BIOLOGY OF REPRODUCTION 56, 837-846 (1997)
Morphological Evidence of Apoptosis and the Prevalence of Apoptotic versus
Mitotic Cells in the Membrana Granulosa of Ovarian Follicles during
Spontaneous and Induced Atresia in Ewes
P.D. Jolly, P.R. Smith, D.A. Heath, N.L. Hudson, S. Lun, L.A. Still, C.H. Watts, and K.P. McNatty'
AgResearch, Wallaceville Animal Research Centre, Upper Hutt, New Zealand
ABSTRACT
Apoptosis is a process by which granulosa cells are thought
to be deleted during ovarian follicular atresia. The aims of the
present studies, using sheep as the experimental model, were to
determine 1) whether morphological changes in cells composing
the membrana granulosa during the process of atresia conformed with the general criteria of apoptotic cell death as assessed using tissue sections stained with hematoxylin and eosin;
2) whether cells classified as apoptotic on the basis of their morphology contained fragmented DNA using an in situ 3' end-labeling technique; and 3) the degree of apoptosis and mitosis
within the granulosa cell populations of large antral follicles (3 mm in diameter) during both spontaneous and experimentally
induced atresia using stereological methods.
The results showed that most degenerate granulosa cells in
follicles undergoing atresia display the morphological characteristics of apoptosis, suggesting that this is the most common
pathway of cell deletion. Typical features were cells containing
nuclei with marginated chromatin; cells with a single small
densely staining nucleus (pyknotic appearance); cells with multiple smaller, densely staining nuclear fragments; and densely
staining membrane-bound bodies (apoptotic bodies) either singly or in clusters. Cells with morphological features more typical
of oncosis or necrosis were sometimes observed, but mainly during the later stages of atresia. All cells classified as apoptotic on
the basis of morphological criteria contained fragmented DNA
as measured by 3' end-labeling.
Apoptotic bodies and/or cells were found in all follicles examined, including those classified as healthy. The overall prevalence of apoptotic cells plus apoptotic bodies expressed as a
percentage of the total granulosa cell number per follicle varied
from 0.02% to 0.20% in healthy follicles, varied from 0.21%
to 2.00% in follicles in early (primary) atresia, and was > 2.0%
in follicles in later (secondary) atresia. Percentages of mitotic
cells in healthy follicles were > 0.5% in all but one of those
examined and were < 1.0% in all follicles classified as atretic.
Both morphological and 3' end-labeling results indicated that
apoptotic cells were widely disseminated throughout the membrana granulosa, including the cell layer adjacent to the basement membrane. Collectively, these observations indicate that
during early atresia, apoptosis occurs randomly and is not limited to specific areas within follicles. Our finding that apoptotic
cell death and mitosis occur simultaneously within the same follicle is consistent with the notion that atresia is determined by
a dynamic equilibrium between cell division, differentiation,
and death.
INTRODUCTION
Recent studies have established that granulosa cell death
during the process of ovarian follicular atresia in ewes can
Accepted October 3, 1996.
Received July 9, 1996.
'Correspondence: AgResearch, Wallaceville Animal Research Centre,
P.O. Box 40063, Upper Hutt, New Zealand. FAX: 64 (4) 528-6605;
e-mail: [email protected]
837
occur by apoptosis, a genetically regulated process of selective cell deletion [1]. This was accomplished by demonstrating oligonucleosomes (a biochemical marker of
apoptosis) in DNA isolated from granulosa cells. This finding is consistent with observations in cows [2] and other
mammalian and avian species (reviewed in [3]), as well as
with the mechanism of cell death demonstrated for other
ovarian cell types in ewes: namely, theca interna cells [4]
and luteal cells during regression of the corpus luteum [58], as well as ovarian surface epithelial cells and cells of
the apical follicle wall during the process of ovulation [9,
10]. From these data it appears that a common mechanism
underlies the fate of ovarian cells in ewes, as in other species.
. In our recent study, oligonucleosomes were evident in
ovine granulosa cell DNA purified from gonadotropin-dependent follicles ( 3 mm in diameter) either undergoing
spontaneous atresia or undergoing atresia induced by injecting ewes with charcoal-treated bovine follicular fluid
(bFF) [1]. Treatment of ewes with bFF at 12-h intervals
induced oligonucleosome formation, loss of extant aromatase activity, and loss of cAMP response to LH within 24
h in granulosa cells from all follicles examined. Results
also indicated that granulosa cells began to die by apoptosis
before there was an appreciable decrease in the capacity of
the granulosa cells to respond to gonadotropins or to synthesize estradiol-17B [1]. Thus it appears that an increase
in the incidence of granulosa cell apoptosis is a very early
event during atresia.
In addition to the detection of oligonucleosomes in isolated DNA, the occurrence of apoptosis may be inferred
from characteristic morphological appearances of degenerating cells [11, 12], together with the detection of fragmented DNA in single cells in situ using 3' end-labeling
methods [13, 14]. The morphological features of cells undergoing apoptosis have been consistently described for numerous cell types under a wide range of physiological and
experimental conditions [11, 12, 15-17]; these include cell
shrinkage and the condensation of chromatin that is either
marginated around the nuclear membrane (marginated
chromatin) or fragmented into multiple dense basophilic
masses. Cells may then break up into multiple discrete
membrane-bounded structures containing variable amounts
of condensed chromatin and/or cytoplasm (apoptotic bodies) that are phagocytosed by neighboring cells or macrophages, or are extruded into body cavities such as the follicular antrum. Alternatively, cells may shrink into a single
dense rounded mass with a densely basophilic (pyknotic)
nucleus [12, 15, 17].
While histological descriptors such as nuclear pyknosis,
karyorrhexis, and the formation of "densely staining" bodies are frequently applied to cellular changes that occur
during follicular atresia, these are not necessarily specific
for apoptotic cell death [18]. The morphological features of
838
JOLLY ET AL.
granulosa cell death during follicular atresia have not been
specifically described in the context of changes known to
occur during the process of apoptosis. In this study we
asked whether changes in the morphological appearance of
cells within the membrana granulosa during the process of
follicular atresia in ewes conformed with the characteristic
morphological features of apoptotic cell death. In addition,
we sought biochemical evidence to determine whether cells
or cell debris that were classified as apoptotic on the basis
of their morphological appearance contained fragmented
DNA, using an in situ 3' end-labeling technique (terminal
deoxynucleotidyl transferase-mediated dUPT nick end-labeling; TUNEL) [13]. Finally, we sought to quantitate the
degree of apoptosis and mitosis occurring within the granulosa cell layer of large antral follicles ( 3 mm in diameter) in ewes during both spontaneous and induced atresia
in a histological study using stereological methods [19, 20].
We did this using the same model system, the bFF-treated
ewe, in which we have previously demonstrated the presence of oligonucleosomes in granulosa cell DNA in relation
to other biochemical and morphometric measures of granulosa cell function and follicular atresia [1].
MATERIALS AND METHODS
Experimental Design, Animals, and Collection of Ovaries
Experiments reported here were conducted in accordance
with the 1987 Animal Protection (Codes of Ethical Conduct) Regulations of New Zealand with the approval of the
Animal Ethics Committee of the Wallaceville Animal Research Centre. The animals used in this study were 4- to
7-yr-old parous New Zealand Romney ewes, grazing pasture during the breeding season and running with vasectomized rams fitted with marking harnesses for the detection
of estrus.
Experiment 1 was designed to identify apoptotic cells
and to determine the relationship between the morphological appearance of cells and/or structures after hematoxylin
and eosin staining and their propensity to label using TUNEL. The objective was to determine whether granulosa
cells that were classified as apoptotic on the basis of their
histological appearance contained fragmented DNA. Ovaries were collected immediately after ewes (n = 3) in the
midluteal phase of the estrous cycle were killed. All folli3 mm in diameter were individually dissected free
cles
from extraneous tissue using a stereomicroscope. Follicle
diameters were measured to the nearest 0.1 mm. Thereafter
follicles were placed in Bouin's fixative for 12 h, processed
for histology, and embedded in paraffin wax. Processing
for TUNEL was also performed on sections containing
large follicles (- 3 mm in diameter) from midluteal-phase
ewe ovaries (n = 3) that had been perfused in situ via the
aorta with 4% (w:v) phosphate-buffered paraformaldehyde,
processed for histology, and embedded in paraffin as previously described [21]. Follicles and ovaries were sectioned
at 5 Lm and mounted on slides coated for TUNEL with
3-aminopropyltriethoxy-silane (Sigma Chemical Company,
St. Louis, MO) or stained with hematoxylin and eosin.
In Situ 3' End-Labeling (TUNEL)
Tissue sections were processed for TUNEL using a modification of the method described by Gavrieli et al. [13].
Sections were deparaffinized, rehydrated, and incubated in
double-strength SSC buffer (30 mM citrate, 300 mM NaCI,
pH 7.0) if fixed in Bouin's, or in 0.01 M citrate (pH 6.0)
if fixed in paraformaldehyde, at 80°C for 30 min. Sections
were then washed in sterile water and incubated with proteinase K (0.5-10 mg/ml in 20 mM Tris, 2 mM CaCl 2 , pH
7.2) in a humidified chamber at 37C for 30 min, washed
again in sterile water, immersed in 2% aqueous hydrogen
peroxide for 5 min, washed, and then pretreated with terminal deoxynucleotidyl transferase (TdT) buffer (Gibco
BRL, Grand Island, NY) for 5 min at room temperature
(RT). DNA was then labeled at 3'-ends by incubating sections with a mixture of biotinylated dATP (5 mM biotin14-dATP; Gibco), unlabeled dATP (25 mM; Gibco), and
TdT enzyme (0.5 u/ml; Gibco) in TdT buffer in a humidified chamber at 37C for 90 min. The labeling reaction
was stopped by washing slides in TBS buffer (50 mM Tris,
150 mM NaCl, pH 7.6) three times, 5 min each, at RT.
Slides were then incubated with 2% BSA (Sigma) in sterile
water at RT for 5 min, washed in TBS buffer, and incubated
with avidin-biotinylated horseradish peroxidase complex
(Dako Corporation, Carpinteria, CA) at RT for 30 min.
Slides were washed again in TBS (three times, 5 min each),
then incubated with 3,3' diaminobenzidine (DAB; 0.5
mg/ml in TBS buffer containing 0.002% hydrogen peroxide) at RT for 5-10 min, rinsed under running water,
mounted using 90% glycerol in TBS, and sealed with clear
nail varnish. Positive control slides were treated with DNase I (Boehringer Mannheim NZ Ltd., Auckland, New Zealand; 1-100 enzyme units/ml in DNase buffer (30 mM Tris,
140 mM sodium cacodylate, 4 mM MgC12, 0.1 mM dithiothreitol, pH 7.2) at RT for 10 min before the labeling
reaction. Negative control slides were incubated with the
labeling reaction solution as above, but TdT enzyme was
omitted. The incubation in double-strength SSC or citrate
prior to proteinase-K digestion resulted in both time (0-30
min)- and temperature (37-80°C)-dependent increases in
the degree of labeling evident for structures containing
highly condensed chromatin (i.e., apoptotic and pyknotic
cells, and apoptotic bodies). Proteinase-K digestion conditions that resulted in the greatest degree of specific labeling
without marked loss of tissue morphology were determined
for each tissue by titration of enzyme concentration and/or
incubation time.
Sections were first examined unstained by light microscopy using x60 and x100 oil-immersion objectives. Labeled cells and/or structures were individually identified
and either photomicrographed or captured as digitized images using a high-resolution video camera that was mounted on the microscope and connected to a personal computer
running NIH-Image version 1.57 software (NIH, Bethesda,
MD). Slides were then stained with hematoxylin with or
without eosin, and the same fields were reexamined to determine the micromorphological features of the labeled
structures. In addition, some sections were treated in the
reverse order. That is, they were stained with or without
hematoxylin and eosin and individual pyknotic and/or
apoptotic structures were identified; they were then destained by immersion in acid/alcohol (70% ethanol, I %
HCI) for 5-10 min, processed for TUNEL, and reexamined
to determine whether cells or structures classified as apoptotic or pyknotic were labeled.
Experiment 2 was designed to determine the percentage
of cells undergoing apoptosis vs. mitosis within the granulosa cell layer of large antral follicles ( 3 mm) in ewes
during both spontaneous and induced atresia, using semiquantitative histological procedures. Ewes (n = 15) were
injected with prostaglandin (125 mg Estrumate; PitmanMoore, Upper Hutt, New Zealand) twice, 10 days apart,
APOPTOSIS IN OVINE GRANULOSA CELLS
and examined for signs of estrus every 6 h beginning 24 h
after the second prostaglandin injection. From 57 + 0.5 h
SEM: designated TO) after the onset of estrus,
(mean
ewes were injected subcutaneously with either 5 ml of bFF
or saline and killed at TO or at 12 or 24 h later (groups
T12, and T24, respectively; n = 3 ewes per group). Ewes
in group T24 were given a second injection of saline or
bFF 12 h after the first. The bFF used had been pretreated
with dextran-coated charcoal to remove steroids, as previously described [22], and the resulting extract was filtered
through 0.45-mm filter units (Falcon; Becton Dickinson &
Co., Cockeysville, MD) and stored in 5-ml amounts at
-20°C prior to use. Jugular venous blood samples were
collected from each ewe at TO and 6 and 12 h later, then
every 12 h until death, and plasma was removed and stored
frozen at -20C for determination of FSH and LH concentrations by RIA.
Ovaries were removed from ewes immediately after
death, and all follicles - 3 mm in diameter were individually dissected free from extraneous tissue using a stereomicroscope. Follicle diameters were measured to the nearest 0.1 mm with the aid of an eyepiece graticule. Follicles
were then placed in Bouin's fixative for 12 h, processed for
histology, randomly oriented, and embedded in plastic
(Technovit 7100; Kulzer GmbH, Wehrheim, Germany).
Follicles were serially sectioned at a thickness of 30 }pm,
and every fourth section was mounted on a glass slide and
stained with hematoxylin and eosin. The time between
2
death and immersion of follicles in fixative was 13
SEM). The occurrence of recent ovulation
min (mean
was confirmed from the presence of developing corpora
lutea in the ovaries of all ewes.
Morphometric Studies
Stereological methods were used to estimate follicular
volumes and cell numbers as described in detail by Gundersen and Jensen [19] and West and Gundersen [20], and
reviewed by Mayhew [23]. Systematic random sampling
(i.e., with a constant and known periodicity from a random
starting point) was carried out in all three dimensions and
at each level of the sampling scheme (i.e., selection of sections, and dissectors or points within sections). The first
and last sections in which the granulosa cell layer was completely transected were identified for each follicle, and a
systematic random sample of these and the intervening sections (n = 7 to 10 sections per follicle) was selected for
examination. The volumes of the membrana granulosa, theca interna, and follicular antrum were estimated for each
follicle using the Cavalieri principle by point counting on
projected images (x50 magnification) of all selected sections. The membrana granulosa was defined as granulosa
cells forming a clearly identifiable celr layer in apposition
to the basement membrane, including cells that may not
have been attached but were in close proximity to underlying attached cells. In areas where there was a significant
detachment of cells and/or cellular debris from the membrana granulosa into the follicular antrum (as in the case
of some atretic follicles), an additional reference volume
was estimated; this included the membrana granulosa plus
dissociated cells within a distance of approximately the
thickness of the membrana granulosa from its antral border
(referred to as the total granulosa cell volume). This served
as the reference volume for estimating the number of apoptotic cells and apoptotic bodies in atretic follicles that
839
would have been underestimated if the membrana granulosa
only (as defined above) had been considered.
The largest three follicles from each ewe were selected
for cell counting using the optical dissector method [18].
An Olympus (Tokyo, Japan) BH microscope fitted with
x60 and x 100 high numerical aperture oil-immersion objectives, a linear depth gauge unit (Mitutoyo Corporation,
Minato-ku, Tokyo, Japan), and a motorized stage modified
for stereology (Bico Productions, Glostrup, Copenhagen,
Sweden) was used as previously described [24]. Systematic
placement of a square counting frame on the reference volume was accomplished by moving the microscope stage in
a raster pattern from a random start point outside the follicle. The dissector volume (determined by the level of
magnification, the size of the counting frame, and depth of
optical sectioning) was varied, depending upon the prevalence of the cell type or body being counted. For very lowprevalence events, the entire cross-sectional area of the
membrana granulosa or total granulosa cell volume was
examined within a depth of 5 ptm on each section selected
for an individual follicle.
Sampling strategies were designed to optimize efficiency, and estimates of stereological sampling variation (coefficients of error) were calculated as previously described
[19, 20]. Where the prevalence of particles counted was
very low and coefficients of error exceeded 15%, a minimum prevalence representing the limit of detection of the
chosen sampling strategy was assigned for the purposes of
data analysis.
Definition of Cell Types
Using morphological criteria similar to those previously
described [11, 15, 16], an attempt was made to differentiate
and quantitate cells that were actively undergoing apoptosis
at the time of fixation, as distinct from apoptotic bodies
that were considered to represent residual products of the
apoptotic process. Apoptotic cells were defined as cells
with nuclei containing condensed chromatin that either was
marginated into sharply delineated, densely staining masses
aligned with the nuclear membrane (marginated chromatin),
was shrunken into a single regularly shaped, dense, homogeneously staining mass (pyknotic appearance), or was
fragmented into multiple homogeneously and densely staining masses (multiple fragments) clustered together that appeared to have originated from a single cell and that were
situated among and apparently not internalized by neighboring viable cells. Apoptotic bodies were considered to be
remnants of apoptotic cell death not clearly recognizable as
originating from a single cell. They were defined as discrete
membrane-bounded structures containing variable amounts
of condensed chromatin and/or cytoplasm that were situated
singly or in clusters between apparently viable cells, internalized by neighboring cells or macrophages, or present in
the follicular antrum. Only discrete structures that were
clearly identifiable using a x60 oil-immersion objective
were counted. Apoptotic bodies that occurred in clusters
immediately adjacent to one another were counted as a single event.
In addition, granulosa cells that exhibited the morphological features characteristic for prophase, metaphase, anaphase, or telophase stages of mitosis [25] were identified
and counted. Cells in telophase were counted as a single
cell if one or both of the newly divided cells fell within the
dissector.
840
JOLLY ET AL.
FIG. 1. Cells with characteristic morphological features of apoptosis in the membrana granulosa of ovine follicles 3 mm in diameter. These include:
(a and b) cells with nuclei containing marginated chromatin (small arrows); (c and d) cells with a single small nucleus with distinctly dense homogeneously staining chromatin (pyknotic appearance; small arrows) (e and f) cells containing multiple smaller, densely staining nuclear fragment (small
arrows); and (g and h) discrete membrane-bounded structures containing variable amounts of condensed chromatin and/or cytoplasm (apoptotic bodies;
small arrows). Tissue was fixed in Bouin's solution and stained with hematoxylin and eosin. Curved arrows denote basement membrane separating the
membrana granulosa from theca interna (b and f. a-h) x950.
Histological Assessment of Ovarian Follicles
In both experiment 1 and experiment 2, the criteria of
Hay et al. [26] were used to classify follicles into one of
four groups depending upon the degree of degeneration evident in the granulosa cell layer in hematoxylin- and eosinstained sections. Follicles classified as healthy had a membrana granulosa that appeared compact and well organized,
with closely apposed cells, numerous mitotic figures, and
only occasional or rare pyknotic cells, or "densely staining" (atretic) bodies. Primary atresia was characterized by
the presence of a moderate number of degenerate cells and/
or atretic bodies, widely distributed along the antral border
of the membrana granulosa. Secondary atresia was characterized by the presence of numerous degenerate cells and/
or atretic bodies within and/or along the antral border of
the membrana granulosa, and relatively few mitotic cells.
Tertiary atresia was characterized by widespread disintegration of the membrana granulosa and rarity or absence of
mitotic cells.
Hormone Assays
FSH. FSH concentrations in plasma were determined by
RIA using antiserum (NIADDK-anti-oFSH-1; AFPC5288113), iodinated tracer (NIADDK-oFSH-I-1; AFP5679C), and reference preparation (USDA-oFSH-19SIAFP-RP-2; AFP-4117A) supplied by the National Hor-
mone and Pituitary Program, Baltimore, MD. The internal
standards and standard curve samples were prepared in
plasma from a hypophysectomized ewe. All samples from
this study were processed in a single assay, which had a
sensitivity of 0.2 ng/ml and an intraassay coefficient of
variation of 14.8%.
LH. LH concentrations in plasma were determined by
RIA using an antiserum raised at Wallaceville [27]. The
iodinated tracer (NIADDK-oLH-I-3; AFP-9598B) and reference preparation (NIAMDD-oLH-24; AFP-0754) was
supplied by the National Hormone and Pituitary Program.
The internal standards and standard curve samples were
prepared in plasma from a hypophysectomized ewe. All
samples from this study were processed in a single assay,
which had a sensitivity of 0.2 ng/ml and an intraassay coefficient of variation of 13.4%.
Statistical Procedures
Statistical procedures used in the calculation of stereological estimates are described in detail elsewhere [19, 20].
Counts and proportions of follicles within each classification (e.g., healthy vs. atretic, as detailed above) within ewe
were analyzed for effects of time, treatment, and time by
treatment interactions, or stage of atresia, using log-linear
and logistic regression models and analyses of deviance.
Approximate F-tests using variation due to animal (nested
APOPTOSIS IN OVINE GRANULOSA CELLS
841
within level of time and treatment) as the error term were
constructed [28] to test for all effects because of overdispersion of the data. Parameter estimates (e.g., means for
each group) were derived from these models and backtransformed with appropriate confidence intervals for presentation of results. The General Factorial ANOVA procedure of SPSS [29] was used for all analyses of variance
and covariance. Data expressed as percentages were transformed by taking their square root to conform with assumptions of normality and constant variance. The effect
of stage of atresia on total granulosa cell number, follicular
volumes, and volume ratios was determined by analysis of
covariance, adjusting for variation due to follicle diameter
where significant. Standard linear regression procedures
were used to determine the relationship between the percentages of cells undergoing apoptosis vs. mitosis.
RESULTS
Experiment 1: Morphological and In Situ 3' End-Labeling
Evidence of Granulosa Cell Apoptosis in Ovine Follicles
Apoptotic cells and apoptotic bodies were evident in
both the membrana granulosa and theca interna of follicles
2 3 mm in diameter collected from ewes during the midluteal phase of the estrous cycle. These included cells with
nuclei containing marginated chromatin (Fig. 1, a and b),
cells with a single small densely staining nucleus (pyknotic
appearance, Fig. 1, c and d), and cells containing multiple
smaller, densely staining nuclear fragments (Fig. 1, e and
f). Apoptotic bodies, consistent with the appearance of
structures resulting from the fragmentation of apoptotic
cells, were also clearly evident (Fig. 1, g and h). These
were situated singly or in clusters between apparently viable cells, were internalized by neighboring cells, or were
present in the follicular antrum, or lumina of lymphatic,
capillary, or small blood vessels in the theca interna.
TUNEL clearly revealed that cells and subcellular structures with the above-mentioned morphological features
contained fragmented DNA (Fig. 2). The labeling procedure was specific: no labeling was observed in any of the
negative control slides (e.g., Fig. 3), whereas all cells were
labeled in DNase-treated positive controls (data not shown).
However, both the intensity of labeling and the clarity of
nuclear morphology varied with method of fixation and
with the prelabeling digestion conditions employed. Nuclear morphology was very distinct in Bouin's-fixed tissue, but
condensed chromatin labeled only faintly (if at all) unless
high levels of proteinase K (20 mg/ml) were used. Unfortunately, when high levels of proteinase K were used, much
of the surrounding cellular and tissue architecture was destroyed. At moderate levels of proteinase-K digestion (2
mg/ml), labeling was most obvious in the cytoplasm or in
cytoplasmic remnants of dead or dying cells, presumably
due to leakage of DNA fragments from degenerating nuclei
in dying cells, and/or lysosomal breakdown of phagocytosed nuclear fragments within healthy cells. Although condensed chromatin in apoptotic cells and apoptotic bodies
was clearly labeled in sections from follicles that had been
fixed in 4% paraformaldehyde (Fig. 3), cellular and nuclear
morphology in these sections was much less distinct than
in the Bouin's-fixed tissue. In paraformaldehyde-fixed follicles, the prevalence of labeled structures was higher than
in apoptotic cells and/or apoptotic bodies identified by hematoxylin and eosin staining. This was presumably due to
specific labeling of DNA fragments within cytoplasmic
remnants of apoptotic cells, or of phagocytosed nuclear
FIG. 2. In situ 3' end-labeling of apoptotic cells and/or apoptotic bodies
(brown coloration), counterstained with hematoxylin and eosin (a-e) or
hematoxylin alone (f), in the membrana granulosa and theca interna of
3 mm in diameter fixed in Bouin's solution. Labeling
ovine follicles
consistent with the presence of fragmented DNA was evident in cells
containing marginated chromatin (large arrowheads), a single small (pyknotic) nucleus (small arrowheads), multiple small nuclear fragments (solid
arrow), and apoptotic bodies (open arrow), shown here after internalization by neighboring cells. x1370.
remnants within the cytoplasm of apparently healthy cells,
that did not stain with hematoxylin.
During primary and secondary atresia, apoptotic cells
were disseminated throughout the granulosa and theca cell
layers (Fig. 4), though in some follicles they appeared to
be more prevalent in some regions of follicle wall than in
others. Within the membrana granulosa, apoptotic cells and
apoptotic bodies were evident in all layers from the basement membrane (Fig. 1, b and f) to the antral cavity but
were usually more prevalent among cells close to the an-
842
JOLLY ET AL.
FIG. 3. a) In situ 3' end-labeling of apoptotic cells and/or apoptotic
bodies (brown coloration) in the membrana granulosa of a paraformaldehyde-fixed follicle in the tertiary stage of atresia. b) An adjacent section
of the same follicle processed identically except that the labeling enzyme
(TdT) was omitted (negative control). Sections counterstained with hematoxylin. x500.
trum. Occasionally, apoptotic cells were observed immediately adjacent to cells undergoing mitosis. Clusters containing what appeared to be cells undergoing apoptosis as
well as apoptotic bodies were also occasionally evident
(e.g., Fig. lh), and these became increasingly prevalent
with advancing degree of atresia. In tertiary-atretic follicles,
the antral layers of what remained of the membrana granulosa consisted almost entirely of pyknotic nuclei and apoptotic bodies, though cells with marginated chromatin and
multiple nuclear fragments (presumably earlier stages of
apoptotic degeneration) were still evident in basal layers.
On the basis of morphological and in situ 3' end-labeling
data, apoptosis appeared to be the predominant form of cell
death evident in the membrana granulosa, at least during
the primary and secondary stages of atresia in ewes. However, cells that exhibited morphological features more typical of oncosis (cell swelling, irregular clumping of chromatin, and/or karyolysis) were occasionally seen in 9 of 46
Bouin's-fixed follicles (-20%): in these cells 3' end-labeling was either very faint or undetectable in both nuclear
and cytoplasmic compartments (Fig. 5). Cells with these
features occurred singly among other apparently healthy
cells, and in the absence of lymphocyte or neutrophil infiltration. During the later stages of atresia, cells exhibiting
similar features were also occasionally evident along the
antral border of the membrana granulosa or within the antral cavity.
TABLE 1. Proportion of follicles -3 mm in diameter per ewe classified
as healthy by histological criteria in ewes injected with either saline or
bFF twice, 12 h apart, in relation to time from first injection.a
Time from first injection
Group
Saline
bFF
Oh
0.64 (0.36, 0.85)
(n = 23)
12 h
0.29 (0.12, 0.57)
(n = 24)
0.29 (0.08, 0.64)
(n= 14)
24 h
0.23 (0.07, 0.54)
(n = 19)
0.00
b
(n = 17)
a Data presented as the mean with 95% confidence interval in parentheses.
b
Confidence intervals could not be estimated.
FIG. 4. Disseminated pattern of apoptotic cells and/or apoptotic bodies
evident within the membrana granulosa and theca interna of a follicle in
the secondary stage of atresia by in situ 3' end-labeling: (a) X38, (b)
X180. Arrows denote basement membrane. Ovary fixed in paraformaldehyde and counterstained with hematoxylin.
Experiment 2: Effect of bFF on Plasma FSH
Concentrations and Follicle Health
Plasma FSH concentrations varied with time and treatment in the same manner that we have reported previously
[1]. In bFF-treated ewes, plasma FSH concentrations decreased to basal levels (-50% of levels at TO, mean +
SEM: 0.29 ± 0.06 ng/ml) within 12 h and thereafter remained low, but they did not vary significantly with time
in saline-treated control ewes. Plasma LH concentrations
did not vary significantly (p > 0.05) with either time or
treatment (overall mean + SEM: 0.43 ± 0.03 ng/ml).
The total number of follicles 2 3 mm in diameter per
ewe did not vary significantly with either time or treatment,
and averaged 6.5 (5.5, 7.6: 95% confidence interval). However, the proportion of these follicles that were classified
histologically as healthy varied as a function of both time
and treatment (time by treatment interaction, p = 0.015).
The proportion of follicles per ewe that were classified as
healthy decreased over time in control ewes, but it decreased to a greater extent in ewes that had been treated
with bFF (Table 1). By T24, all follicles - 3 mm in diameter present in bFF-treated ewes were classified as atretic. Follicles - 8 mm in diameter all showed evidence of
luteinization or had become cystic and were excluded from
analyses. No significant time or treatment effects (or interactions) were evident for the proportion of follicles per ewe
APOPTOSIS IN OVINE GRANULOSA CELLS
843
that were classified as being in the primary, secondary, or
tertiary stages of atresia (data in the last two categories
were pooled as too few tertiary-atretic follicles were present
for separate analysis). The overall mean proportions of follicles per ewe classified in the primary and secondary or in
the tertiary stages of atresia were 0.25 (0.15, 0.38) and 0.44
(0.32, 0.57), respectively (means and 95% confidence intervals).
The diameter of the largest follicle per ewe measured at
dissection (excluding follicles
8 mm in diameter) also
did not vary significantly with either time or treatment and
averaged 5.3 + 0.2 mm (mean ± SEM) overall.
Prevalence of Apoptotic Cells and Mitotic Cells in
Healthy and Atretic Follicles
The stereological methods used in this study resulted in
unbiased estimates of the total number of granulosa cells
and of the volumes of the membrana granulosa, theca interna, and follicular antrum for each follicle; these estimates
had coefficients of error (mean + SEM) of 0.070 ± 0.002,
0.068 ± 0.002, 0.032 + 0.001, and 0.078 + 0.003, respectively. Estimates of the prevalence of cells classified as
apoptotic or mitotic (expressed as percentage of total granulosa cell number for each follicle) above 0.02% were determined with coefficients of error < 15% (mean ± SEM:
0.096 + 0.005 and 0.099
0.005, respectively). Prevalences c 0.02% were too low for precise estimates to be
determined using the sampling methods chosen and were
assigned this minimum value for the purposes of statistical
analysis.
The prevalence of cells classified as apoptotic or mitotic
varied with stage of atresia (both p < 0.001), and this effect
was similar for control and bFF-treated ewes (atresia status
by treatment interactions, both p > 0.10). Data were therefore pooled across treatments and summarized by atresia
status (Table 2). There was only one follicle in the tertiary
stage of atresia among the follicles chosen for stereological
analysis, and data from this follicle were excluded from
analyses. Estimates of the prevalence of mitotic cells, apoptotic cells, and apoptotic bodies for this follicle were s
0.02%, 3.65%, and 71.59%, respectively.
Apoptotic cells and apoptotic bodies were observed in
the membrana granulosa of all follicles examined. However, the prevalence of apoptotic cells and apoptotic bodies
observed in follicles classified as healthy was very low (0.03% on average). Evidence of granulosa cell apoptosis
increased markedly with advancing stages of atresia; concomitantly there was a marked decrease in the prevalence
of cells undergoing mitosis (Table 2). In follicles in primary
and secondary stages of atresia, the maximum percentages
of cells observed to be undergoing apoptosis at the time of
FIG. 5. Degenerating granulosa cell (arrow) exhibiting morphological
features of oncosis (cell swelling, irregular clumping of chromatin, and/or
karyolysis) in an atretic ovine follicle. Bouin's fixative, hematoxylin staining, 1280.
fixation were 0.5% and 1.5%, respectively. Increases in the
prevalence of apoptotic cells with advancing stage of atresia were associated with decreases in total granulosa cell
number and in the volumes of the membrana granulosa and
theca interna (Table 3). The ratios of membrana granulosa
to theca interna volume, as well as of membrana granulosa
to follicular antrum volume, also decreased with advancing
stage of atresia (Table 3). Objective criteria (based on the
overall prevalence of apoptotic structures observed [apoptotic cells + apoptotic bodies; expressed as a percentage of
total granulosa cell number]) resulting in discrimination of
follicles that was similar to the subjective classification according to stage of atresia used in this study were 0.020.20% (healthy), 0.21-2.00% (primary atresia), and >
2.00% (secondary atresia). The percentages of mitotic cells
were > 0.5% in all but one follicle classified as healthy
and < 1% in all follicles classified as atretic.
As observed in experiment 1, apoptotic cells appeared
to be disseminated throughout the membrana granulosa and
were evident in all layers extending from the basement
membrane to the antral cavity. Apoptotic bodies were more
prevalent toward the antrum and within the antral cavity
proper. There was a highly significant inverse relationship
within individual follicles between the prevalence of cells
undergoing apoptosis and the prevalence of cells undergoing mitosis (Fig. 6: R2 = 0.64).
TABLE 2. Prevalence of apoptotic cells, mitotic cells, and apoptotic bodies
within the membrana granulosa of ovine follicles, classified by stage of atresia.a
Stage of atresia
Cell type
Mitotic cells
Apoptotic cells
Apoptotic bodies
Healthy
Primary
Secondary
0.99 (0.84, 1.15)
[0.18-1.77]
0.02 (0.01, 0.03)
[0.02-0.13]
0.03 (0.02, 0.06)
[0.02-0.29]
0.50 (0.36, 0.66)
[0.1 5-0.89]
0.24 (0.14, 0.37)
[0.04-0.54]
0.47 (0.33, 0.62)
[0.15-0.89]
0.29 (0.20, 0.39)
[0.16-0.691
0.89 (0.67, 1.13)
[0.42-1.51]
2.26 (1.64, 2.98)
[1.00-4.31]
a Data expressed as the mean percentage of total granulosa cell number for each
follicle, (95% confidence interval), and [range].
844
JOLLY ET AL.
' An
4.u
t
1.5
o
U)
1.0
Bo
0
-
O
o0
.
O
O Ce
0.5
nn
v.v
>q
4
0
I
0.0
0.5
1.0
1.5
2.0
Apoptotic Cells (sqrt %)
FIG. 6. Relationship between the prevalence of cells undergoing mitosis
3 mm in diameter (exvs. apoptosis within individual ewe follicles
pressed as the square root of the percentage of total granulosa cell number). Follicles were classified by stage of atresia: healthy (circles), primary
(diamonds), secondary (upright triangles), or tertiary (inverted triangle).
0.06 x; R2 = 0.64.
Regression: y = 1.05 0.04 - 0.55
DISCUSSION
Results of this study extend classic histological descriptions of granulosa cell degeneration during follicular atresia
in ewes (nuclear pyknosis, hyperchromatosis, karyorrhexis,
and the formation of atretic bodies) [26, 30, 31] by relating
specific morphological features of granulosa cell death to
the physiological process of apoptosis. This relationship is
supported in this study by in situ histochemical evidence
of DNA fragmentation (a hallmark feature of apoptosis).
These results are also supported by the demonstration of
oligonucleosome formation in granulosa cell DNA purified
from follicles collected from similarly treated ewes in another study, conducted using the same animal model system, the bFF-treated ewe [1]. Morphological evidence of
apoptotic cell death has previously been described for theca
interna cells during the process of atresia in ewes [4]. Our
results provide histochemical evidence that supports this
finding. These morphological, in situ histochemical, and
biochemical findings in ewes are also consistent with results of our recent studies in cows ([2]; unpublished results).
Quite remarkably, morphological features of degenerating granulosa cells identical to those described and depicted
here were first published more than 100 yr ago by a German
scientist, Walther Flemming, who studied the involution of
ovarian follicles in rabbits [32]. Flemming's study has been
brought to light in a recent review [17], and it has been
proposed that this was the first study to introduce the concept of spontaneous cell death as a physiological event.
Flemming named the process he described "chromatolysis," and the drawings he used to illustrate his findings
clearly depict all of the morphological features considered
to be characteristic of the process now known as apoptosis
(reproduced in [17]). These include cells with a single
shrunken and densely basophilic nucleus (pyknotic appearance). While such cells do not display the classically described features of apoptosis (margination of chromatin and/or nuclear fragmentation [11, 15]), they probably represent a morphological variant of apoptotic cell death that
has been described for a variety of cell types [12, 15, 33,
34]. Consistent with this view, our data clearly show the
presence of fragmented DNA in cells with this morphological appearance, which also agrees with the findings of others using an in vitro model of induced apoptosis in cultured
cells [34].
We have previously demonstrated the occurrence of
apoptosis among ovine granulosa cells that were classified
as healthy by morphometric criteria and that retained the
capacity to synthesize estradiol and to respond to both FSH
and LH by producing the intracellular second messenger,
cAMP [1]. We concluded from that study that an increase
in the prevalence of granulosa cell apoptosis is a very early
event in the process of ovarian follicular atresia, evident
before other changes in morphological or biochemical indices of follicular status or granulosa cell function became
apparent. In the present study, we determined the prevalence of apoptotic cells and apoptotic bodies in relation to
the prevalence of cells undergoing mitosis during both
spontaneous and induced follicular atresia, using the same
animal model system. As in our previous study, a high proportion (0.64) of follicles present at the beginning of the
experiment (approximately 60 h postestrus) were classified
as healthy. In both studies, treatment with bFF resulted in
a rapid decline in circulating FSH concentrations and an
increase in the degree of granulosa cell apoptosis (presum3 mm in
ably due to the onset of atresia) in all follicles
diameter within 24 h.
The stereological methods used in the present study provided unbiased estimates of the absolute number (hence
prevalence when expressed as a percentage of total number
of granulosa cells) of apoptotic cells and apoptotic bodies
detected by light microscopy in hematoxylin- and eosinstained sections. To our knowledge, similar data have not
previously been reported. One notable finding was the presence of apoptotic cells and/or apoptotic bodies (albeit very
few) in all follicles that were classified as healthy. This is
consistent with similar histological findings in cows (unpublished results), as well as with descriptions of occasional pyknotic nuclei and atretic bodies in ovine follicles classified as healthy by others [26]. The prevalence of apoptotic
cells and apoptotic bodies increased markedly as atresia
progressed and was highly correlated with a decrease in the
prevalence of cells undergoing mitosis. These changes were
reflected by decreases in total granulosa cell number as well
TABLE 3. Effect of stage of atresia on mean granulosa cell numbers and intrafollicular volumes in ovine
3 mm in diameter.
follicles
Pooled
Stage of atresia
Variable
Granulosa cell no. (x106)
3
Membrana granulosa volume (mm )
3
)
(mm
volume
Theca interna
Membrana granulosa:theca interna volume ratio
Membrana granulosa:antrum volume ratio
Healthy
Primary
Secondary
SEM
Probabilityb
6.33
3.33
2.39
1.37
0.118
6.16
2.55
2.20
1.18
0.090
5.17
2.30
1.96
1.21
0.081
0.17
0.07
0.05
0.03
0.002
<0.001
<0.001
0.011
0.010
<0.001
a Means adjusted for variation due to follicular diameter.
b
Significant effect of stage of atresia.
APOPTOSIS IN OVINE GRANULOSA CELLS
as in estimates of the volume of the membrana granulosa
in atretic follicles. Although point prevalences of apoptotic
cells evident in atretic follicles averaged < 1%, the values
obtained are consistent with rates of cell loss on the order
of 10% per day if the visible duration of apoptotic cell
death discernible histologically is approximately 2 h [12].
The prevalences of cells undergoing mitosis determined
in our study are similar to previously reported estimates for
ovine follicles [31, 35]. The prevalences of cells undergoing apoptosis estimated for atretic follicles in our study are
similar to those determined in histological studies of regression of the ventral prostate gland after castration in rats
[16], and to spontaneous rates of cell death in a wide range
of tumor types [36]. However, they are considerably lower
than prevalences of granulosa cells with subdiploid levels
of DNA determined by flow cytometry in studies on porcine [37] and bovine follicles [38]. These differences may
be explainable by differences between species, in criteria
used to define stage of atresia (thus in the actual stage of
atresia at which follicles were examined), and by the likelihood that flow cytometric methods fail to distinguish between apoptotic cells and apoptotic bodies. From the single
follicle in the tertiary stage of atresia that was examined in
our study, we noted a very high prevalence of apoptotic
bodies (72%, expressed as a percentage of total granulosa
cell number). Apoptosis of a single cell may give rise to
many apoptotic bodies, which appear to accumulate in the
follicular antrum as atresia progresses, while at the same
time the number of healthy granulosa cells decreases. Estimates of degree of cell death that do not distinguish between single cell events (e.g., apoptotic cells as defined in
our present study) and the multiple products of apoptosis
(e.g., apoptotic bodies) may grossly overestimate the actual
prevalence of cell death. This comment may also be relevant to histological studies reporting prevalences of pyknotic nuclei, if these do not distinguish between apoptotic
cells and apoptotic bodies.
It appears from our observations that apoptosis is the
predominant form of cell death occurring in the membrana
granulosa of ovarian follicles, at least during the initial
stages of follicular atresia in ewes. However, cells with
morphological features more typical of oncosis [17] or necrosis were occasionally seen, particularly during the later
stages of atresia. Oncosis has recently been reported among
granulosa cells in rat follicles treated with interleukin-lpconverting enzyme-related protease inhibitors in vitro that
suppressed internucleosomal DNA cleavage associated with
apoptosis but failed to prevent extensive cellular degeneration [39]. Oncosis is thought to represent a mechanism of
cell death, distinctly different from apoptosis, that is associated with disruption of cellular ATP generation or increases in the permeability of the plasma membrane as typically
occurs following ischemia or toxic injury [17]. Whether
oncosis plays a significant physiological role in granulosa
cell deletion during follicular atresia remains to be determined. Alternatively, cells displaying features similar to
those described for oncosis may have been undergoing necrotic degeneration secondary to apoptotic cell death, for
which the term apoptotic necrosis has been proposed [17].
In addition to specific labeling of cells undergoing apoptosis, TUNEL-positive cells have been reported in hepatic
tissue undergoing autolysis or necrosis induced by cytotoxic drugs in rats [40]. These authors questioned the specificity of the TUNEL assay for apoptotic cell death and emphasized the importance of concurrent demonstration of the
distinct micro-morphological features of apoptosis in de-
845
generating cells. In our study we demonstrated both specific
labeling and the distinct morphological features of apoptosis in the same cells by dual processing of tissue sections
and with the aid of photomicrographs and image analysis.
Moreover, our morphological findings in this study are consistent with our demonstration of oligonucleosome formation in ovine granulosa cells using the same bFF-treated
ewe model [1]. Taken together, our results provide solid
evidence of apoptosis among ovine granulosa cells during
follicular atresia in ewes.
From both morphological and in situ 3' end-labeling results in our study, apoptotic cells appeared to be widely
disseminated throughout the membrana granulosa, including the cell layer adjacent to the basement membrane. However, a higher prevalence of apoptotic cells and apoptotic
bodies was observed in the antral layer of the membrana
granulosa, particularly in more markedly atretic follicles.
This need not reflect a higher incidence of apoptosis among
antral cells but may result from the extrusion of apoptotic
debris toward the follicular antrum where it progressively
accumulates. We have observed a similar distribution of
apoptotic cells in bovine follicles (unpublished results) and
suggest that during the early stages of atresia, the incidence
of apoptosis within the membrana granulosa (excluding the
cumulus oophorus, which was not systematically examined
in our studies) is essentially random. The widespread distribution of apoptotic cells within the membrana granulosa,
together with the relatively low point prevalence of apoptotic cell death and concurrent evidence of ongoing mitosis
in atretic follicles, is consistent with the notion that atresia
is not a sudden coordinated process involving concerted
death of all granulosa cells [41] but rather a process determined by a dynamic equilibrium between factors regulating
granulosa cell division, differentiation, and death [42].
ACKNOWLEDGMENTS
We thank the NIADDK, NIH, Bethesda, MD, for ovine pituitary gonadotropins and RIA reagents; Dr. Heywood Sawyer for his advice; Mr.
D. Jensen and Mr. R. Bailey for care of experimental animals; Mr. G.
De'ath for statistical advice; Mr. A. Barkus and Ms. R. Kiel for preparation
of figures; Miss T. Smith for technical assistance; and Ms. S. Swaney for
secretarial assistance.
REFERENCES
1. Jolly PD, Tisdall DJ, De'ath G, Heath DA, Lun S, Hudson NL,
McNatty KP Granulosa cell apoptosis, aromatase activity, cyclic adenosine 3',5'-monophosphate response to gonadotropins, and follicular
fluid steroid levels during spontaneous and induced follicular atresia
in ewes. Biol Reprod 1996; 56:830-836.
2. Jolly PD, Tisdall DJ, Heath DA, Lun S, McNatty KP. Apoptosis in
bovine granulosa cells in relation to steroid synthesis, cyclic adenosine
3',5'-monophosphate response to follicle-stimulating hormone and luteinizing hormone, and follicular atresia. Biol Reprod 1994; 51:934944.
3. Hsueh AJW, Billig H, Tsafriri A. Ovarian follicle atresia: a hormonally
controlled apoptotic process. Endocr Rev 1994; 15:707-724.
4. O'Shea JD, Hay MF, Cran DG. Ultrastructural changes in the theca
interna during follicular atresia in sheep. J Reprod Fertil 1978; 54:
183-187.
5. Sawyer HR, Niswender KD, Braden TD, Niswender GD. Nuclear
changes in ovine luteal cells in response to PGF2 ,. Domest Anim
Endocrinol 1990; 7:229-238.
6. McGuire WJ, Juengel JL, Niswender GD. Protein kinase C second
messenger system mediates the antisteroidogenic effects of prostaglandin F2, in the ovine corpus luteum in vivo. Biol Reprod 1994;
51:800-806.
7. Murdoch WJ. Temporal relationships between stress protein induction,
progesterone withdrawal and apoptosis in corpora lutea of ewes treated with prostaglandin-F2,. J Anim Sci 1995; 73:1789-1792.
846
JOLLY ET AL.
8. Rueda BR, Wegna JA, Marion SL, Wahlen DD, Hoyer PB. Internucleosomal DNA fragmentation in ovine luteal tissue associated with
luteolysis: in vivo and in vitro analyses. Biol Reprod 1995; 52:305312.
9. Ackerman RC, Murdoch WJ. Prostaglandin-induced apoptosis of
ovarian surface epithelial cells. Prostaglandins 1993; 45:475-485.
10. Murdoch WJ. Programmed cell death in preovulatory ovine follicles.
Biol Reprod 1995; 53:8-12.
11. Kerr JFR, Wyllie AH, Currie AR. Apoptosis: a basic biological phenomenon with wide ranging implications in tissue kinetics. Br J Cancer 1972; 26:239-257.
12. Bursch W, Taper HS, Lauer B, Schulte-Hermann R. Quantitative histological and histochemical studies on the occurrence and stages of
controlled cell death (apoptosis) during regression of rat liver hyperplasia. Virchows Arch B Cell Pathol 1985; 50:153-166.
13. Gavrieli Y, Sherman Y, Bensasson SA. Identification of programmed
cell death in situ via specific labeling of nuclear DNA fragmentation.
J Cell Biol 1992; 119:493-501.
14. Palumbo A, Yeh J. In situ localization of apoptosis in the rat ovary
during follicular atresia. Biol Reprod 1994; 51:888-895.
15. Wyllie AH, Kerr JFR, Currie AR. Cell death: the significance of apoptosis. Int Rev Cytol 1980; 68:251-306.
16. English HF, Kyprianou N, Isaacs JT Relationship between DNA fragmentation and apoptosis in the programmed cell death in the rat prostate following castration. Prostate 1989; 15:233-250.
17. Majno G, Joris I. Apoptosis, oncosis, and necrosis: an overview of
cell death. Am J Pathol 1995; 146:3-15.
18. Kerr JFR, Winterford CM, Harmon BV. Apoptosis-its significance
in cancer and cancer therapy. Cancer 1994; 73:2013-2026.
19. Gundersen HJG, Jensen EB. The efficiency of systematic sampling in
stereology and its prediction. J Microsc 1987; 147:229-263.
20. West MJ, Gundersen HJG. Unbiased stereological estimation of the
number of neurons in the human hippocampus. J Comp Neurol 1990;
296:1-22.
21. Tisdall DJ, Hudson N, Smith P, McNatty KP. Localization of ovine
follistatin and alpha and beta(A) inhibin mRNA in the sheep ovary
during the oestrous cycle. J Mol Endocrinol 1994; 12:181-193.
22. Henderson KM, Prisk MD, Hudson N, Ball K, McNatty KP, Lun S,
Heath D, Kieboom LE, McDiarmid J. Use of bovine follicular fluid
to increase ovulation rate or prevent ovulation in sheep. J Reprod
Fertil 1986; 76:623-635.
23. Mayhew TM. The new stereological methods for interpreting functional morphology from slices of cells and organs. Exp Physiol 1991;
76:639-665.
24. Smith P, Braw-Tal R, Corrigan K, Hudson NL, Heath DA, McNatty
KP. Ontogeny of ovarian follicle development in Booroola sheep fetuses that are homozygous carriers or non-carriers of the Fec(B) gene.
J Reprod Fertil 1994; 100:485-490.
25. Bloom W, Fawcett DW. A Textbook of Histology. Philadelphia, PA:
W.B. Saunders Co.; 1975.
26. Hay MF, Cran DG, Moor RM. Structural changes occurring during
atresia in sheep ovarian follicles. Cell Tissue Res 1976; 169:515-529.
27. McNatty KP, Hudson N, Henderson KM, Gibb M, Morrison L, Ball
K, Smith P. Differences in gonadotrophin concentrations and pituitary
responsiveness to GnRH between Booroola ewes which were homozygous (FF), heterozygous (F+), and non-carriers (++) of a major
gene influencing their ovulation rate. J Reprod Fertil 1987; 80:577588.
28. McCullagh P, Nelder JA. Generalized Linear Models. London: Chapman and Hall; 1989.
29. SPSS. SPSS Advanced Statistics 6.1. Chicago, IL: SPSS Inc.; 1994.
30. Brand A, deJong WHR. Qualitative and quantitative micromorphological investigations of tertiary follicle population during the oestrous
cycle in sheep. J Reprod Fertil 1973; 33:431-439.
31. Turnbull KE, Braden AWH, Mattner PE. The pattern of follicular
growth and atresia in the ovine ovary. Aust J Biol Sci 1977; 30:229241.
32. Flemming W. Uber die bildung von richtungsfiguren in saugethiereiern beim untergang graaf'scher follikel. Arch Anat Entwickl 1885;
221-244.
33. Faa G, Leddacolumbano GM, Ambu R, Congiu T, Coni P, Riva A,
Columbano A. An electron microscopic study of apoptosis induced
by cycloheximide in rat liver. Liver 1994; 14:270-278.
34. Kressel M, Groscurth P. Distinction of apoptotic and necrotic cell
death by in situ labeling of fragmented DNA. Cell Tissue Res 1994;
278:549-556.
35. Cahill LP, Mauleon P Influences of season, cycle and breed on follicular growth rates in sheep. J Reprod Fertil 1980; 58:321-328.
36. Sarraf CE, Bowen ID. Proportions of mitotic and apoptotic cells in a
range of untreated experimental tumours. Cell Tissue Kinet 1988; 21:
45-49.
37. Guthrie HD, Cooper BS, Welch GR, Zakaria AD, Johnson LA. Atresia
in follicles grown after ovulation in the pig: measurement of increased
apoptosis in granulosa cells and reduced follicular fluid estradiol-173.
Biol Reprod 1995; 52:920-927.
38. Blondin P, Dufour M, Sirard M. Analysis of atresia in bovine follicles
using different methods: flow cytometry, ELISA, and classic histology. Biol Reprod 1996; 54:631-637.
39. Flaws JA, Kugu K, Trbovich AM, DeSanti A, Tilly KI, Hirshfield
AN. Interleukin-lb-converting enzyme-related proteases (IRPs) and
mammalian cell death: dissociation of IRP-induced oligonucleosome
endonuclease activity from morphological apoptosis in granulosa cells
of the ovarian follicle. Endocrinology 1995; 136:5042-5053.
40. Grasl-Kraupp B, Ruttkay-Nedecky B, Koudelka H, Bukowska K,
Bursch W, Schulte-Hermann R. In situ detection of fragmented DNA
(TUNEL Assay) fails to discriminate amongst apoptosis, necrosis and
autolytic cell death: a cautionary note. Hepatology 1995; 21:14651468.
41. Hirshfield AN. Rescue of atretic follicles in vitro and in vivo. Biol
Reprod 1989; 40:181-190.
42. Tsafriri A, Braw RH. Experimental approaches to atresia in mammals.
Oxf Rev Reprod Biol 1984; 6:226-265.