Structural changes occurring during atresia in sheep ovarian follicles

Summary The structural changes that characterize primary, secondary and tertiary atresia in sheep Graafian follicles have been studied by means of histological, histochemical and ultrastructural techniques.In primary atresia vacuoles representing swollen endoplasmic reticulum are prominent along the antral border together with disorganized granulosa cells containing pyknotic nuclei. Phagocytic cells, which increase in number as atresia progresses, were seen within the membrana granulosa and are considered to be transformed granulosa cells. Even in follicles classified as nonatretic, a few antral vacuoles and occasional pyknotic nuclei are present.During secondary atresia there is a large increase in the number of cells with pyknotic nuclei; many of these nuclei had been extruded and had fused to form the characteristic Feulgen-positive atretic bodies found along the edge of the antral cavity. These bodies usually have a diameter of up to 15 m but occasionally reached as much as 400 m. A second area of degeneration is frequently present in the membrana granulosa, two or three cell layers from the basal lamina, and it is at this level that exfoliation of granulosa cells occurs in tertiary atresia. In contrast to the membrana granulosa, there are during secondary atresia, only slight indications of degeneration in the cumulus.In tertiary atresia the membrana granulosa is highly disorganized; the atretic bodies are often fewer in number than at earlier stages. The basal lamina remains essentially intact. It is at this stage that the first clear signs of degeneration occur in the theca interna. Despite some disintegration of the cumulus, the integrity of the oocyte is maintained and its nucleus remains vesicular.Changes in the thecal microcirculation may play a key role in atresia: adjacent to the basal lamina of non-atretic follicles, there is a well-developed capillary network which is significantly reduced as atresia progresses.The authors are greatly indebted to Dr. H.M. Dott and Mr. G.C. Foster for carrying out the analysis with the Quantimet image analysing computer. The skilled technical assistance of Mrs. Linda Collins is also gratefully acknowledged     

Mechanism of granulosa cell death during follicular atresia depends on follicular size

Changes in granulosa cell lysosomal and mitochondrial functions in relation to follicular size and to the stage of atresia were studied by fluorescent emission spectra and intensity using flow cytometry. Antral follicles were grouped by size in two groups: small, 3-6 mm and large, >6mm in diameter, and classified into three stages of atresia: non-atretic, initially atretic and advanced atretic. Differences in Rhodamine 123 (Rh123) and Acridine Orange (AO) fluorescent intensity indicated that changes in mitochondrial function are the primary mechanism of granulosa cell death in atretic follicles 3-6 mm in diameter, while its role in granulosa cell death in >6 mm atretic follicles seemed to be less important. However, modifications in lysosomal function (shown by a decrease in fluorometric intensity of AO incubated granulosa cells) were mainly associated with cell death in large atretic follicles. Our results support the hypothesis that the pathway of granulosa cell death during follicular atresia depends on the state of energy metabolism or on the production of hypoxic conditions related to follicular size. Changes in mitochondrial membrane potential and production of permeability transition pores were the main changes found in small follicles, while lysosomal function destabilization seemed to be the major cause of granulosa cell death during atresia in large follicles.     

Light microscopic observations of alterations in staining of the zona pellucida of porcine follicular oocytes: Possible early indication of atresia

Serial sections of porcine ovaries were examined in an attempt to detect early signs of oocyte degeneration/atresia using special staining. Porcine ovaries were fixed in Bouin's fixative and embedded in paraffin using routine techniques. Serial sections (8 μm) were mounted on glass slides and stained with Shorr's S3 and hematoxylin stain. Several criteria were used for examining general histology of the antral follicles: condition of the granulosa layer, antral cavity, the oocyte and its surrounding zona pellucida, and the cumulus layers. A change in the staining characteristic of the zona pellucida was the single most striking observation in all ovaries examined. In presumably healthy follicles, the zona pellucida was uniformly stained green, the granulosa layer was intact with fewer than three pyknotic cells per section, and a uniform basement membrane (stained green) separated the intact theca layers from the remainder of the follicle. In those follicles showing some degree of degenerative changes in the follicular wall, the zona pellucida was stained a bright orange. In the last stages of degeneration, follicles exhibited many pyknotic nuclei throughout the granulosa layers, the layer of granulosa cells was in many cases separated from the basement membrane, and the antrum was infiltrated with lymphocytes. In these follicles, the zona pellucida was always stained orange. Frequently, the zona pellucida was partially stained orange before any detectable changes could be seen in other elements of the follicular wall. Additionally, many non-antral (primary) follicles exhibited oocytes with orange-stained zonae pellucidae. In terminal stages of follicular degeneration, collapsed follicles were infiltrated by connective tissue elements stained bright orange and green. These structures very often contained dying oocytes always with bright orange-stained zonae pellucidae. Scattered throughout the ovarian stroma were many orange-stained remnants of zonae pellucidae. It is thought that perhaps the characteristic staining of the zona pellucida with Shorr's S3 stain may give an early, previously undetectable indication of follicular atresia.     

Ultrastructure of follicular atresia in the rat

Observations were made on the sequence of morphologic changes in atresia of medium-sized preantral follicles in the rat. Ultrastructural studies indicated that in both control and hormonally treated animals granulosa cell changes, including nuclear condensation and alterations in cytoplasmic organelles, occurred prior to effects on the oocyte. In more advanced stages of atresia, extensive disruption of granulosa cell cytoplasm was associated with loss of microvilli and cytoplasmic vacuolization in oocytes. The findings are consistent with the view that follicular atresia begins with alterations in granulosa cells, effects on the oocyte occurring later in the atretic process.     

Histogenesis of atretic ovarian follicles in a seasonally breeding bird

Histologic examination of ovaries from a non-migratory population of scrub jays (Aphelocoma coerulescens) disclosed a marked annual cycle in the incidence of atresia. Atretic follicles became more common as the nesting season progressed and were most abundant immediately after the cessation of breeding. Atresia involved a dissociation of granulosa cells and movement of these cells into the follicle. Subsequently, granulosa cells showed steatogenesis and ultimately disappeared simultaneously with the invasion of the follicle by ex-thecal gland cells. The data suggest that the diverse histology of avian atretic follicles reflects different stages in the process of atresia rather than multiple origins. Ovarian stromal glands apparently arise both from ex-thecal gland cells of atretic follicles and stromal connective tissue. A possible secretory role of atretic follicles is considered.     

Expression of tPA, LH receptor and inhibin α, β_A subunits during follicular atresia in rats

By using DNA 3'-end labeling, immunocytochemistry and mRNA in situ hybridization detection techniques, the expression of inhibin subunits and LH receptor in the granulosa cells and tissue-type plasminogen activator (tPA) in the oocytes has been studied in relation to follicular development and atresia. The results demonstrated that: (ⅰ) tPA activity in the oocytes of normal developing follicles is undetectable, and increases significantly in the follicle undergoing atresia; (ⅱ) the production of inhibin subunits in granulosa cells is negatively correlated with the expression of oocyte tPA activity, indicating that they may be an important regulator of oocyte tPA production and follicular development; (ⅲ) in atretic follicles, granulosa cells do not express LH receptor and inhibin subunits. It is therefore suggested that tPA may play a role in oocyte self-destruction and clearance in some of atretic follicles, and inhibin of granulosa-origin might be an inhibitory factor for the translation of tPA in     

Expression of tPA, LH receptor and inhibin α, βA subunits during follicular atresia in rats

By using DNA 3′-end labeling, immunocytochemistry and mRNA insitu hybridization detection techniques, the expression of inhibin subunits and LH receptor in the granulosa cells and tissue-type plasminogen activator (tPA) in the oocytes has been studied in relation to follicular development and atresia. The results demonstrated that: (i) tPA activity in the oocytes of normal developing follicles is undetectable, and increases significantly in the follicle undergoing atresia; (ii) the production of inhibin subunits in granulosa cells is negatively correlated with the expression of oocyte tPA activity, indicating that they may be an important regulator of oocyte tPA production and follicular development; (iii) in atretic follicles, granulosa cells do not express LH receptor and inhibin subunits. It is therefore suggested that tPA may play a role in oocyte self-destruction and clearance in some of atretic follicles, and inhibin of granulosa-origin might be an inhibitory factor for the translation of tPA in the oocyte.     

Cholesterol side-chain cleavage cytochrome P450 and 3beta-hydroxysteroid dehydrogenase expression and the concentrations of steroid hormones in the follicular fluids of different phenotypes of healthy and atretic bovine ovarian follicles

Bovine ovarian antral follicles exhibit either one or the other of two patterns of granulosa cell death in atresia. Death can commence either from the antrum and progress toward the basal lamina (antral atresia) or the converse (basal atresia). In basal atresia, the remaining live antrally situated cells appeared to continue maturing. Beyond that, little is known about these distinct patterns of atresia. Healthy (nonatretic) follicles also exhibit either one or the other of two patterns of granulosa cell shape, follicular basal lamina ultrastructure or location of younger cells within the membrana granulosa. To examine these different phenotypes, the expression of the steroidogenic enzymes cholesterol side-chain cleavage cytochrome P450 (SCC) and 3beta-hydroxysteroid dehydrogenase (3beta-HSD) in granulosa cells and concentrations of steroid hormones in follicular fluid were measured in individual histologically classified bovine antral follicles. Healthy follicles first expressed SCC and 3beta-HSD in granulosa cells only when the follicles reached an approximate threshold of 10 mm in diameter. The pattern of expression in antral atretic follicles was the same as healthy follicles. Basal atretic follicles were all <5 mm. In these, the surviving antral granulosa cells expressed SCC and 3beta-HSD. In examining follicles of 3-5 mm, basal atretic follicles were found to have substantially elevated progesterone (P < 0.001) and decreased androstenedione and testosterone compared to healthy and antral atretic follicles. Estradiol was highest in the large healthy follicles, lower in the small healthy follicles, lower still in the antral atretic follicles, and lowest in the basal atretic follicles. Our findings have two major implications. First, the traditional method of identifying atretic follicles by measurement of steroid hormone concentrations may be less valid with small bovine follicles. Second, features of the two forms of follicular atresia are so different as to imply different mechanisms of initiation and regulation.     

Effects of follicular atresia and size on the developmental competence of bovine oocytes: a study using the well-in-drop culture system

The effect of granulosa cell (GC) apoptosis and follicle size on the competence of bovine oocytes were studied using a well-in-drop (WID) oocyte/embryo culture system, which allows identification of follicular origin. Hatching rates of blastocysts did not differ (P>0.05) between oocytes cultured in the WID system (13%) and those cultured in the conventional group system (16%). Hatching rates of blastocysts were higher (P<0.05) in early atretic (17%) than in non-atretic (8%) and late atretic follicles (10%) of the same size (4-8mm), and in 6-8mm (22%) than in 4-5mm follicles (15%) at the early atretic stage. More oocytes (P<0.05) from late atretic (17%) than from non-atreteic (7%) or early atretic follicles (9%) of the same size (4-8mm) were arrested at Grade 1 cumulus expansion (only cells in the peripheral two layers began to expand). Similarly, more oocytes from 2 to 3mm follicles (30%) than from 6 to 8mm follicles (21%) at the same (late) atretic stage had Grade 1 cumulus expansion (P<0.05). Hatching blastocyst percentages of oocytes with Grade 3 (all layers of the cumulus except corona radiate cells expanded) or Grade 4 (full) cumulus expansion were higher in early atretic (20%) than in non-atretic (13%) or late atretic follicles (12%). Hatching blastocyst percentages of oocytes from follicles at the early atretic stage increased as cumulus expanded from Grade 2 (9%) to Grade 4 (27%). Regardless of the degree of follicle atresia, 72-76% of the floating cells in the follicular fluid (FF) were undergoing apoptosis. The floating cell density in FF was highly (r=0.6-0.7) correlated with oocyte developmental potency. In conclusion, the WID culture system was as efficient as group culture and allowed identification of follicular origin. Furthermore, the developmental potential of oocytes was affected by GC apoptosis, follicle size and cumulus expansion, and the floating cell density in FF could be used as a simple and non-invasive marker of oocyte quality.     

Effects of follicular size and FSH on granulosa cell apoptosis and atresia in porcine antral follicles

The purpose of this study was to establish a culture model for isolated intact porcine antral follicles and investigate the relationship between granulosa cell apoptosis and follicular atresia. Small (<3 mm), medium (3–5 mm) and large (>5 mm) healthy porcine follicles were isolated and cultured in serum‐free TCM199 with or without follicular stimulating hormone (FSH). Microscopic identification of healthy follicles was confirmed by histology. A spontaneous onset of apoptotic cell death in granulosa cells was observed from cultured antral follicles. The apoptotic rate of granulosa cells from small follicles cultured for 24 hr was higher than those of large and medium follicles, accompanied with high FasL mRNA abundance in granulosa cells. Supplementation with 3 or 5 IU/ml FSH significantly inhibited the percentage of granulosa cells that became apoptotic. FSH did not significantly alter estradiol secretion from cultured follicles. Progesterone secretion significantly decreased after culture for 48 hr, coinciding with the morphological changes observed. FasL and Fas mRNA were expressed in the healthy, early atretic, and progressed atretic porcine follicles regardless of follicular size. However, FasL but not Fas mRNA levels increased during follicular atresia. Addition of FSH significantly decreased FasL rather than Fas mRNA levels in granulosa cells and could attenuate apoptosis. Small follicles seemed to be more susceptible to atresia as compared to medium and large follicles. Mol. Reprod. Dev. 77: 670–678, 2010. © 2010 Wiley‐Liss, Inc.     

Expression of tPA, LH receptor and inhibin a,βA subunits during follicular atresia in rats

By using DNA 3′-end labeling, immunocytochemistry and mRNA insitu hybridization detection techniques, the expression of inhibin subunits and LH receptor in the granulosa cells and tissue-type plasminogen activator (tPA) in the oocytes has been studied in relation to follicular development and atresia. The results demonstrated that: (i) tPA activity in the oocytes of normal developing follicles is undetectable, and increases significantly in the follicle undergoing atresia; (ii) the production of inhibin subunits in granulosa cells is negatively correlated with the expression of oocyte tPA activity, indicating that they may be an important regulator of oocyte tPA production and follicular development; (iii) in atretic follicles, granulosa cells do not express LH receptor and inhibin subunits. It is therefore suggested that tPA may play a role in oocyte self-destruction and clearance in some of atretic follicles, and inhibin of granulosa-origin might be an inhibitory factor for the translation of tPA in the oocyte. Project supported by the National Natural Science Foundation of China (Grant Nos. 39770099 and 39770290), Chinese Academy of Sciences and Rockefeller Foundation/WHO HPR research project.     

Size-frequency analysis of atresia in cycling rats

The purpose of this study was to delineate when, during follicular growth, the alternative developmental pathways leading to ovulation or atresia diverge. By using computerized image analysis techniques, random samples of healthy and atretic follicles in ovaries of cycling rats were subjected to size-frequency analysis. The vast preponderance of atretic follicles were of the early antral size class (approximately 300-350 micron diameter, 800-1000 granulosa cells in the largest cross-section); atretic small follicles (less than 250 granulosa cells in the largest cross-section) were rare. Follicles in early stages of atresia were uncommon in ovaries of animals killed at estrus, but were found with great frequency in ovaries of animals killed the following day (metestrus). These results suggest that, under normal cyclic conditions, there may be only one major branching point during follicular development when growing follicles become susceptible to atresia. The alternative developmental pathways leading to ovulation and atresia may not diverge until the penultimate stage of growth, immediately preceding the final transformation into a preovulatory follicle.     

Steroidogenic activity of atretic follicles in the cycling hamster ovary and relation to ultrastructural observations

To evaluate the relation between the steroidogenic activity and cell proliferation of individual follicles in mature hamster ovaries during the estrous cycle, the localization of enzymes involved in estrogen biosynthesis and bromodeoxyuridine (BrdU) incorporation were examined immunohistochemically. Moreover, granulosa cells from the early atretic follicle were examined by scanning and transmission electron microscopy. Immunoreactivity for aromatase was localized in the granulosa cells of healthy developing follicles and Graafian follicles, as well as in newly formed granulosa lutein cells. In the healthy follicles of an ovulation cycle, intensity of aromatase immunoreactivity was suddenly decreased on day 3. The theca interna cells of healthy developing follicles were immunopositive for 17-hydroxylase/C17–C20 lyase (17-lyase) from day 2 to the morning of day 4, but on the evening of day 4 most theca interna cells were immunonegative except for only a few cells of the large Graafian follicles. BrdU incorporation was observed in the granulosa cells of healthy developing follicles, in the endothelial cells of capillaries around the developing follicles, and of newly formed corpora lutea. Very early morphological signs of atresia was the pyknotic change of a few granulosa cells lining the antral cavity. In that follicle, the number of BrdU-incorporating granulosa cells was suddenly decreased whilst immunoreactivity of aromatase and 17-lyase were gradually decreased. These data suggest that the mechanism of the loss of aromatase activity from the granulosa cells of atretic follicles appears to differ from that in cycling follicles. Even in the early stage of atresia, the granulosa cells showed remarkable morphological change characteristic of apoptosis, as visualized by scanning and transmission electron microscopy. Cessation of granulosa cell proliferation may occur earlier than apoptotic change and the degeneration of the granulosa cells becomes rapid once atresia starts.     

Multiparametric study of atresia in ewe antral follicles: histology, flow cytometry, internucleosomal DNA fragmentation, and lysosomal enzyme activities in granulosa cells and follicular fluid

The differential quantitative participation of apoptosis and necrosis in ewe antral follicles of two different sizes, separated in four stages of atresia using macroscopic, histologic, and esteroid quantification methods was assessed. Annexin V binding and propidium iodide (PI) uptake was used to detect healthy live cells (Annexin V negative/PI negative), early apoptotic cells (Annexin V+/PI-), and necrotic or late apoptotic cells (PI+). Additionally we used internucleosomal DNA fragmentation as a quantitative estimate of apoptosis. Presence and distribution of lysosomal enzymes in follicular fluid and granulosa cells was used as a measure of necrotic cell death. DNA flow cytometry and gel electrophoresis were positively correlated with the progression of atresia, small atretic follicles tend to have higher percentages of internucleosomal cleaved DNA than follicles >6 mm. Annexin/PI binding also indicates that apoptosis and necrosis increase with atresia progression, generally apoptosis outweighs necrosis in small follicles. Acid phosphatase and glucosaminidase in follicular fluid of 3-6 mm follicles showed no significant modifications between healthy and initially atretic follicles, and only a small, but significant increase in activity in advancedly atretic follicles. On the contrary, lysosomal enzyme activity in follicles >6 mm showed positive correlation between atresia stages and the activities of acid phosphatase and glucosaminidase in follicular fluid. A similar size-differential behavior was found in free or membrane-bound lysosomal enzyme activity of granulosa cells. Necrosis, but principally apoptosis, were present during all stages of follicular maturation indicating that growth and maturation of ovarian follicles involves a continuous renewal of granulosa cells, regulated by apoptosis. Mechanisms regulating this equilibrium may participate in the final destiny, whether ovulation or atresia of ovarian follicles.     

Histological and histochemical studies of post-ovulatory and pre-ovultory atretic follicles in Mustelus canis

The atresia of post-ovulatory and pre-ovulatory follicles of the viviparous smooth dogfish, Mustelus canis, is compared for approximately the first fourth of an 11 month gestation. A thick collagenous sheath and numerous tubules in the theca identify the large, folded stage A post-ovulatory follicle. In stage B the tubules have been filled by cells to form “islands.” In stage C the entire structure is greatly diminished, adjacent islands tend to fuse, the collagenous sheath is virtually gone and the granulosa is degenerating. Preovulatory follicles from large, yolky oocytes pass through four stages beginning with yolk phagocytosis by granulosa cells of the villi (stage I), which are long and granular in stage II; villi fuse, theca cells increase greatly, fill with granules (stage III), encroach on the granulosa and disperse it into small groups of cells which finally disappear (stage IV) leaving a mass of thecal cells. A special type of pre-ovulatory follicle from small non-yolky oocyte atresia exhibits prominent thecal tubules and an unusual arrangement of granulosa cells. This follicle appearrs to enlarge during the summer, becoming multilobed; few granules are present. The distribution of lipid in frozen sections, stained by Oil red O, is described for all types of follicles. Schultz and Lewis and Lobban tests for steroids were made on frozen sections with corresponding results. Positive green tests indicating the presence of steroids or possible steroidogenesis were limited to: (1) one post-ovulatory follicle, in the islands; (2) four stage III and seven late stage IV pre-ovulatory yolky atretic follicles; (3) two special atretic follicles. The special atretic follicle appears to be a unique feature of M. canis and it is suggested tentatively that it may be related to viviparity.     

Aspects of cellular proliferation during follicular atresia in the dog ovary

Summary Autoradiography after pulse labelling with [³H] thymidine was applied to investigate the proliferation processes in the granulosa and theca related to follicular atresia of the dog ovary during metestrus.The number of proliferating cells depends on the follicle type and its atretic stage. There is less proliferation in smaller secondary follicles than either in larger ones or tertiary follicles. While in early atresia tertiary follicles show the highest labelling indices, in advanced atresia the larger secondary follicles are those with the highest values. For each follicle type a decline in the labelling indices can be observed from early to terminal atresia. Tertiary follicles show a precipitous decrease in the labelling index between early and advanced atresia. There is a continuous gradient of proliferation from the center of the follicle over the peripheral granulosa to the theca. In tertiary follicles, an inverse correlation between labelling and necrosis of granulosa cells can be observed.     

An enzymatic method for dissociation of intact follicles from the hamster ovary: histological and quantitative aspects

An enzymatic method was developed to collect intact follicles at different stages of development from cyclic hamsters to study ovarian folliculogenesis under various circumstances. Ovaries from 6 adult hamsters on each day of the cycle (Day 1 = ovulation) were collected, corpora lutea and large preantral and antral follicles were dissected, and follicles saved. Minced ovaries were then incubated with a mixture of collagenase, DNAse and pronase at 37 degrees C for 20 min to disperse intact follicles. Histological studies with 2191 isolated follicles revealed 10 different stages of follicular development (depending on the number of granulosa cell layers surrounding the oocyte and development of the antrum). Of the total follicular population, 14% showed signs of atresia, with 50% of those having 1-3 layers of granulosa cells (Stages 1-3); a second peak of 18% was observed in antral follicles (Stages 8-10). No signs of thecal cells were evident until the follicles reached Stage 6 (7-8 layers of granulosa cells), which possibly accounts for reduced atresia in this class and beyond. Ultrastructural study revealed that there were no signs of morphological damage to the basement membrane or to other subcellular organelles in the small preantral follicles. The presence of subnuclear lipid droplets in follicles with 3 layers of granulosa cells provided evidence for potential steroidogenesis by small follicles. The number of Stage 1-10 follicles was remarkably constant throughout the estrous cycle (460 +/- 34 per animal on Day 1 vs. 492 +/- 66 on Day 4). The usefulness of this method in analyzing follicular kinetics is illustrated in experiments involving hypophysectomy and the effects of unilateral ovariectomy. This procedure offers an improved method to study the factors responsible for the growth and the differentiation of small preantral follicles in the mammalian ovary.     

Changes in expression of anti-apoptotic protein, cFLIP, in granulosa cells during follicular atresia in porcine ovaries

Follicular selection is performed in mammalian ovaries, as most follicles undergo atresia during follicular development and growth. Follicular regression is indicated to begin with granulosa cell apoptosis. To reveal the molecular mechanisms of the selection, we examined the changes in the levels of cellular-Flice like inhibitory protein (cFLIP) expression in porcine granulosa cells. cFLIP is the homologue of intracellular apoptosis inducer (procaspase-8/Flice), and has two alternative splicing isoforms: cFLIP short form (cFLIP(S)) and long form (cFLIP(L)). By competing with caspase-8, cFLIP inhibits apoptosis initiated by death receptors. The changes in the levels of cFLIP(S) and cFLIP(L) mRNA and protein expression in granulosa cells were determined by RT-PCR and Western blotting, respectively. cFLIP(L) mRNA and protein were highly expressed in granulosa cells of healthy follicles and decreased during atresia. cFLIP(S) mRNA levels in granulosa cells were low and showed no change among the stages of follicular development, and its protein level was extremely low. We examined the changes in the localization of cFLIP mRNAs in pig ovaries by in situ hybridization and found that cFLIP(L) is abundant in granulosa cells of healthy follicles in comparison with those of atretic follicles. Immunohistochemical analyses demonstrated that the cFLIP protein is highly expressed in the granulosa cell of healthy follicles but weakly expressed in that of atretic follicles. We presumed that cFLIP, especially cFLIP(L), plays an anti-apoptotic role in the granulosa cells of healthy follicles of pig ovaries, and that cFLIP could be a major survival factor that determines whether growth or atresia occurs in porcine follicles.     

Histological study of follicular atresia in the ovary of the domestic hen (Gallus domesticus)

Semi-serial (1 in 20) sections of ovaries were studied and only two types of atresia were identified--non-bursting and bursting. Smaller, non-yolky follicles (less than 1 mm diameter) showed non-bursting atresia. Atresia in follicles greater than 1 mm diameter was invariably of the bursting type which involved the rupture of the follicular wall, and the extrusion of yolk and cellular debris through the rupture site into the stroma. However, this rupture site was small and consequently was not visible in every section but it could always be seen when the follicle was followed in semi-serial sections. The mitotic index of granulosa cells in bursting atretic follicles was much lower than that for normal follicles. The most common criteria for distinguishing non-bursting atretic follicles were the extremely shrunken, irregularly shaped oocytes and the separation of the granulosa from the theca. In bursting atretic follicles, reliable indications were the presence in the ooplasm of some cells or cellular debris, and disorganization of the yolk and granulosa tissue. The presence of pycnotic nuclei in the granulosa cells was not a consistent feature of all atretic follicles of the hen.