oocytes
Hartshorne et al.
from 15 weeks of gestation (Hartshorne et al., 1999). However, many
oocytes in these early stages of MPI were present until at least 19
weeks of gestation (as shown in Fig. 1 and in agreement with Kurilo,
1981; Speed, 1985; Roig et al., 2005a, b).
In order to test whether oocytes entering MPI earlier or later have
an increased likelihood of being selected for cell death, we used terminal deoxynucleotidyltransferase-mediated dUTP nick-end labelling
(TUNEL) as a marker of DNA breaks in oocytes at different stages
of MPI at known gestational ages. TUNEL highlights breaks in DNA
that occur during the later stages of apoptosis. It is not a specific
marker of apoptosis, but may offer insights into the state of DNA.
TUNEL was used successfully as a correlate of cell death in human
oocytes by Modi et al. (2003), who found that 3 –7% of oocytes
were apoptotic between weeks 13– 23 in normal ovaries, rising to
.50% in Turner’s syndrome 45XO ovaries, where extensive prenatal
loss of oocytes occurs. Like Modi et al. (2003), Albamonte et al.
(2008) observed low levels (10%) of TUNEL-positive germ cells
throughout the early second trimester, but they found a higher incidence (20%) at 18 –20 weeks, which in their study coincided with
the switch from a bcl-2/bax balance to bax dominance, as bcl-2,
which inhibits apoptosis, became undetectable at this stage.
However, Abir et al. (2002) reported conflicting results, noting overexpression of bcl-2 in all samples at all ages, with TUNEL positivity
undetectable before 23 weeks.
Unfortunately, tissue sections do not allow reliable identification of
MPI stages. For this, oocyte spreads are essential. We therefore
applied TUNEL on oocytes that had been spread on slides, enabling
accurate cytogenetic determination of their stage of MPI. Spreads
confer increased accessibility of reagents because the cytoplasm has
been removed, in contrast with sections of tissues embedded in
wax or other media. Consequently, the extent of TUNEL staining in
our data and that of others using this method (DeFelici et al., 2008)
is higher than that reported for tissue sections. Figure 3 shows that
pre-leptotene oocytes are predominantly TUNEL-positive, as
expected owing to dsDNA breaks during reorganization in preparation for meiosis (Mahadevaiah et al., 2001). dsDNA breaks are
also prevalent in the leptotene stage. However, we found that leptotene, and particularly zygotene-stage oocytes that persisted at 18 and
19 weeks of gestation, were more likely to be TUNEL-positive than
those at 17, 16 and 15 weeks of gestation. TUNEL-positive oocytes
comprised at most half of the total of leptotene- and zygotene-stage
oocytes until 17 weeks. These novel data therefore tend to support
the idea that oocytes entering late into meiosis may be more likely
to succumb to DNA damage and possible demise. If the observed
TUNEL coincides with the DNA fragmentation of cell death, it is possible that oocyte elimination in human fetal ovaries may take place at
the zygotene stage rather than at the diplotene stage (as in mice).
Further investigation of human oocytes using specific markers of apoptosis is necessary to test this idea.
Our data also confirm that oocytes are found in the pre-leptotene
and leptotene stages of MPI across a wide range of gestations, supporting the idea of gradual non-synchronous entry of human oogonia into
MPI. This lack of synchrony adds complexity to the interpretation of
human data from fixed preparations. Clarity will only be achieved
when individual live oocytes can be followed through MPI in vitro.
Too few pachytene- and diplotene-stage oocytes were analysed to
draw conclusions regarding the completion of MPI.
Oogenesis and cell death in prenatal ovaries
809
Figure 2 Relative intensities of alkaline phosphatase (AP) staining in spread ovarian cells at different gestational ages. Staining intensity is represented
by þ (weak) to þþ þþ (strongest). The distribution of staining intensity remained similar across the gestational ages examined. The numbers of
fetuses examined from weeks 13 to 19 of gestation were 1, 3, 2, 2, 4, 4, 1, respectively. The absolute numbers of cells staining positive for AP declined
from gestational week 16 to 19 (data not shown).
Interestingly, a more extensive study using the same method in mice
found different results (Fig. 4, details in Ghafari et al., 2007). In mice,
oocytes reaching the pachytene stage early (16 dpc) were more likely
to be positive for TUNEL, whereas those in pachytene on Day 18
were more likely to be TUNEL-negative. These significant results contrasted with our data in humans (Fig. 3) but showed some similarities
to dsDNA break detection in spermatogenesis (Qin et al., 2004).
Co-staining for cleaved poly-(ADP-ribose) polymerase-1 (PARP-1), a
marker of apoptosis, showed that mouse oocytes in the pachytene
stage, either early (15.5 dpc) or late (18 dpc), were less likely to be
apoptotic than those on Days 16 or 17 (Fig. 4). From this, we conclude that, in mice, there may be an optimal schedule for oocyte progression, deviation from which may increase the chances of apoptosis.
We also note that the apoptotic marker, cleaved PARP-1, and TUNEL
staining were not coincident, demanding cautious interpretation. It is
also highly likely that different pathways of apoptosis are active and
their interaction and relative prevalence at different times are not
yet fully understood (DeFelici et al., 2008).
Thus, we offer some additional evidence that the fate of oocytes
may be affected by the timing of their entry into meiosis. The transition time through the stages of MPI may reflect, to some extent,
oocyte developmental competence (Roig et al., 2005a, b), and
further studies are needed to track individual oocytes in vitro and
gain detail on apoptotic decision-making at the cellular level. Although
the question of apoptotic mechanisms in vitro has not yet been
addressed for human fetal oocytes, progress has been made in mice
by DeFelici et al. (2008) using short-term cultures. Their results
clearly show that several pathways of apoptosis are active in mouse
fetal oocytes, and they also present evidence of autophagy and necrosis (DeFelici et al., 2008). Importantly, they note that inhibitors of caspases, enzymes that mediate the cascade leading to DNA destruction,
are only partially effective in inhibiting oocyte apoptosis, indicating that
caspase-independent mechanisms must also be operational. They also
confirm our observation in human oocytes that DNA fragmentation
identified by TUNEL is not always associated with apoptosis
(Ghafari et al., 2007). Currently, the interactions of the many possible
pathways of cell death involved in prenatal oocyte selection are not
understood. Their further study will prove vital in understanding the
different selection criteria applied to oocytes at a cellular level.
Mechanisms affecting oocyte
survival decisions: the changing
ovarian environment
It is unclear to what extent the balance for individual oocyte decisions
to continue development or to initiate cell death might relate to local
effects of the developing ovarian somatic cell compartment upon
oocyte survival (e.g. endocrine milieu, intercellular signalling, extracellular matrix), or to the inherent qualities of oocytes in MPI (e.g.
errors in synapsis which may delay meiotic progression or cause
arrest). We are interested in comparing these intrinsic and extrinsic
possibilities at different stages of development, thereby exploring tolerance for deviation from the normal programme of development.
A range of paracrine factors, identified in the human fetal ovary,
may affect oocyte survival, at one stage or another (e.g. Bennett
et al., 1996; Schilling and Yeh, 1999; Martins da Silva et al., 2004;
Coutts et al., 2008). For some paracrine factors, progress has been
made in identifying the originating cell type and the locations of receptors; however, the local ovarian environment is complex and constantly changing both through fetal life and subsequently. There are
a number of so-called ‘survival factors’, without which germline cells
will die, for example, SCF, which is essential during PGC migration
and for the survival of early oocytes (Moniruzaman et al., 2007,
mice; Gu et al., 2009, mice). SCF, insulin-like growth factor I (IGF-1)
and leukocyte inhibitory factor support the survival of post-migratory
germline cells (Morita et al., 1999, mice). Superimposed upon these
are the pro- and anti-apoptotic mechanisms and other cell death pathways that are functional at particular stages (see, e.g. Quenby et al.,
1999; Vaskivuo et al., 2001; Fulton et al., 2005). So, for example, in
mice, an absence of kit signalling activates the Fas cell death
pathway within pre-follicular oocytes (Sakata et al., 2003, mice), and
810
Hartshorne et al.
Figure 3 The proportions of TUNEL-positive and TUNEL-negative oocytes at different stages of MPI, as identified by SCP3 staining, in human fetal
ovaries at different gestational ages. Notice, at 18 and 19 weeks gestation, TUNEL-positive oocytes are more prevalent, in contrast to earlier ages. The
numbers of fetuses examined from weeks 13 to 19 of gestation were 1, 2, 2, 2, 2, 2, 1, respectively.
Continuous lines represent TUNEL-positive. Broken lines represent TUNEL-negative.
tumour necrosis factor alpha promotes apoptosis around the time of
follicle formation, detected by TUNEL, in oocytes, interstitial and
granulosa cells (Morrison and Marcinkiewicz, 2002, rat).
Within nests, inter-oocyte communication is mediated via cytoplasmic bridges, but after nest breakdown, inter-follicular communication might continue, for example, via molecules produced by
granulosa cells (DaSilva-Buttkus et al., 2009). The developing ovary
is also influenced by the developing fetal endocrine system, and transcripts of some steroidogenic enzymes are present in the ovary from
at least 15 weeks of gestation (Voutilainen and Miller, 1986), whereas
the capacity to metabolize androgen to estrogen is present from
around 12 weeks of gestation (George and Wilson, 1978). Estrogen
and particularly progesterone inhibit follicular assembly in rats, possibly
via inhibiting apoptosis in oocytes (Kezele and Skinner, 2003).
However, the progesterone may be at least partially endocrine
rather than local, because an accelerated transition from primordial
to primary follicles, owing to a lack of progesterone, was noted
when ovaries had been removed from the in vivo environment
Oogenesis and cell death in prenatal ovaries
811
Figure 4 TUNEL and cleaved PARP-1 staining for apoptosis in mouse oocytes. TUNEL indicates DNA breaks, and cleaved PARP-1 is a marker of
apoptosis. (A) The proportions of TUNEL-positive and TUNEL-negative B6CBf2 mouse oocytes at different stages of MPI between 14.5 and 21 days
post-coitum. (B) The proportions of cleaved PARP-1-positive and cleaved PARP-1-negative B6CBf2 mouse oocytes at different stages of MPI between
14.5 and 21 days post-coitum. a versus b, c versus d, P , 0.05. Each oocyte was classified as being in a particular stage of MPI by core protein (COR-1)
staining and was also categorized as staining for either TUNEL or cleaved PARP-1, or both or neither (as described in Ghafari et al., 2007). In the
graphs presented, TUNEL-positive oocytes may be either positive or negative for cleaved PARP-1, and PARP-1-positive oocytes may be either positive
or negative for TUNEL. A total of 1960 oocytes from 24 fetal mice (3 per time point) were analysed.
812
Hartshorne et al.
Figure 4 Continued.
(Kezele and Skinner, 2003). Possible extra-gonadal sources of such
steroids have been identified in human fetal tissues (Pezzi et al.,
2003). Experiments where ovaries of a particular genotype are
transplanted into severe combined immune-deficient mice can be
applied to demonstrate whether local action of a gene is needed,
so, for example, local presence of the Fas gene within the ovary is
Oogenesis and cell death in prenatal ovaries
Figure 5 Expansion of human fetal ovarian tissue in vitro. Mean percentage (+SEM) for tissue from week 14 (A) and week 15 (B) of
gestation. The presence of growth factors (SCF 10 ng/ml and IGF-1
5 ng/ml) increased the tendency for outgrowth of 14 weeks of gestation tissue but did not significantly affect tissue of 15 weeks of gestation. Note the different scales on the y-axis. The data result from
five fetuses at 14 weeks and four at 15 weeks of gestation. The
time points analysed were: 5, 8, 11, 15, 19, and 22 days in culture.
necessary for the normal elimination of oocytes in transplanted ovaries
(Moniruzaman et al., 2007).
Against the background of this complex local environment, intrinsic
oocyte factors may also affect the prospects for survival. Some investigators have noted an increased incidence of abnormal synapsis in genetically abnormal fetuses (Barlow et al., 2002, Trisomy 21; Roig et al.,
2005a, b, Trisomy 18), as expected, and have also associated certain
chromosomal abnormalities with increased apoptosis (Modi et al.,
2003). We have reported disruption of SC and recombination in individual abnormal oocytes in normal human fetuses (Tease et al., 2002).
However, when examining individual mouse oocytes in MPI, we
found only a minor association between apoptotic molecular markers
and the normal or abnormal appearance of the SC in normal mice
(Ghafari et al., 2007). We have not observed any apparent association
between SC appearance and TUNEL staining in the limited numbers of
human oocytes that we have analysed (Ghafari F, Hartshorne GM,
unpublished results).
In contrast to the relative lack of association in our experience
between apoptotic markers and meiotic abnormality in individual
oocytes, we noted significant differences in the behaviour of ovarian
tissue samples in vitro, according to the gestational age. This could
influence the environment that is experienced by oocytes entering
MPI at different times during gestation.
813
We found that cultures of ovarian fragments from fetuses at 14
weeks of gestation were prone to expansion in vitro, with cells
moving away from the initial fragment and eventually covering much
of the membrane with clusters and aggregates (Hartshorne et al.,
1999). AP staining suggested that germ cells were mainly concentrated
in the initial tissue fragment and in clusters that formed in the outgrowth (data not shown). The curious ability of oocytes to migrate
during fetal life, even during MPI, has already been noted by previous
authors (Blandau et al., 1963, mice; Motta et al., 1997). Similar but
smaller aggregates were observed by Zhang et al., (1995), whose
methods were similar; however, their tissue donors were slightly
older, aged 16 –20 weeks. Since the pioneering studies on culture of
human fetal ovarian tissue (Baker and Neal, 1974), interest has
renewed recently and other methods have been reported. For
example, Wu et al. (2002) used minced tissue in dishes, whereas
Biron-Shental et al. (2004) used organ culture techniques, as applied
for adult tissue. Both of these groups observed follicle survival and
growth in vitro from tissue of 16–23 and 22–23 weeks of gestation,
respectively. Roig et al. (2006) and Sadeu et al. (2006), respectively, cultured ovarian samples of 17–22 weeks of gestation and 20–22 weeks
of gestation in tubes. Roig et al. (2006) confirmed meiotic initiation and
progression similar to that observed in our culture system (Hartshorne
et al., 1999). Interestingly, Sadeu et al. (2006) showed clear ultrastructural evidence of relatively slow follicle growth in vitro to primary follicles
in their tubal culture system, which compared with faster growth to a
more advanced follicular stage observed with adult ovaries cultured
similarly (Sadeu and Smitz, 2008). However, these authors did not
describe the cellular outgrowth that we observed, and which may be
a feature of the type of culture system employed. Optimization of
culture methods for fetal ovaries will provide a valuable tool with
which to study hormonal and paracrine influences upon key aspects
of human ovarian development.
The mechanism of movement of the germline cells in our system is
currently under study. This capacity for expansion was significantly
enhanced by the inclusion of growth factors in the culture medium
(SCF 10 ng/ml and IGF-1 5 ng/ml) (Fig. 5), and qualitative observations
suggested more AP-positive cells in cultures with SCF than those
without. In contrast, ovarian tissue of 15 weeks of gestation expanded
far less in vitro, and the stimulatory effect of growth factors was no
longer significant (Fig. 5). We have also noted that later gestational
ages (up to 23 weeks), cultured under similar conditions, tend to
round up in vitro and develop a dense surface epithelium, with little
or no attachment to the substrate (data not shown).
It is well known that mouse PGCs can be cultured successfully and
their numbers increased in cultures supplemented with growth factors
including SCF (kit ligand) (De Felici et al., 2005). It is also known that
c-kit is present on human oogonia and oocytes from 14 –21 weeks of
gestation (Robinson et al., 2001) and that the distribution of both c-kit
and SCF in the human germline varies according to the status of the
cell (Hoyer et al., 2005). For example, Hoyer et al. (2005), using
immunohistochemistry, found that PGCs and oogonia stain positive
for SCF and c-kit, whereas unenclosed oocytes or those enclosed in
primordial follicles stained only faintly or not at all for SCF and c-kit.
c-kit became detectable again at the surface of oocytes in growing follicles (Hoyer et al., 2005). However, Abir et al. (2004) found c-kit
expression to be evident in oocytes from the primordial follicle
stage only. We interpret the growth of 14 week gestation samples
814
Hartshorne et al.
Figure 6 Numbers of oocytes in MPI in human ovarian tissue after different times in culture. The top panel shows tissues from 14 weeks of gestation (n ¼ 108 cells) and the bottom panel shows tissues from 15 weeks of gestation (n ¼ 139 cells). The left side presents results from tissue that
was cultured in basal conditions (MEMa, 5% fetal calf serum, penicillin/streptomycin), whereas cultures presented on the right additionally included
SCF (10 ng/ml) and IGF-1 (5 ng/ml). Data from five fetuses at 14 weeks and four at 15 weeks of gestation. Stage of MPI was determined according to
SCP3 staining.
in culture as probably due to the proliferation of PGCs. However, the
dramatic difference between the behaviour of 14 weeks and 15 weeks
of gestation tissue in vitro is surprising. We know that precursor cells
remain in fetal ovaries, long after this time (Fig. 2), so perhaps they
proliferate less, or undergo apoptosis more at later stages.
We then assessed the progression of MPI in the same tissues cultured under different conditions using SC proteins as a specific
marker of MPI (as described in Hartshorne et al., 1999). The results
are shown in Fig. 6. Oocyte numbers markedly declined during the
first week in culture, probably owing to the shock of environmental
change, as noted by others (Byskov et al., 1997, mice; Roig et al.,
2006). Thereafter, oogenesis recovered with numbers of leptotene
and zygotene oocytes increasing during the following week of
culture. Recovery was more pronounced in 14 week than 15 week
samples, and was evident with or without growth factor supplementation, until the second week in vitro. Thereafter, cultures with
growth factors were more likely to maintain recovery, whereas
those without tended to slow or decline in oocyte numbers. It is
not yet clear whether the growth factors support the precursor
cells, the oocytes directly, or have an indirect effect via the somatic
environment. Further study will require the application of specific
markers to distinguish between oocytes and precursor cells at different stages, and the use of tissue samples from a wider range of gestational ages.
Factors affecting apoptosis
during MPI
In humans, MPI starts from around 13 weeks, and entry to meiosis
continues until at least 19 weeks of gestation, with different stages
of MPI, oogonia and degenerating oocytes present concurrently
(Fig. 1). Many researchers have proposed that oocyte cell death
during MPI is stage-specific, affecting the pachytene stage in particular,
and linked to the theme that oocytes that cannot complete the pachytene stage consequently arrest and undergo cell death (Modi et al.,
2003). Pachytene arrest is certainly functional in males (Guichaoua
et al., 2005); however, the situation in females seems more complex
and the controls of genetic integrity less stringent (Roig et al., 2004).
In addition, histological methods of identifying cells at different
stages of MPI, and distinguishing them from somatic and degenerating
cells, are fraught with difficulty, raising queries about the comparability
and interpretation of earlier studies which raise many conflicting data.
More recently, the application of molecular techniques has improved
the confidence with which oocytes can be identified. Our work in
mice confirmed that of McClellan et al. (2003), that apoptosis
occurs in all stages of oocytes (Ghafari et al., 2007); however,
similar analyses of human oocytes are not yet available.
We have also shown (Ghafari et al., 2009) that the lack of p53 gene
in knockout mice is associated with a changed profile of progression
through MPI at 15.5 and 16 dpc, possibly associated with reduced surveillance and an increased proportion of visually abnormal axial
elements. Therefore, a gene that is profoundly involved in the
control of the genetic integrity of cells and a regulator of apoptosis
(Pietsch et al., 2008) is implicated in both the rate of meiotic progression and the quality of the resulting oocytes, possibly adding
weight to the idea that failure of MPI owing to the timing of entry
or the rate of progress might predispose to elimination. p63 gene is
likely to be a dominant player here (Suh et al., 2006); however, its
role in response to exogenous stimuli, such as radiation, does not
exclude a role of p53 in the absence of such stimuli. p53 gene
815
Oogenesis and cell death in prenatal ovaries
product was detected in half of oogonia in human ovaries at 8 weeks
of gestation and all at 10–12 weeks of gestation (Quenby et al., 1999),
and p53 has been reported to increase with in vitro duration of culture
(Wu et al., 2002).
A further issue with human oocyte analyses comprises the genetic
diversity of donors. The impact of genetic background on aspects of
oogenesis has been noted by several authors (Koehler et al., 2002;
Canning et al., 2003; Ibáñez et al., 2005) and we have also documented differences between B6CB mice of F1 and F2 generations in terms
of meiotic progression (Ghafari et al., 2009). For humans, where the
supply of tissue is relatively limited, inter-individual variability can be
accommodated by controlled experiments utilizing tissue culture replicates. Therefore, much of our future work is aimed at characterizing
tissue responses in vitro and modelling prenatal oogenesis in a
manner that facilitates functional analyses.
Discussion
Even after many decades of study, oogenesis retains many intriguing
mysteries. The concepts presented here highlight some of the knowledge that we need in order to understand the processes involved. In
this article, we have presented new data on the timing of prenatal
oogenesis and apoptosis, through TUNEL staining and identification
of meiotic progress at different gestational ages in human, with comparative data in mice. We have also presented preliminary data on
the variable capacity for human oocytes and ovarian tissue to move
and expand in vitro. The large capacity for oocyte movement at
early stages is a striking finding that we shall follow up.
We emphasize the need to study human material, because the
results obtained using model animals differ from humans in our experience. An improved understanding of oocyte formation and selection
holds the key to many humans applications, with potential to
improve our ability to control unwanted fertility or promote fertility
treatments, suspend or accelerate the menopause, protect the germline from toxic agents and pharmacological or radiological destruction
and better understand the origins of maternally inherited genetic diseases, such as Trisomy 21 Down syndrome.
One reason why progress in this area has been slow is the difficulty
of accessing prenatal ovarian tissue for research. Ethical and procedural issues make completing the governance requirements a protracted task, even where the legal and regulatory framework is
permissive. Additional hurdles have included a limited array of sufficiently informative research techniques. However, recent progress in
molecular markers for MPI, and for oocytes and stem cells, has
helped immensely in the identification and classification of oocytes.
Similar progress in apoptotic markers and many other pathways relevant to ovarian cellular functions now leaves us poised for substantial
and rapid progress towards understanding oogenesis. The final piece
of the jigsaw is the establishment of culture methods that support
human oogenesis and folliculogenesis in vitro, thereby permitting
experimentation under controlled conditions, on live cells. The basic
model has been developed, and optimization is under way. Oocytes
are known to be sensitive to perturbations in culture, and local
ovarian signals in vivo are crucially important for survival and determination of cell fates. However, a live-cell model is the only way to study
the decisions of individual oocytes responding to selection criteria
applied for entry to the ovarian reserve.
In this way, factors affecting both the quality and the quantity of
oocytes will become accessible for study. The concept of a trade-off
between quantity and quality of oocytes, where resources are
limited, is attractive but has not been substantiated. However, it is
important to understand fully how individual local cell death decisions
accumulate to result in reproductive function at the level of the whole
organism. This systems biology approach cannot readily be tackled
until more basic molecular information is available, but it remains on
the horizon as a potentially fruitful research area.
Clearly, the internal environment of the pregnant woman during key
stages of fetal ovary development has the potential to directly affect
both the fertility of her future daughter (by controlling the size of
the ovarian reserve) and the quality of the oocyte that will eventually
become her grandchild (by influencing the extent of selection and
apoptosis). Prenatal oogenesis thus extends the well-known concept
of fetal origins of adult disease (Barker, 1998) to include the next generation, whose building blocks are dependent upon events up to 80
years earlier. This unique concept helps to underline the importance
of gaining a full understanding of prenatal oogenesis for the benefit
of future generations.
Materials and Methods
Tissues
Mice were housed in licensed premises at the University of Warwick and
killed at different gestational ages or within 2 days of birth. Ovaries were
dissected from fetuses or neonates under microscopic observation and
placed in culture medium. Full details are available in Ghafari et al. (2007).
Human fetal ovaries were obtained from second-trimester terminations
of pregnancy in patients who consented to their use in research. The pregnancies were normal (i.e. no abnormalities had been identified) and were
dated by foot length measurement, with additional information from ultrasound scan and/or dates of last menstrual period, if available. The
researchers had no contact with the patients. Tissues were maintained
at 48C until ovaries were removed aseptically, and placed into culture
medium for transport to the laboratory. For details, see Hartshorne
et al. (1999).
Detection of SC proteins
The SC is present uniquely during MPI and can be detected using specific
antibodies to its component parts on cells spread on microscope slides
(Hultén et al., 2001; Tease et al., 2002). The orientation of the SC proteins
often provides sufficient detail to be confident of the stage of MPI. Cell
spreads were prepared by chopping ovarian tissue to a fine suspension.
A drop of suspension on a microscope slide was mixed with a drop of
hypotonic solution (0.3% Lipsol) and allowed to stand for 30 min. Cells
were fixed by adding two drops of 1% formaldehyde in 0.4% sodium
dodecyl sulphate (SDS) for 20 min. Cells were blocked for 30 min in
PBT buffer before primary antibodies to either synaptonemal complex
protein 3 (SCP3) or core (COR-1) protein were added at dilutions of
1:1000. Anti-SCP3 antibody, raised in rabbit, was kindly provided by
Crista Heyting, Wageningen University, The Netherlands, and polyclonal
mouse anti-hamster COR-1 antibody was kindly provided by Barbara
Spyropoulos and Peter Moens, University of Toronto. Fluorescently
labelled secondary antibodies were applied at a dilution of 1:200, and
the mounted slides were viewed under fluorescence microscopy. Full
details are available in Hartshorne et al. (1999) and Ghafari et al. (2007).
816
Detection of apoptotic nuclear markers
Cleaved PARP-1
The primary antibody against cleaved PARP-1 (rabbit anti-mouse, polyclonal antibody) was applied at a concentration of 1:50 in PBT, simultaneously
with the anti-COR-1 antibody described earlier. Slides were placed in a
moist chamber at 48C overnight. Fluorescently labelled secondary antibodies were then applied and visualized as described.
Hartshorne et al.
Funding
The authors are grateful to the following organizations which have
contributed funding to this research programme: the Royal Society,
The Wellcome Trust, Wellbeing of Women, The Sanofi Winthrop
Foundation, The University of Warwick, Centre for Reproductive
Medicine, University Hospitals Coventry and Warwickshire NHS Trust.
References
TUNEL
DNA cleavage was detected by TUNEL using the Apop Tag Fluorescein
Direct In Situ Apoptosis Detection Kit (Intergen, USA). We applied it to
study DNA breaks on the same micro-spread oocytes that had already
been assessed using antibodies to COR1 and cleaved PARP-1. All
washing and incubation processes were therefore performed in the
dark. Slides were washed in PBT, and then processed according to the
manufacturer’s protocol for TUNEL. Positive control slides were treated
with 0.9 mg/ml DNAse 1 in DN buffer for 10 min at room temperature.
Slides were mounted with Vectashield mounting medium without DAPI.
Enzymatic detection of AP
AP is a cytoplasmic marker routinely used for detecting germline cells. We
applied the method of Buehr and McLaren (1993) to suspensions of
ovarian tissue prepared by fine chopping, followed by spreading on a
microscope slide and air drying. The same staining method was also
applied to membranes containing cultured cells.
Slides or membranes were fixed for 20 – 30 s in buffered acetone,
washed in tap water and allowed to dry. Slides were stained with AP substrate for 15 min, rinsed in dH2O and mounted in aqueous mountant. The
stain comprised 25 mg Fast Texas Red and 5 mg alpha-naphthyl phospiate
in 44.6 ml dH2O; 0.3 ml 10% MgCl2 and 5 ml 4.5% borax. These ingredients were mixed in order and used immediately. The presence of AP
resulted in red staining.
Culture of fetal ovarian tissue
Fragments of ovarian tissue, 0.3 0.3 0.3 mm3, were placed individually on permeable polyester membranes (Costar) in pre-prepared tissue
culture wells containing 1 ml culture medium in 12-well plates. The
tissue was supported near to the surface of the medium, and any empty
wells were filled with medium to maintain humidity. The culture
medium was minimal essential medium (MEM) a, supplemented with
insulin, transferrin and selenium, and 2.5 – 5% serum, with or without
SCF (10 ng/ml) and IGF-1 (5 ng/ml). Growth factors were prepared in
PBS containing 0.1% bovine serum albumin. Control cultures received
the appropriate vehicle only. The medium was refreshed by changing
that under the membrane every 2 – 4 days. Cultures were maintained at
378C under 5% CO2 in air for 22 days.
Acknowledgements
The authors wish to acknowledge the assistance and support of all
those who have contributed to the collection of tissues for research,
particularly staff of the British Pregnancy Advisory Service and the John
Radcliffe Hospital. We also wish to thank in particular Dr Charles
Tease and Professors Maj Hulten, Anne Grete Byskov and Ursula
Eichenlaub-Ritter, whose valuable discussions have helped to formulate some of the ideas presented. Dr Kate Watkins and Laura Evans
are gratefully acknowledged for their contribution to Fig. 1.
Abir R, Orvieto R, Dicker D, Zukerman Z, Barnett M, Fisch B. Preliminary
studies on apoptosis in human fetal ovaries. Fertil Steril 2002;78:259–264.
Abir R, Fisch B, Jin S, Barnnet M, Kessler-Icekson G, Ao A. Expression of
stem cell factor and its receptor in human fetal and adult ovaries. Fertil
Steril 2004;82(Suppl. 3):1235 – 1243.
Albamonte MS, Willis MA, Albamonte MI, Jensen F, Espinosa MB,
Vitullo AD. The developing human ovary: immunohistochemical
analysis of germ-cell-specific VASA protein, BCL-2/BAX expression
balance and apoptosis. Hum Reprod 2008;23:1895– 1901.
Alton M, Taketo T. Switch from BAX-dependent to BAX-independent
germ cell loss during the development of fetal mouse ovaries. J Cell
Sci 2007;120:417 – 424.
Baker TG. A quantitative and cytological study of germ cells in human
ovaries. Proc R Soc Lond B Biol Sci 1963;158:417 –433.
Baker TG, Neal P. Oogenesis in human fetal ovaries maintained in organ
culture. J Anat 1974;117:591– 604.
Barker DJ. Mothers, Babies and Health in Later Life. Glasgow: Churchill
Livingstone, 1998.
Barker DJP, Clark PM. Fetal undernutrition and disease in later life. Rev
Reprod 1997;2:105 – 112.
Barlow AL, Tease C, Hultén MA. Meiotic chromosome pairing in fetal
oocytes of trisomy 21 human females. Cytogenet Genome Res 2002;
96:45 – 51.
Baum JS, St George JP, McCall GK. Programmed cell death in the germline.
Semin Cell Dev Biol 2005;16:245– 259.
Bendsen E, Byskov AG, Andersen CY, Westergaard LG. Number of germ
cells and somatic cells in human fetal ovaries during the first weeks after
sex differentiation. Hum Reprod 2006;21:30 –35.
Bennett RA, Osathanondh R, Yeh J. Immunohistochemical localization of
transforming growth factor-alpha, epidermal growth factor (EGF), and
EGF receptor in the human fetal ovary. J Clin Endocrinol Metab 1996;
81:3073 – 3076.
Bergeron L, Perez GI, Macdonald G, Shi L, Sun Y, Jurisicova A, Varmuza S,
Latham KE, Flaws JA et al. Defects in regulation of apoptosis in
caspase-2-deficient mice. Genes Dev 1998;12:1304 – 1314.
Biron-Shental T, Fisch B, Van Den Hurk R, Felz C, Feldberg D, Abir R.
Survival of frozen-thawed human ovarian fetal follicles in long-term
organ culture. Fertil Steril 2004;81:716 –719.
Blandau RJ, White BJ, Rumery RE. Observations on teh movements of the
living primordial germ cells in the mouse. Fertil Steril 1963;14:482– 489.
Bowles J, Koopman P. Retinoic acid, meiosis and germ cell fate in
mammals. Development 2007;134:3401– 3411.
Buehr M, McLaren A. Isolation and culture of primordial germ cells. Meth
Enzymol 1993;225:58 – 77.
Byskov AG, Lintern-Moore S. Follicle formation in the immature mouse
ovary: the role of the rete ovarii. J Anat 1973;116:207 – 217.
Byskov AG, Guoliang X, Andersen CY. The cortex-medulla oocyte growth
pattern is organized during fetal life: an in-vitro study of the mouse
ovary. Mol Hum Reprod 1997;3:795 – 800.
Canning J, Takai Y, Tilly JL. Evidence for genetic modifiers of ovarian
follicular endowment and development from studies of five inbred
mouse strains. Endocrinology 2003;144:9– 12.
Oogenesis and cell death in prenatal ovaries
Carlson BM. Human Embryology and Developmental Biology, 3rd edn. Mosby
Inc., St Louis, 2004.
Coccia ME, Rizzello F. Ovarian reserve. Ann N Y Acad Sci 2008;1127:
27 – 30.
Coutts SM, Childs AJ, Fulton N, Collins C, Bayne RA, McNeilly AS,
Anderson RA. Activin signals via SMAD2/3 between germ and
somatic cells in the human fetal ovary and regulates kit ligand
expression. Dev Biol 2008;314:189 – 199.
Cummings AM, Kavlock RJ. Function of secual glands and mechanism of
sex differentiation. J Toxicological Sci 2004;29:167 –178.
DaSilva-Buttkus P, Marcelli G, Franks S, Stark J, Hardy K. Inferring
biological mechanisms for spatial analysis: prediction of a local
inhibitor in the ovary. Proc Natl Acad Sci USA 2009;106:456 – 461.
De Felici M, Klinger GF, Farini D, Scaldaferri ML, Iona S, Lobascio M.
Establishment of oocyte population in the fetal ovary: primordial germ
cells proliferation and oocyte programmed cell death. RBM Online
2005;10:182 – 191.
DeFelici M, Lobascio AM, Klinger FG. Cell death in fetal oocytes: many
players for multiple pathways. Autophagy 2008;4:240 – 242.
Djahanbakhch O, Ezzati M, Zosmer A. Reproductive ageing in women.
J Pathol 2007;211:219 – 231.
Faddy MJ, Gosden RG, Oktay K, Nelson JF. Factoring in complexity and
oocyte memory - can transformations and cyperpathology distort
reality? Fertil Steril 1999;71:1170– 1172.
Farini D, Scaldaferri ML, Iona S, LaSala S, DeFelici M. Growth factors
sustain primordial germ cell survival, proliferation and entering
into meiosis in the absence of somatic cells. Dev Biol 2005;285:
49 – 56.
Forabosco A, Sforza C. Establishment of ovarian reserve: a quantitative
morphometric study of the developing human ovary. Fertil Steril 2007;
88:675– 683.
Fowler PA, Flannigan S, Mathers A, Gillanders K, Lea RG, Wood J,
Maheshwari A, Bhattacharya S, Collie-Duguid ESR, Baker PJ et al.
Gene expression analysis of human fetal ovarian primordial follicle
formation. J Clin Endocrinol Metab 2009;94:1427– 1435.
Fulton N, Martins da Silva SJ, Bayne RAL, Anderson RA. Germ cell
proliferation and apoptosis in the developing human ovary. J Clin
Endocrinol Metab 2005;90:4664 – 4670.
George FW, Wilson JD. Conversion of androgen to estrogen by the
human fetal ovary. J Clin Endocrinol Metab 1978;47:550– 555.
Ghafari F, Gutierrez CG, Hartshorne GM. Apoptosis in mouse fetal and
neonatal oocytes during meiotic prophase one. BMC Dev Biol 2007;
7:87.
Ghafari F, Pelengaris S, Walters E, Hartshorne GM. Influence of p53 and
genetic background on prenatal oogenesis and oocyte attrition in
mice. Hum Reprod 2009;24:1460 – 1472.
Gilchrist RB, Lane M, Thompson JG. Oocyte-secreted factors: regulators
of cumulus cell function and oocyte quality. Hum Reprod Update 2008;
14:159– 177.
Gondos B. Comparative studies of normal and neoplastic ovarian germ
cells: 1. Ultrastructure of oogonia and intercellular bridges in the fetal
ovary. Int J Gynecol Pathol 1987;6:114 – 123.
Gu Y, Runyan C, Shoemaker A, Surani A, Wylie C. Steel factor controls
primordial germ cell survival and motility from the time of their
specification in the allantois, and provides a continuous niche
throughout their migration. Development 2009;136:1295– 1303.
Guichaoua MR, Perrin J, Metzler-Guillemain C, Saias-Magnan J, Giorgi R,
Grillo JM. Meiotic anomalies in infertile men with severe
spermatogenic defects. Hum Reprod 2005;20:1897– 1902.
Harman D. Aging: a theory based on free radical and radiation chemistry.
J Gerontol 1956;11:298 – 300.
817
Hartshorne GM, Barlow AL, Child TJ, Barlow DH, Hultén MA.
Immunocytogenetic detection of normal and abnormal oocytes in
human fetal ovarian tissue in culture. Hum Reprod 1999;14:172–182.
Henderson SA, Edwards RG. Chiasma frequency and maternal age in
mammals. Nature 1968;218:22 – 28.
Hoyer PE, Byskov AG, Mollgard K. Stem cell factor and c-Kit in human
primordial germ cells and fetal ovaries. Mol Cell Endocrinol 2005;
234:1 –10.
Hultén MA, Barlow AL, Tease C. Meiotic studies in humans. In:
Rooney DE (ed). Human Cytogenetics: Constitutional Analysis, A Practical
Approach. 3rd ed. Oxford: Oxford University Press, 2001, 211 – 236.
Hultén MA, Patel SD, Tankimanova M, Westgren M, Papadogiannakis N,
Jonsson AM, Iwarsson E. On the origin of trisomy 21 Down syndrome.
Mol Cytogenet 2008;1:21.
Hunt PA, Hassold TJ. Human female meiosis: what makes a good egg go
bad? Trends Genet 2008;24:86 – 93.
Hussein MR. Apoptosis in the ovary: molecular mechanisms. Hum Reprod
Update 2005;11:162 – 178.
Ibáñez E, Albertini DF, Overström EW. Effect of genetic background and
activating stimulus on the timing of meiotic cell cycle progression in
parthenogenetically activated mouse oocytes. Reproduction 2005;
129:27– 38.
Jones KT. Meiosis in oocytes: predisposition to aneuploidy and its
increased incidence with age. Hum Reprod Update 2008;14:143– 158.
Keefe DL, Marquard K, Liu L. The telomere theory of reproductive
senescence in women. Curr Opin Obstet Gynecol 2006;18:
280 – 285.
Kezele P, Skinner MK. Regulation of ovarian primordial follicle assembly
and development by estrogen and progesterone: endocrine model of
follicle assembly. Endocrinology 2003;144:3329 – 3337.
Koehler KE, Cherry JP, Lynn A, Hunt PA, Hassold TJ. Genetic control of
mammalian meiotic recombination. I. Variation in exchange frequencies
among males from inbred mouse strains. Genetics 2002;162:
297 – 306.
Krysko DV, Diez-Fraile A, Criel G, Svistunov AA, Vandenabeele P,
D’Herde K. Life and death of female gametes during oogenesis and
folliculogenesis. Apoptosis 2008;13:1065 – 1087.
Kurilo FL. Oogenesis in antenatal development in man. Hum Genet 1981;
57:86– 92.
Lintern-Moore S, Peters H, Moore GPM, Faber M. Follicular development
in the infant human ovary. J Reprod Fertil 1974;39:53 – 64.
Lobascio AM, Klinger FG, Scaldaferri ML, Farini D, DeFelici M. Analysis of
programmed cell death in mouse fetal oocytes. Reproduction 2007;
134:241– 252.
Lumey LH, Stein AD. In utero exposure to famine and subsequent fertility:
The Dutch Famine Birth Cohort Study. Am J Public Health 1997;
87:1962– 1966.
Mahadevaiah SK, Turner JMA, Baudat F, Rogakou EP, DeBoer P,
Blanco-Rodriguez J, Jasin M, Keeney S, Bonner WM, Burgoyne PS.
Recombinational DNA double-strand breaks in mice precede
synapsis. Nat Genet 2001;27:271– 276.
Maheshwari A, Fowler PA. Primordial follicular assembly in humans–
revisited. Zygote 2008;16:285 – 296.
Martins da Silva SJ, Bayne RAL, Cambray N, Hartley PS, McNeilly AS,
Anderson RA. Expression of activin subunits and receptors in the
developing human ovary: activin A promotes germ cell survival and
proliferation before primordial follicle formation. Dev Biol 2004;
266:334– 345.
Matzuk MM, Burns KH, Viveiros MM, Eppig JJ. Intercellular communication
in the mammalian ovary: oocytes carry the conversation. Science 2002;
296:2178– 2180.
818
McClellan KA, Gosden R, Taketo T. Continuous loss of oocytes
throughout meiotic prophase in the normal mouse ovary. Dev Biol
2003;258:334 – 348.
Modi DN, Sane S, Bhartiya D. Accelerated germ cell apoptosis in sex
chromosome aneuploid fetal human gonads. Mol Hum Reprod 2003;
9:219 – 225.
Moniruzaman M, Sakamaki K, Akazawa Y, Miyano T. Oocyte growth and
follicular development in KIT-deficient Fas-knockout mice. Reproduction
2007;133:117 – 125.
Morita Y, Manganaro TF, Tao XJ, Martimbeau S, Donahoe PK, Tilly JL.
Requirement of phosphatidylinositol-30 -kinase in cytokine-mediated
germ cell survival during fetal oogenesis in the mouse. Endocrinology
1999;140:941 – 949.
Morrison LJ, Marcinkiewicz JL. Tumor necrosis factor alpha enhances
oocyte/follicle apoptosis in the neonatal rat ovary. Biol Reprod 2002;
66:450 – 457.
Motta PM, Makabe S. Elimination of germ cells during differentiation of the
human ovary: an electron microscopic study. Eur J Obstet Gynecol Reprod
Biol 1986;22:271 – 286.
Motta PM, Makabe S, Nottola SA. The ultrastructure of human
reproduction. I. The natural history of the female germ cell: origin,
migration and differentiation inside the developing ovary. Hum Reprod
Update 1997;3:281– 295.
Murdoch WJ, VanKirk EA, Vonnahme KA, Ford SP. Ovarian responses to
undernutrition in pregnant ewes, USA. Reprod Biol Endocrinol 2003;1:6.
Ohno S, Klinger HP, Atkin NB. Human oogenesis. Cytogenetics 1962;
1:42 – 51.
Oktem O, Oktay K. Stem cells, a perspective on oocytes. Ann NY Acad Sci
2008;1127:20 – 26.
Pepling ME. From primordial germ cell to primordial follicle: mammalian
female germ cell development. Genesis 2006;44:622 –632.
Pepling ME, Spradling AC. Mouse ovarian germ cell cysts undergo
programmed breakdown to form primordial follicles. Dev Biol 2001;
234:339– 351.
Perez GI, Robles R, Knudson CM, Flaws JA, Korsmeyer SJ, Tilly JL.
Prolongation of ovarian lifespan into advanced chronological age by
Bax-deficiency. Nat Genet 1999;21:200 – 203.
Peters H, Himelstein-Braw R, Faber M. The normal development
of the ovary in childhood. Acta Endocrinol (Copenh) 1976;82:617 – 630.
Peters H, Byskov AG, Grinsted J. Follicular growth in fetal and prepubertal
ovaries of humans and other primates. Clin Endocrinol Metab 1978;
7:469 – 485.
Pezzi V, Mathis JM, Rainey WE, Carr BR. Profiling transcript levels for
steroidogenic enzymes in fetal tissues. J Steroid Biochem Mol Biol 2003;
87:181 – 189.
Pietsch EC, Sykes SM, McMahon SB, Murphy ME. The p53 family and
programmed cell death. Oncogene 2008;27:6507 – 6521.
Polani PE, Crolla JA. A test of the production line hypothesis of
mammalian oogenesis. Hum Genet 1991;88:64– 70.
Qin J, Richardson LL, Jasin M, Handel MA, Arnheim N. Mouse strains with
an active H2-Ea meiotic recombination hot spot exhibit increased levels
of H2-Ea-specific DNA breaks in testicular germ cells. Mol Cell Biol 2004;
24:1655 – 1666.
Quenby SM, Gazvani MR, Brazeau C, Neilson J, Lewis-Jones DI, Vince G.
Oncogenes and tumour suppressor genes in first trimester human fetal
gonadal development. Mol Hum Reprod 1999;5:737– 741.
Rabinovici J, Jaffe RB. Development and regulation of growth and
differentiated function in human and subhuman primate fetal gonads.
Endocr Rev 1990;11:532 – 557.
Ratts VS, Flaws JA, Klop R, Sorenson CM, Tilly JL. Ablation of bcl2 gene
expression decreases the number of oocytes and primordial follicles
Hartshorne et al.
established in the postnatal female mouse gonad. Endocrinology 1995;
136:3665– 3668.
Reddy P, Liu L, Adhikari D, Jagarlamudi K, Rajareddy S, Shen Y, Du C,
Tang W, Hamalainen T, Peng SL et al. Oocyte-specific deletion of
Pten causes premature activation of the primordial follicle pool.
Science 2008;319:611 – 613.
Robinson LLL, Gaskell TL, Saunders PTK, Anderson RA. Germ cell specific
expression of c-kit in the human fetal gonad. Mol Hum Reprod 2001;
7:845 – 852.
Rodrigues P, Limback D, McGinnis LK, Plancha CE, Albertini DF. Multiple
mechanisms of germ cell loss in the perinatal mouse ovary. Reproduction
2009;137:709 – 720.
Roig I, Liebe B, Egozcue J, Cabero L, Garcia M, Scherthan H.
Female-specific features of recombinational double-stranded DNA
repair in relation to synapsis and telomere dynamics in human
oocytes. Chromosoma 2004;113:22– 33.
Roig I, Robles P, Garcia R, Martinez-Flores I, Cabero L, Egozcue J, Liebe B,
Scherthan H, Garcia M. Chromosome 18 pairing behaviour in human
trisomic oocytes. Presence of an extra chromosome extends bouquet
stage. Reproduction 2005a;129:565 – 575.
Roig I, Robles P, Garcia R, Martin M, Egozcue J, Cabero L, Barambio S,
Garcia M. Evolution of the meiotic prophase and of the chromosome
pairing process during human fetal ovarian development. Hum Reprod
2005b;20:2463– 2469.
Roig I, Garcia R, Robles P, Cortvrindt R, Egozcue J, Smitz J, Garcia M.
Human fetal ovarian culture permits meiotic progression and
chromosome pairing process. Hum Reprod 2006;21:1359 – 1367.
Sadeu JC, Smitz J. Growth differentiation factor-9 and anti-Mullerian
hormone expression in cultured human follicles from frozen thawed
ovarian tissue. Reprod BioMed Online 2008;17:537– 548.
Sadeu JC, Cortvrindt R, Ron-El R, Kasterstein E, Smitz J. Morphological and
ultrastructural evaluation of cultured frozen-thawed human fetal ovarian
tissue. Fertil Steril 2006;85(Suppl. 1):1130 – 1141.
Sakata S, Sakamaki K, Watanabe K, Nakamura N, Toyokuni S,
Nishimune Y, Mori C, Yonehara S. Involvement of death receptor Fas
in germ cell degeneration in gonads of Kit-deficient Wv/Wv mutant
mice. Cell Death Differ 2003;10:676 –686.
Schilling B, Yeh J. Expression of transforming growth factor (TGF)-beta1,
TGF-beta2, and TGF-beta3 and of type I and II TGF-beta receptors during
the development of the human fetal ovary. Fertil Steril 1999;72:147–153.
Skinner MK. Regulation of primordial follicle assembly and development.
Hum Reprod Update 2005;11:461– 471.
Speed RM. The prophase stages in human foetal oocytes studied by light
and electron microscopy. Hum Genet 1985;69:69– 75.
Suh EK, Yang A, Kettenbach A, Bamberger C, Michaelis AH, Zhu Z,
Elvin JA, Bronson RT, Crum CP, McKeon F. p63 protects the female
germ line during meiotic arrest. Nature 2006;444:624 – 628.
Tease C, Hartshorne GM, Hultén MA. Patterns of meiotic recombination
in human fetal oocytes. Am J Hum Genet 2002;70:1469 – 1479.
Tingen CM, Bristol-Gould SK, Kiesewetter SE, Wellington JT, Shea L,
Woodruff TK. Prepubertal primordial follicle loss in mice is not due
to classical apoptotic pathways. Biol Reprod Mar 2009;4: epub ahead
of print.
Vaskivuo TE, Anttonen M, Herva R, Billig H, Dorland DM, TeVelde ER,
Stenback F, Heikinheima M, Tapanainen JS. Survival of human ovarian
follicles from fetal to adult life: apoptosis, apoptosis-related proteins,
and transcription factor GATA-4. J Clin Endocr Metab 2001;
86:3421 – 3429.
Voutilainen R, Miller WL. Developmental expression of genes for the
stereoidogenic enzymes P450scc (20,22-desmolase), P450c17 (17
alpha-hydroxylase/17,20-lyase), and P450c21 (21-hydroxylase) in the
human fetus. J Clin Endocrinol Metab 1986;63:1145 – 1150.
Oogenesis and cell death in prenatal ovaries
Wandji SA, Srsen V, Nathanielsz PW, Eppig JJ, Fortune JE. Initiation of
growth of baboon primordial follicles in vitro. Hum Reprod 1997;
12:1993– 2001.
Wartenberg H, Ihmer A, Schwarz S, Miething A, Viebahn C. Mitotic arrest
of female germ cells during prenatal oogenesis. A colcemid-like,
non-apoptotic cell death. Anat Embryol 2001;204:421– 435.
Wu J, Chen Y, Tan L. Expression of Fas, p53 and AFP in development of
human fetal germ cells in vitro. Zygote 2002;10:333– 340.
Yang MY, Fortune JE. The capacity of primordial follicles in fetal bovine
ovaries to initiate growth in vitro develops during mid-gestation and is
819
associated with meiotic arrest of oocytes. Biol Reprod 2008;
78:1153– 1161.
Zachos NC, Billiar RB, Albrecht ED, Pepe GJ. Developmental regulation of
baboon fetal ovarian maturation by estrogen. Biol Reprod 2002;
67:1148– 1156.
Zhang J, Liu K, Xi LP, Liu B, DeMattina M. Extracorporeal development
and ultrarapid freezing of human fetal ova. J Assist Reprod Genet 1995;
12:361– 368.
Submitted on February 19, 2009; resubmitted on June 23, 2009; accepted on July 3,
2009