Size-Frequency Analysis of Atresia in Cycling Rats

Transcription

1 BOLOGY OF REPRODUCTON 38, (1988) Size-Frequency Analysis of Atresia in Cycling Rats ANNE N. HRSHFELD Department of Anatomy University of Maryland School of Medicine Baltimore, Maryland 2121 ABSTRACT 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 3-35 m diameter, 8-1 granulosa cells in the largest cross-section); atretic small follicles (<25 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. NTRODUCTON Degenerating follicles are a prominent feature in ovaries of adult cycling rats. Follicles are believed to degenerate (undergo atresia) because their developmental stage is out of synchrony with the hormonal events of the estrous cycle. Follicles which degenerate may have failed to develop adequate receptors for specific hormones (Richards et a!., 1978), they may have been deprived of critical hormonal support at a crucial stage of development (Hirshfield, 1986b), or they may have been exposed to deleterious hormones (atretogenic factors) at sensitive stages of development (Ryan, 1981). To understand follicular atresia, we must correlate specific stages of follicular growth with the hormonal requirements necessary to sustain viability at each stage. A first step is to determine at which stages of fo!!icu!ar development follicles are most likely to undergo atresia under norma! physiological conditions. Once these vulnerable stages of growth have been identified, hormonal conditions can be manipulated to identify the specific factors responsible for follicular demise. Accepted February 8, Received August 6, Supported in part by NH Grant HD13642 and Research Career Development Award HD593. Two alternative hypotheses concerning the vulnerable stages of follicular growth are prevalent in the literature. Some authors (Eng!e, 1927; Knigge and Leathern, 1956; Byskov, 1974, 1979; Ryan, 1981) maintain that atresia can occur at any stage of deve!- opment, suggesting that those follicles which ultimately ovulate have escaped atresia many times during the course of growth. Others maintain that the phenomenon of atresia is limited to a particular stage of development (Richards and Midg!ey, 1976; Richards et al., 1978) and that there is a single period of vulnerability during the growth process. Quantitative empirical data pertinent to these hypotheses are scarce. Many size-frequency studies of fo!!icu!ar development ignore atretic follicles entirely (Welschen and Rutte, 1971; Hage et al., 1978; Hirshfield and Midg!ey, 1978; Hoak and Schwartz, 198; Lintern-Moore et al., 1981). Of the few studies that quantitate atretic follicles, most report only on atresia in antra! follicles (Mandi and Zuckerman, 1952; Braw and Tsafriri, 198; Taya and Sassamoto, 1981; Osman, 1985, Hirshfie!d, 1986b); others fail to categorize atretic follicles by size, indicating only the overall incidence of atresia of all sizes taken together (Byskov, 1974; Kohut et al., 1985). Only a handful of studies provide more detailed data on the size distribution of atretic follicles. Some of these studies indicate that atresia occurs with considerable 1181 on 8 February 218

2 1182 HRSHFELD frequency in follicles of all size classes (Mandl and Zuckerman, 1951, Gosden et a!., 1983), while others report negligible atresia of small follicles (Faddy et al., 1976). The purpose of this study was to determine if the course of fo!!icular development in cycling rats contains multiple branching points when follicles are likely to undergo atresia, or if atresia is confined to a unique stage of growth. My results suggest that, in the cycling rat, atresia is restricted to a discrete developmental period that occurs near the final stages of follicu!ar growth. MATERALS AND METHODS Animals and Preparation of Tissues Sprague-Daw!ey-derived, young adult (15-16 g) female rats were obtained from Dominion Laboratories (Dublin, VA). Vaginal smears were taken daily to determine estrous cyc!icity. Rats, hypophysectomized at 21 days of age, were obtained from Zivic- Miller Laboratories (A!ison Park, PA). Al! rats were housed under light- and temperature-controlled conditions (12L:12D; lights on at 7 h) and fed Purina Rat Chow (Ralston-Purina Co., St. Louis, MO) and water ad!ibitum. Rats were killed by an overdose of ether. Hypophysectomized rats were killed 1 days post surgery. Adult rats were killed on each day of the estrous cycle. Ovaries were removed, fixed in Kah!e s fixative (Gurr, 1956), embedded in Paraplast (VWR Scientific, Baltimore, MD), serially sectioned at 7 m, then stained with Weigert s Hematoxy!in followed by Picric Acid-Methy!ene Blue counterstain (Luna, 1968). Morphometry Sample selection. To determine the size-frequency distribution of atretic follicles, 2 sections of ovaries from 32 adult rats (randomly chosen without regard to the day of the estrous cycle) were scanned at 1 x magnification. Al! sections of atretic follicles (see below) containing a portion of a fully grown oocyte were located and subjected to detailed morphometric analysis at 25 X magnification. To determine the size-frequency distribution of healthy follicles in cycling rats, 1 random sections from ovaries of 7 cycling rats were scanned, and a!! healthy follicles with a portion of the oocyte included in the section were selected for morphometric analysis. To determine the effect of the day of the estrous cycle on the size-frequency distribution of atretic follicles, a systematic stratified sampling procedure was used (Williams, 1977). This method of sampling was necessary because of the nonrandom distribution of follicles within the ovary (small follicles are concentrated in the medullary region of the follicle; unpublished observations). The sample consisted of 18 tissue sections: 3 ovarian sections were randomly chosen from the middle three slides of 5 ovaries from each of the 4 days of the estrous cycle. All atretic follicles displaying a portion of the oocyte were measured by image analysis as described below. To determine if pyknotic granulosa cells are adequate markers of atresia in small follicles, random sections of ovaries from hypophysectomized immature rats and from infant rats were scanned. Small follicles containing pyknotic granulosa cells were located and their granu!osa cells counted. mage Analysis A video-based image analysis system (mage Technology Corporation, Deer Park, NY) was used for semi-automated quantitative analysis of fol!icular size. This instrument is equipped with a digitizer pad, which permits manual input, and with Boo!ean operators, which perform logical functions between its two planes of memory. Measurements were performed at a magnification of 16 X with a green filter in the light path that intersifies the contrast between stained tissue and spaces within tissues (antrum, intercellular space). The feature-specific mode of the image analyzer was used to compute the perimeter and diameter of the follicular cross-sections and area of the granulosa layer as follows. The grey level discriminator was adjusted for each follicle so that all solid tissue was selected but oocytes (which stain lightly) and antral spaces were excluded. The enhanced image was then saved in one memory plane. The follicle was delineated from the surrounding tissue by tracing around the basement membrane of the projected image using the mouse and a digitizer pad. The computer first measured the area within the tracing of the follicle, as well as its circumference, longest diameter, and shortest diameter. Then, using Boo!ean logic, the computer measured the area of the enhanced image (stored in the first memory plane) on 8 February 218

3 ATRESA N CYCLNG RATS 1183 within the tracing of the circumference of the follicle (stored in the second memory plane). This is equivalent to the area of the granulosa layer. For each follicle in the study, therefore, the following data were collected: subjective classification of atresia, overall area of the fo!licu!ar crosssection, granulosa cell area, circumference of the follicle, and average follicular diameter. Estimation of the Number of Granulosa Cells in Follicles The granu!osa cell area was used to estimate the number of granulosa cells in the follicular crosssection. This estimate was obtained by multiplying the granu!osa cell area (in i.im2) by the constant (the average area of a single granulosa cell). The fol!icular diameter was used to calculate the original number of granulosa cells that should have been present in atretic follicles before cell loss began to occur. This estimate was obtained by using the following equation: Lm(y) = ( ) + [(1.8647)(Ln x)] where y = number of granulosa cells, and x = follicular diameter in jim (Hirshfield and Midgley, 1978). RESU Validation of Sampling Procedure LTS Quantifiable parameters such as fo!!icu!ar diameter or the number of granulosa cells in the cross-section, provide the most precise descriptions of follicle size. These measurements are usually taken at what is assumed to be the LCS; peripheral sections through the follicle would be considerable smaller than those taken from the central portion. n previous studies, the nucleolus of the oocyte was used as a marker for the LCS of follicles (Hirshfie!d and Midgley, 1978); every section of serially sectioned ovaries was scanned to locate those sections containing nucleoli. n this study, two constraints prevented the use of the nucleolus as a criterion for choosing sections of follicles to measure: 1) atretic oocytes often lack nucleoli, and 2) random sampling was used (when every section is not scanned, the likelihood of encountering the oocyte nucleolus of any given follicle is greatly reduced). For these reasons, the selection criterion for choosing the largest cross-section was broadened to include any section containing a portion of the oocyte. t was therefore necessary to determine if the oocyte was an acceptable marker of the LCS-that is, if the follicular sections chosen by this method were Criteria for Distinguishing A tretic Follicles A medium or large follicle was considered atretic if it had more than 3 pyknotic granulosa cells in the largest cross-section (LCS). A small follicle (less than 5 granulosa cells in the LCS) was considered atretic if it had more than 1 pyknotic granulosa cell in the LCS. Follicles that met these minimum criteria were further subdivided subjectively into the following categories: incipient atresia (appeared healthy under low magnification; pyknotic granu!osa cells only discovered after close examination under higher magnification); obvious atresia, (clearly atretic under low magnification, usually a large number of pyknotic granu!osa cells, overall structure of follicle still intact); advanced atresia (thinning or disruption of the cumulus oophorus, marked loss of granulosa cells, number of pyknotic bodies decreasing; endstage atresia (fragments of oocyte floating free within a cavity, true granulosa and theca cells no longer identifiable). w z w cn ( z SAMPLE NUMBER FG. 1. Location of oocyte and oocyte nucleolus in complete serial sections of follicles. Each of the 5 lines on the graph represents the serial sections from a single follicle. Follicular size is expressed on the vertical axis as the area (in tm2) occupied by granulosa cells. For each follicle, the first histological serial section containing granulosa cells was considered Section 1. Filled triangles represent those sections that contained portions of the oocyte. Open triangles represent those sections that contained the oocyte nucleolus. on 8 February 218

4 1184 HRSHFELD close enough to the center of the follicle to give an accurate indication of follicular size. For this purpose, the granu!osa cell area of every serial section from 5 healthy antra! follicles was measured and the locations of oocyte and oocyte nucleolus were noted. There was no significant difference in granulosa cell area between sections containing the nucleolus of the oocyte versus sections containing other portions of the oocyte (Fig. 1), suggesting that any portion of the oocyte is as accurate an indicator of the LCS as the nucleolus of the oocyte. n 4 of the 5 follicles examined, both markers were closely correlated with the largest area of the granu!osa layer in the follicle. However, in one very large follicle (extending through 52 serial sections), the greatest granulosa cell areas were found in peripheral sections; antral fluid occupied most of the area in cross-sections through the center of the follicle. Classification of A tretic Follicles by mage Analysis Qualitative indicators of atresia were closely correlated with quantifiable loss of granulosa cells. Follicles judged subjectively to be in incipient atresia had approximately the same number of granulosa cells in the LCS as healthy follicles of similar size (circumference used as a measure of size). However, follicles judged to be in obvious atresia had fewer granulosa cells in the largest cross-section than healthy follicles of equivalent size (circumference). n healthy follicles, estimates of the number of granulosa cells based on the granulosa cell area were highly correlated with estimates based on the diameter of the follicle (correlation coefficient 89% by regression analysis). The ratio of the number of granu!osa cells actually present in an atretic follicle (based on the granulosa cell area) versus the number that should have been present before cell loss began (based on the follicular diameter) will henceforth be referred to as the thinning ratio. This ratio represents the proportion of the granulosa layer remaining intact. The thinning ratio was closely correlated with subjective classification of the degree of atresia. The more atretic a follicle was judged to be, the lower the thinning ratio (Fig. 2). Follicles in incipient atresia had thinning ratios of.9 or higher. Follicles in obvious atresia had ratios of approximately.7 to.9; most follicles in advanced atresia had ratios of between.5 and.7. Follicles in endstage atresia had thinning ratios of less than.5. Size/Frequency of A tretic Follicles A Analysis Small healthy follicles greatly outnumbered large healthy follicles in the adult rat ovary (Fig. 3). The greatest number of follicles seen was in the smallest size class examined (approximately 5-1 cells in the largest cross-section, 3 jim circumference, 1 jim diameter). There were more than 4 times as many follicles with a circumference of approximately 45 jim in the sample than follicles with a circumference of 15 pm. The frequency distribution of atretic follicles in the adult presented a striking contrast to that of healthy follicles (Fig. 3). The vast preponderance of atretic follicles were between 8 and 12 pm in circumference, with highest concentrations at 1 pm circumference. Atretic smaller follicles (4-7 pm circumference) were extremely rare. The overall number of atretic follicles (all categories of atresia) per stratified sample did not vary significantly with the day of the estrous cycle. There were an average of 11.7 follicles in all stages of atresia per nine-section sample (Fig. 4a). n contrast, the overall number of follicles in incipient atresia (>8% intact granulosa layer) did U #{149} U #{149} U #{149} OD.n #{149}.. S. U.U #{149} U (Thousands) CRCUMFERENCE (sm) FG. 2. Thinning ratio (a quantifiable indicator of atresia) correlated with a subjective assessment of atresia. With an algorithm based upon data from healthy follicles, the diameter of the largest crosssection (LCS) of a follicle was used to predict the number of granulosa cells that should have been found in the same LCS. The thinning ratio reflects the degree of disparity between the predicted number of granulosa cells and the number actually seen (y axis). Follicular size is represented on the x axis in terms of circumference of the LCS. Each marker represents one follicle whose stage of atresia was scored subjectively as follows: incipient atresia (.), obvious atresia (n), advanced atresia (s),endstage atresia (s). on 8 February 218

5 ATRESA N CYCLNG RATS 1185 u,3 w U- o U 4 - Ui 41 -i-i V% CRCUMFERENCE (THOUSANDS) (jlm) FG. 3. Size-frequency distribution of healthy and atretic follicles in ovaries of adult cycling rats. Size is expressed as the circumference of the follicle. Only those atretic follicles with a thinning ratio of.7 or greater (see Results) were included in this illustration. Under normal cyclic conditions, most small growing follicles seem to be relatively impervious to atresia. Atresia of small follicles appears to be rare in the cycling rat. The possibility exists that small follicles of cycling rats do undergo atresia but that the symptoms may differ significantly from those seen in larger follicles. Pyknotic granulosa cells may not be a sensitive indicator of atresia in smaller follicles. Nevertheless, pyknotic granulosa cells are common in small follicles of hypophysectomized rats and infant rats, suggesting that this symptom of atresia is indeed manifested in small as well as large follicles. Engle (1927) and Byskov (1978) have stated that degeneration of the oocyte is the first sign of atresia in very small follicles. While frequently saw this symptom of atresia vary significantly with the days of the estrous cycle (Fig. 4b). There were only.6 follicles in incipient atresia per nine-section sample in the estrous ovaries, but by the following day (metestrus) there were approximately 6 follicles per sample. The size distribution of follicles in incipient atresia varied according to the estrous cycle as well (Fig. 5). Atretic ninth generation follicles (>1 granulosa cells originally in the LCS; Hirshfield and Schmidt, 1987) were absent in ovaries of animals killed on proestrus and estrus, but present in ovaries of rats killed on metestrus and diestrus. n contrast to adult cycling rats, ovaries of hypophysectomized immature rats and infant rats had many small atretic follicles (with <1 granulosa cells in the LCS). These follicles were readily recognized by the presence of apoptotic bodies in the granulosa layer (Figs. 6 and 7), as we!! as by degenerative changes in the oocyte. Although there were many primordial and very small primary follicles (<2 granulosa cells in the LCS) in these ovaries, no atretic follicles were identified in this size cohort. in ovaries of prepubertal rats and in hypophysectow 4 () ii a- Cl, ii C.) U- C.) - ii - 4 DSCUSSON These observations suggest that the developmental pathways leading to ovulation and atresia most often do not diverge until the penultimate stage of follicu!ar growth. n the cycling rat, there may be only one major branching point during follicu!ar development when growing follicles become susceptible to atresia. DAY OF CYCLE FG. 4. The average number of atretic follicles per stratified random sample of ovaries from each day of the estrous cycle. Follicles combined from all categories of atresia (incipient through endstage) are shown in Panel a, and only follicles in incipient atresia are shown in Panel b. Each bar represents the mean ± SEM of 5 rats. on 8 February 218

6 1186 HRSHFELD C/) Ui C-) U- U- Ui z 8 4. * Estrus * ts S / : *,, S Proestrus 8 / \,lt : : 4 J;.: i/..\, 2 5, ( number of granuloso cells Generation S t licles with a circumference of about 1 pm have an average diameter of 32 pm, and slightly more than 8 granulosa cells in their largest cross-section. Thus, this size is equivalent to the small antral or early antral stage of follicular development (Hirshfield, 1982). Therefore, atresia at this early antral stage of development accounts for the vast preponderance of follicular attrition in the cycling rat. Previous studies have suggested a mechanism and rationale for atresia at the early antral stage of follicular development (Schwartz, 1974; Welschen and Dullaart, 1976; Hoak and Schwartz, 198; DePaolo et al., 1981; Hirshfield, 1986a,b). Atresia during this stage of development appears to be a component of the regulatory process that controls the number of follicles selected for ovulation. The early antral follicle consists of cells approaching the final stages of differentiation. At this stage, the proliferative potential of the granulosa cells has already become nearly exhausted, and both granulosa and theca are poised to undergo functional maturation (Hirshfield and Schmidt, 1987). Only one more round of cell division is required to bring these follicles to full preovulatory size. However, granulosa cells of most early antral follicles apparently encoun- FG. 5. Size-frequency distribution of follicles in early atresia as a function of the day of the estrous cycle. These data were obtained by the stratified random sampling procedure (see Materials and Methods). The dotted line (repeated in each panel) represents the overall size distribution of all atretic follicles in the entire sample with >8% intact granulosa layer. Follicles from all rats killed on a particular day of the estrous cycle were pooled to give the size distribution for that day (solid lines). Two descriptions of follicular size are given on the horizontal axes: estimated original number of granulosa cells in the largest cross-section (based on follicular diameter; log scale), and granulosa cell generations (a primordial follicles = first generation; a preovulatory follicle = 1th generation; see -irshfield and Schmidt, 1987). mized rats, seldom saw small follicles with degenerating oocytes in ovaries of adult cycling rats. Byskov also states that in the mouse, as well, necrotic oocytes are numerous during the first days after birth but are rare in later life. Moreover, my results concur with those of Faddy et al. (1976, 1983), whose model of ovarian follicular dynamics demonstrated negligible death rates in follicles with 1-3 layers of granulosa cells in several strains of adult mice. n the present study, most atretic follicles were in the 1-pm circumference size class. Healthy fol- FG. 6. Small, one-layer atretic follicle of hypophysectomized, untreated, immature rat. This follicle has 29 granulosa cells in the largest cross-section (LCS), one of which is obvious pyknotic. Note that the oocyte has grown to full size. Follicles with fully grown oocytes and only a single layer of granulosa cells are characteristic of ovaries of hypophysectomized rats, but are rare in intact immature or adult rats (X 4). on 8 February 218

7 ATRESA N CYCLNG RATS 1187 FG. 7. Small, two-layer atretic follicle of hypophysectomized, untreated immature rat. The follicle on the right has 53 granulosa cells in the largest cross-section, 2 of which are pyknotic. The follicle on the left appears to be healthy (X 4). ter some sort of metabolic roadblock that causes them to die before this final round of cell division can take place. High levels of follicle-stimulating hormone (FSH) appear to be able to sustain early antral follicles, permitting their granulosa cells to complete the last round of cell division and the final steps of differentiation. This allows them to acquire the potential to luteinize in response to luteinizing hormone (LH). Negative feedback from maturing antral follicles apparently causes the high postovulatory concentrations of FSH to fall on the morning of estrus (DePaolo et al., 1981). As the number of follicles reaching a critical maturity increases, the magnitude of this negative feedback signal must increase as well. t is by this classic global feedback regulatory mechanism that the appropriate number of follicles is probably selected for ovulation (Lacker et al., 1987). The wave of atresia that Engle (1927) and have observed at metestrus coincides temporally with the selection of the appropriate number of follicles for ovulation, which is completed by the afternoon of estrus. The necrotic changes that signal atresia at metestrus are probably a consequence of cell death which must have occurred many hours earlier (Trump and Arstila, 1975). f the global feedback loop is broken, for example, by prolonging the endogenous FSH surge with a low dose of pregnant mare s serum gonadotropin, the demise of younger early antral follicles can be prevented, and a supernormal number of large healthy follicles will be seen at metestrus (Hirshfield, 1986b). Demise of young antral follicles can also be prevented by hemicastration. By removing one ovary, approximately half of the follicles already selected for ovulation are lost. n response to this loss, there is a compensatory rise of FSH (Welschen and Dullaart 1974; Welschen et al., 1978) that recruits additional follicles in the remaining ovary for ovulation (Peppler and Greenwald, 197). These additional follicles are probably those that had just reached the metabolic roadblock at the early antral stage of growth and would have died if FSH levels had remained at baseline. t is likely that growing follicles are distributed along a continous maturation gradient and that no two follicles are precisely synchronized with each other with regard to their stage of growth. Therefore, follicles are probably arriving continuously at the roadblock of development and undergoing atresia. Elevated FSH following hemicastration apparently sustains the viability of some of these follicles; this allows them to complete the final round of cell proliferation and undergo final maturation. Within 36 h, the newly recruited follicles mature sufficiently to ovulate (Peppler and Greenwald, 197; Otani and Sasamoto, 1982). n summary, the data reported here indicate that in the cycling rat, follicles are relatively impervious to atresia until they reach the size when functional differentiation of granulosa and theca cells would have occurred (Hirshfield, 1985). t is tempting to postulate a causal link between cell viability and cell differentiation: granulosa cells must differentiate at this early antral size to remain viable. Differentiation is clearly triggered by high concentrations of FSH (Zeleznick et al., 1974; Mueller et al, 1978; Daniel and Armstrong, 1984; Zlotkin et al., 1986). Those follicles that arrive at the early antral stage of development when FSH levels are high are able to complete the final round of cell division and the dramatic phenotypic differentiation that transfroms them into preovulatory follicles. REFERENCES Braw RH, Tsafriri A, 198. Effect of pregnant mare s serum gonadotropin on follicular atresia in the immature rat ovary. J Reprod Fertil 59: Byskov AGS, Cell kinetic studies of follicular atresia in the mouse ovary. J Reprod Fertil 37: Byskov AGS, Follicular Atresia. n: Jones RE (ed.), The Verte- on 8 February 218

8 1188 HRSHFELD brate Ovary. New York: Plenum Press, pp Byskov AGS, Atresia. n: Midgley AR Jr, Sadler WA (eds.), Ovarian Follicular Development and Function. New York: Raven Press, pp Daniel SAJ, Armstrong DT, Site of action of androgens on FSH-induced aromatase activity in cultured rat granulosa cells. Endocrinology 114: DePaolo LV, Anderson LD, Hirshfield AN, Possible existence of a long loop feedback system between FSH and inhibin in female rats. Am J Physiol 24:E Engle ET, A quantitative study of follicular atresia in the mouse. Am J Anat 39: Faddy MJ, Jones EC, Edwards RG, An analytical model for ovarian follicle dynamics. J Exp Zoo! 197: Faddy MJ, Gosden RG, Edwards RG, Ovarian follicle dynamics in mice: a comparative study of three inbred strains and an Fl hybrid. J Endocrinol 96:23-33 Gosden RG, Laing SC, Felicio LS, Nelson JF, Finch CE, mminent oocyte exhaustion and reduced follicular recruitment mark the transition of acyclicity in aging C57BL/6J mice. Biol Reprod 28:255-6 Gurr E, A Practical Manual of Medical and Biological Staining Techniques. London: Leonard Hill Books Ltd Hage AJ, GroenKlevant AG, Welschen R, Follicle growth in the immature rat ovary. Acts Endocrinol 88: Hirshfield AN, Follicular recruitment in long-term hemicastrate rats. Biol Reprod 27:48-53 Hirshfield AN, Patterns of cell proliferation in follicles approaching ovulation. n: Toft DO, Ryan RJ (eds.), Proceedings of the Fifth Ovarian Workshop. Champaign, L: Ovarian Workshops, nc. pp Hirshfield AN, 1986a. Patterns of j3 H] thymidine incorporation differ in immature rats and mature, cycling rats. Biol Reprod 34: Hirshfield AN, 1986b. Effect of a low dose of pregnant mare s serum gonadotropin on follicular recruitment and atresia in cycling rats. Biol Reprod 35: Hirshfield AN, Midgley AR, Morphometric analysis of follicular development in the rat. Biol Reprod 19: Hirshfield AN, Schmidt WA, Kinetic aspects of follicular development in the rat. n: Mahesh VB, Dhindsa DS, Anderson E, Kalra SP (eds.), Regulation of Ovarian and Testicular Function. New York, NY: Plenum Press, pp Hoak DC, Schwartz NB, 198. Blockade of recruitment of ovarian follicles by suppression of the secondary surge of folliclestimulating hormone with porcine follicular fluid. PNAS (USA) 77 : ngram DL, Atresia. n: Zuckerman S (ed.), The Ovary. NY: Academic Press, pp Knigge KM, Leathem JH, Growth and atresia of follicles in the ovary of the hamster. Anat Rec 124: Kohut JK, Jarrell JF, Younglai EV, Does dihydrotestosterone induce atresia in the hypophysectomized immature female rat treated with pregnant mare s serum gonadotropin? Obstet Gynecol 151: Lacker HM, Beers WH, Meuli LE, Atkin E, A theory of follicle selection. Hypothesis and examples. Biol Reprod 37:57-8 Lintern-Moore S, Everitt AV, Mariana JC, Mauleon P, The effect of restricted food intake and refeeding on the ovarian follicle population of the pre-pubertal Wistar rat. Reprod Nutr Dcv 21:611-2 Luna LG, Manual of Histological Staining Methods of the Armed Forces nstitute of Pathology, Third Edition. NY: McGraw-Hill Mandl AM, Zuckerman S Numbers of normal and atretic oocytes in unilaterally spayed rats. J Endocrinol 7: Mandl AM, Zuckerman S, Cyclical changes in the number of medium and large follicles in the adult rat ovary. J Endocrinol 8: Mueller PL, Schreiber JR, Lucky AW, Schulman JD, Rodbard D, Ross GT, Follicle stimulating hormone stimulates ovarian synthesis of proteoglycans in the estrogen-stimulated hypophysectomized immature female rat. Endocrinology 12: Osman P, Rate and course of atresia during follicular development in the adult cyclic rat. J Reprod Fertil 73:261-7 Otani T, Sasamoto 5, Plasma and pituitary hormone changes and follicular development after unilateral ovariectomy in cyclic rats. J Reprod Fertil 65: Peppler RD. Greenwald GS, 197. Effect of unilateral ovariectomy on ovulation and cycle length in 4- and 5-day cycling rats. Am J Anat 127:1-8 Richards JS, Kersey KA, 198. Changes in theca and granulosa cell function in antral follicles developing during pregnancy in the rat: gonadotropin receptors, cyclic AMP, and est. Biol Reprod 21: Richards is, Midgley AR, Protein hormone action: a key to understanding ovarian follicular and luteal cell development. Biol Reprod 14:82-94 Richards JS, Rao MC, reland JJ, Actions of pituitary gonadotrophins on the ovary. n: Crighton CB, Haynes NB, Foxcroft GR, Lamming GE (eds.), Control of Ovulation, 26th Easter School Symposium, University of Nottingham, Nottingham, England. London: Butterworths, pp Ryan Ri, Follicular atresia: some speculation on biochemical markers and mechanisms. n: Schwartz NB, Hunzicker-Dunn (eds.), Dynamics of Ovarian Function. NY: Raven Press, pp Schwartz NB, The role of FSH and LH and of their antibodies on follicle growth and ovulation. Biol Reprod 1: Taya K, Sasamoto S, Changes in FSH, LH, and PRL secretion and ovarian follicular development during lactation in the rat. Endocrinol Jpn 28: Trump BF, Arstila AU, Cellular Reaction to njury. n: Trump BF, LaVia M, Hill RB (eds.), Principles of Pathobiology. London: Oxford University Press, pp Welschen R, Dullaart J, Administration of antiserum against ovine follicle-stimulating hormone or ovine luteinizing hormone at pro-oestrus in the rat: effect on follicular development during the oncoming cycle. J Endocrinol 7:31-6 Welschen R, Dullaart J, de Jong FH, nter-relationships between circulating levels of estradiol-17j3, progesterone, FSH, and LH immediately after unilateral ovariectomy in the cyclic rat. Biol Reprod 18: Welschen t, Rutte M, Ovulation in adult rats after treatment with pregnant mare serum gonadotrophin during oestrus. Acta Endocrinol 68:41-49 Welschen RW, Dullaart J, Serum concentrations of FSH and LH after unilateral ovariectomy in the adult rat. J Endocrinol 63: Williams MA, Quantitative Methods in Biology. New York, NY: North-Holland Publishing Co. Zeleznick AJ, Midgley AR Jr. Reichert LE Jr, Granulosa cell maturation in the rat: increased binding of human chorionic gonadotropin following treatment with hormone in vivo. Endocrinology 95: Zlotkin T, Farkash Y, Orly J, Cell specific expression of immunoreactive cholesterol side chain cleavage cytochrome P-45 during follicular development in the rat ovary. Endocrinology 119:289-2 on 8 February 218