| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Obstetrics and Gynaecology, Diakonessenhuis Utrecht, 3582 KE Utrecht, The Netherlands
b Department of Human Genetics, University Medical Centre Utrecht, 3584 CX Utrecht, The Netherlands
c Department of Biochemistry, Cell Biology and Histology, University of Utrecht, 3584 CM Utrecht, The Netherlands
d Department of Cell Biology, University Medical Centre Utrecht, 3584 CX Utrecht, The Netherlands
e Department of Public Health, Faculty of Medicine, Erasmus University Rotterdam, 3000 DR Rotterdam, The Netherlands
f Department of Obstetrics and Gynaecology, University Medical Centre Utrecht, 3584 CX Utrecht, The Netherlands
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
aging, apoptosis, follicle, granulosa cells, ovary, ovum
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although the decline in follicle number concomitant with age has been established accurately, we have limited knowledge of the changes in follicle quality associated with age and the factors that govern these changes. As follicles "rest" in the ovaries for up to 50 yr, degenerative changes will probably accumulate in oocytes and granulosa cells [4]. According to the production line hypothesis, follicle quality is determined in utero; the best follicles are formed first in fetal life and are recruited first during reproductive life, whereas follicles of poor quality are formed later, and are recruited later. Accordingly, as women age, an increasing proportion of their follicles are of poor quality [5, 6]. The concept that follicle quality declines with age is further complicated by the dynamics of the follicle population. Follicles leave the resting pool on a regular basis and pass through subsequent developmental stages, but most degenerate and very few reach the final stage of ovulation [7]; in addition, at various developmental stages, follicles probably behave differently in response to degeneration-inducing factors [8].
In recent years, cryopreservation of ovarian tissue, in vitro development of immature follicles, and reproductive aging have become major research issues in reproductive biology, and assessment of follicle quality is of paramount importance in these disciplines, both in order to examine the decline in quality during freezing or tissue culture, and to investigate the age-related decline in follicle quality. It is clear that research in these areas will benefit from a better understanding of the cellular mechanisms that determine follicle quality and a method for objectively describing follicle quality in the human ovary.
Follicular degeneration, commonly referred to as atresia, is accompanied by ultrastructural, morphological changes in oocytes and granulosa cells. A number of studies have endeavored to describe these morphological signs of atresia in resting follicles, and a firm agreement for several signs can be found among these studies (Table 1).
|
We carried out a systematic morphometric investigation of the ultrastructure of the human resting follicle pool in a group of young healthy women who were not yet expected to show signs of age-related fertility decline [9] and whose follicles were likely to be of normal quality. Our aims were to 1) investigate the ultrastructural characteristics of the pool of resting follicles, 2) determine the distribution of resting follicles with and without signs of atresia, and 3) develop a method by which follicle quality could be objectively assessed.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Healthy women aged 25–32 yr were recruited from a group of patients who requested tubal sterilization. Each woman provided informed consent and allowed us to collect small ovarian biopsies during surgery. Permission to perform the assessment was also granted by the local ethics committee.
Women were selected on the basis of proven fertility (spontaneous pregnancy achieved within the last 12 mo), a regular menstrual cycle (cycle length 21–35 days, maximum intercycle difference 7 days), no endocrinological pathology, no use of oral contraceptives in the previous 2 mo, and in good general health as assessed by a questionnaire.
Women were scheduled for laparoscopic surgery in the preovulatory phase of the menstrual cycle as determined by transvaginal ultrasonography (dominant follicle >15 mm). With a biopsy forceps (Storz, Munich, Germany), two biopsies of ovarian cortex of 
3 x 3 mm were taken from the ovary that contained the dominant follicle. Biopsies were taken from a part of the ovarian cortex that was devoid of corpora albicantia and antral follicles bulging at the surface, and at a clear distance from the dominant follicle.
Histology
Biopsies were fixed immediately after biopsy by immersion in fresh Karnovsky fixative (2.5% glutaraldehyde and 2% paraformaldehyde in 0.1 M Na-cacodylate buffer, pH 7.3) for 60 min, and postfixed for 3 h in 2% osmium tetroxide in the same buffer. Samples were than stained en bloc with 2% aqueous uranyl acetate for 3 h, followed by dehydration in a graded series of acetone, and embedding in Durcupan ACM (Fluka, Buchs, Switzerland).
Because primordial or primary follicles measure about 25–75 µm in diameter [10], semithin sections (1 µm) were cut on an LKB III microtome at intervals of 50 µm throughout each biopsy, stained with toluidine blue, and screened for the presence of follicles by light microscopy.
Ultrathin sections (50–70 nm), corresponding to selected sections containing follicles, were prepared with a diamond knife on an LKB V ultramicrotome and mounted on single-slot formvar-carbon-coated copper grids. After staining with Reynold lead citrate [11], the sections were examined, and random small follicles with a visible oocyte nucleus were photographed with a transmission electron microscope (CM 10 Philips, Eindhoven, The Netherlands) at 80 kV (microscope magnification for a general view 1650x and 2950x for detail photographs).
Morphometry
Negatives were scanned (Agfa Duoscan; Agfa Gevaert GmbH., Köln, Germany) at a resolution of 1120 pixels per inch and analyzed with an image analysis software program (Scion Image ß release 3b; Scion Corp., Frederick, MD). Further calculations were performed with a statistical software program (SPSS release 8.0; SPSS Inc., Chicago, IL).
We designed the following set of morphometric variables based on the signs of atresia that are presented in Table 1:
The following discrete variables from Table 1 were scored for oocytes and granulosa cells.
Oocytes Nuclear membrane intact, or indented or ruptured; mitochondrial membranes intact with parallel or transverse cristae, intact with arch-like cristae or ruptured; mitochondrial matrix low, moderate, or high density; smooth endoplasmic reticulum (SER) flat or dilated and their location close (associated) or distant (dissociated) to the mitochondria; number of microvilli low or high; the Golgi complex absent or present in the oocyte profile; and the Golgi saccular membranes flat or dilated. Condensation of the oocyte chromatin was not separately scored because this was observed in only 1 out of 182 follicles. Rough endoplasmatic reticulum (RER) was not scored because it was scarce in oocytes.
Granulosa cells Chromatin condensation high or low; electron density of the cytoplasm high or low; mitochondrial membranes intact or ruptured; SER and RER flat or dilated. For further calculations these individual granulosa cell scores were expressed as scores per follicle (i.e., the proportion of granulosa cells with low condensation of the chromatin, the proportion of cells with low density of the cytoplasm, the proportion of cells with intact mitochondria and the proportion of cells with flat SER and RER). The aspect of the Golgi complex in the granulosa cells was not scored because the frequency of Golgi visible in the granulosa cell profiles was very low.
Three variables, which were supposed to add quantitative information on the developmental stage of the follicle were included: follicle stage, the absence or partial presence of the zona pellucida, and the number of zonula adherens [12–15]. Follicle stage definitions were adopted from those of Lintern-Moore et al. [10]: primordial (primary oocyte surrounded by a single layer of flattened granulosa cells), intermediary (primary oocyte surrounded by a single layer of flattened and cuboidal cells), and primary (primary oocyte surrounded by a single layer of cuboidal granulosa cells).
Methods of Analysis
We aimed to construct a follicle quality score from the data of all morphological variables. Therefore, the data matrix of all variables was ordered with a principal components analysis (PCA) [16]. In PCA, a common factor for the variables can be extracted from the correlations between the variables. This common factor, which best explains the total variance of the data matrix, is called the first principal factor. Because we chose morphological variables that reflect follicle quality, we expect that the first principal factor can be assumed to be the best overall reflection of follicle quality.
Because not all variables are correlated closely, the first principal factor will only partly explain the total variance of the data matrix. A second principal factor can be identified, which explains the variance of the variables not yet represented in the first principal factor. The second principal factor reflects another characteristic of follicular morphology that is not correlated to follicle quality.
The results of the PCA can be displayed in a two-dimensional graph or biplot [16] in which the horizontal axis represents the first principal factor and the vertical axis represents the second principal factor. The morphological variables are displayed as vectors. The angles between the vectors indicate the correlations between the variables. The length of the vectors indicates the contribution of the variables to the total variance of the data matrix. The positions of the individual follicles are displayed by markers.
The number of follicles assessed per woman differed a great deal. To remove the undue influence of women with many follicles we weighted the data for the follicles in the PCA with the inverse of the number per woman. Because follicles from the same woman can be expected to be interdependent with respect to quality, developmental stage, or both, we refrained from pooling all follicles. Also, further accounting for subject variance may obscure noteworthy differences in follicle quality per woman. Missing values were few, and therefore, were replaced by the mean of all observations. Classes of discrete variables were included in the PCA individually, except for follicle stage, leading to 38 variables in the PCA. The effect of log-transformation of variables that appeared to be not normally distributed was evaluated, but this had a negligible effect on the outcomes of the PCA.
Because the first principal factor is assumed to be the best reflection of follicle quality, the position of the individual follicles on the horizontal axis of the biplot can be used to rank follicles by quality. To construct a compact formula for a follicle quality score, the number of variables needed to reliably provide this position was reduced by performing a stepwise forward selection, which runs as follows: in step one the variable with the highest explained variance for the first principal factor is selected. In step two, from all remaining variables the variable is selected that adds the highest explained variance in combination with the variable selected in step one, and so on. The selection was continued until the following step would add less than 0.02 explained variance to the model. If one of the classes of a discrete variable was being selected, all other classes of that variable were included in the same step. All analyses were performed with CANOCO 4.0 for Windows [17].
For a more detailed description of the statistical model for the PCA, Yki = (Bk1·Xil + Bk2·Xi2) + residual, where Yki = the standardized observation of variable k on follicle i, Bk1 = the variable score for variable k on the first principal factor (horizontal axis in Fig. 1A), Xil = the object score for follicle i on the first principal factor (horizontal axis in Fig. 1B), Bk2 = the variable score for variable k on the second principal factor (vertical axis in Fig. 1A) and Xi2 = the object score for follicle i on the second principal factor (vertical axis in Fig. 1B). Residuals are minimized by least squares approximation.
|
The position of an individual follicle is determined by the sum of the standardized scores of all variables for that particular follicle. Standardization of variable and object scores is rather arbitrary; for example, in the given model, doubling the B's and halving the X's leads to the same outcome of the model. In our analysis, to account for differences in units and scales, we standardized by rescaling the variable scores by subtracting with the mean score value and dividing by the standard deviation of the score value. These standardized scores are multiplied with the X-coordinates of the morphological variables in the biplot, and the sum of these scores provides the X-coordinate of the follicle. The same applies for providing the Y-coordinate.
For our final score algorithm we rescaled in such a manner that the range of Xi1 runs from 0 to 10 when original (i.e., unstandardized) Yki are used.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
The results of the PCA are displayed in Figure 1, A and B. In Figure 1A the vector numbers correspond to the numbers in Table 3. Variable scores indicating atresia were closely correlated to the first principal factor (horizontal axis); for example, ruptured mitochondrial membranes, the cytoplasmic fraction of lipid droplets, and vacuoles. Variables indicating developmental stage of the follicle; for example, follicle stage and the presence or absence of a zona pellucida, did not correlate well with the first principal factor but were close to the second principal factor (vertical axis). Figure 1B shows the position of all individual follicles in the same biplot. Because the first principal factor represented the variance of all atretic variables, the position of the follicles on the horizontal axis was taken as an objective score for morphological follicle quality. This axis was rescaled, ranging from 0 to 10 points (low to high quality). The distribution of all follicles on this quality scale is shown in Figure 2 and was skewed toward high scores with a median value of 7.3 (10th and 90th percentiles, 4.6 and 8.6). The median follicle score values for each subject are listed in Table 2.
|
|
The outcomes of all variables are summarized in Table 3, in which the explained variance for the first principal factor is presented as follows from univariate analysis; for example, in order of importance: ruptured nuclear membranes (0.41), ruptured (0.41) or intact (0.37) mitochondrial membranes, and flat or associated SER (0.31).
From all follicles 31.5% were at the primordial stage, 54.1% were at the intermediary stage, and 14.4% were at the primary stage. No differences for the areas of the oocytes and the oocyte nuclei or for the nucleus:oocyte ratios were observed for the different follicle stages. The areas of the granulosa cells and granulosa cell nuclei increased significantly from 23.2 to 36.0 µm2 and from 9.5 to 15.2 µm2, respectively, in primordial to primary follicles (ANOVA, P < 0.001). The nucleus:granulosa cell ratio remained constant at 0.42 in primordial to primary follicles.
Free lipid droplets in the cytoplasm were found in small quantities. In the presence of other atretic signs, the amount of lipid droplets did increase, but lipid droplets were located mainly inside multivesicular bodies. A limited number of small multivesicular bodies was sometimes present in follicles without other signs of atresia, however, numbers and sizes of multivesicular bodies clearly increased when more signs of atresia were noted. Their predominant location was the perinuclear area. Vacuoles were located both centrally and peripherally in the cytoplasm and varied greatly in size. The presence of large numbers of vacuoles was always associated with other signs of atresia. The contents of vacuoles could appear as empty or as amorphous material of low electron density. In several follicles vacuoles appeared to come forth from swollen and at times confluenced vesicles of the SER. Furthermore, big vacuoles developed within the membrane of large multivesicular bodies. The Golgi complex, if it was present, accounted for only a small proportion of the cytoplasmic volume. Generally, a Golgi complex consisted of 5–10 orderly stacked flat saccules. A distinct swelling of these saccules could be noted in association with other atretic signs. Numerous mitochondria were present in nearly all follicles. The profile areas of individual mitochondria showed large variation, ranging from 0.05 to 2.23 µm2.
No differences for cytoplasmic fractions of lipid droplets, multivesicular bodies, vacuoles, Golgi, mitochondria, and the mean profile area of mitochondria were observed for the different follicle stages. The number of zonula adherens was not clearly related to atretic signs and showed an increase of 1.5 in primordial follicles to 3.7 in primary follicles (P = 0.06).
In the majority of follicles the oocyte nucleus was centrally placed in the cell with a regular spherical or oval shape. More pronounced indentations and finally rupture of the oocyte nuclear membrane was observed in 32.4% of follicles, and was associated with the accumulation of other atretic signs. Condensation of oocyte chromatin was observed in only 1 out of the 182 follicles. In follicles without signs of atresia the organelles in the oocyte were clustered around the nucleus, and are commonly referred to as Balbiani's vitteline body [18]. Round and elongated mitochondria with parallel or transverse cristae and a matrix of moderate electron density were observed in the majority of follicles (91.8% and 63.7%, respectively). With mounting signs of atresia a reduction of the number of mitochondrial cristae was observed together with a changing density (high or low) of the matrix. Furthermore, mitochondrial shape became irregular and ultimately, mitochondrial membranes ruptured. SER was abundant in all oocytes in vesicular and tubular form. In 65.9% of follicles vesicles were small and tubules were flat and in close association to the mitochondria. In association with other atretic signs, a dilation of SER occurred, and the association with mitochondria was lost. In 25.3% of follicles the assessed cross-section did not contain Golgi complexes in the oocyte. In association with other signs of atresia, Golgi saccules became dilated. The numbers of microvilli were small in most follicles (62.6%). These numbers did not seem to be influenced by follicle health, but by follicle stage: many microvilli were seen in 28% of primordial, 40% of intermediary, and 46% of primary follicles. The (partial) presence of a zona pellucida was also related to follicle stage and increased from 7% to 19% from primordial to primary follicles. No relation with follicle stage was observed for the other variables.
In 79.7% of granulosa cells condensation of the chromatin was low. In the other granulosa cells patches of condensed chromatin along the nuclear membrane were associated with alterations of cell organelles and increased electron density of the cytoplasm. A small proportion of granulosa cells (3.2%) contained ruptured mitochondria. SER and RER were generally flat (88.5%) in granulosa cells, but were found dilated in dark granulosa cells. Dark (condensed chromatin, high electron density of cytoplasm) and light (uncondensed chromatin, low electron density of cytoplasm) granulosa cells were found adjacent in follicles that contained oocytes with and without signs of atresia. Several follicles could be found with nearly destroyed oocytes, still surrounded by a number of light granulosa cells.
The results of the stepwise forward selection for explained variance of the first principal factor are shown in Table 4. The selected variables are also indicated in Table 3. Seven variables were selected, which explained 93% of the variance of the scores based on all variables (SD = 0.44 for the 10-point follicle quality scale). The first selected variable, the presence of ruptured mitochondrial membranes, explained a large proportion of the quality score. Other obvious markers of atresia were not selected because they were closely correlated to the presence of ruptured mitochondrial membranes, and therefore added little to the explained variance. In other words, after each selection step of the analysis the order of importance of the remaining variables changes depending on their correlation with earlier selected variables, which explains why some variables that have a higher explained variance from univariate analysis as listed in Table 3 are not selected, whereas other variables with a lower explained variance in univariate analysis are. This is the case, for example, with the mean area of the oocyte that gains importance over the mean area of the oocyte nucleus after selection step 4 (Table 4).
|
From the seven selected variables, five reflect atretic signs present in the oocyte and two reflect atretic signs in the granulosa cells. From the selected parameters, follicle quality score can be calculated with the algorithm in Table 4, which includes two examples of such calculations for the follicles shown in Figure 3. Figure 4 shows further examples of morphological variables in healthy and atretic situations.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our aim was to describe the morphology of the resting follicle pool and to construct an objective method to use the morphological variables for a follicle quality score. This was attempted by running a PCA on all variables. The first principal factor predominantly explained the variance of atretic signs, which generally correlated well with this factor. Thus, the first principal factor seemed to accurately reflect quality, and the horizontal axis in the PCA biplot was taken as the gradient on which follicles can be placed, according to their quality. The distribution of follicles along this axis on a scale of 0–10 showed that most resting follicles were of good quality, showing little signs of atresia. It should be noted that a very small number of follicles showed such severe atretic changes that morphometry could not be performed reliably, and were therefore excluded from the analysis.
From light-microscopic studies, although at that level more subtle signs of atresia could have remained obscure, similar findings of a low rate of atresia have been reported [2, 7, 24, 25]. Apparently, resting follicles accumulate little damage, which may be associated with their low metabolic rate [22]. The increase in atresia may occur only after the initiation of rapid follicle growth.
The second principal factor grouped variables such as follicle stage and zona pellucida score, suggesting a reflection of follicle development. Follicle development did not correlate well with most atretic signs, indicating that in resting follicles the rate of atresia is not different for primordial, intermediary, or primary follicles.
The combination of atretic signs in follicles suggest a distinct sequence in which degenerative changes take place. The presence of few atretic signs suggest early atresia. These few signs were frequently the same; for example, multivesicular bodies and lipid droplets. Other signs, for example, indentation or rupture of the nuclear membrane, were frequently found in association with the presence of a larger number of atretic signs, suggesting progressive atresia.
Accordingly, the following sequence of degenerative changes in resting follicles can be hypothesized. First, an increase of multivesicular bodies and lipid droplets is noted. Next, the SER dissociates from the mitochondria and a rapid accumulation of vacuoles occurs. Thereafter, the nuclear membrane begins to show indentations and mitochondria become more irregularly shaped with a reduction of cristae and changing density of the matrix. The SER and Golgi show progressive swelling, the SER even transforms to large vacuoles. In terminal phases of atresia mitochondrial membranes rupture, leaving empty remains of mitochondrial structures, nuclear membranes show large indentations or ruptures, and the largest part of the cytoplasm consists of small to very large vacuoles.
According to this sequence of degenerative changes, atresia in resting follicles appears to be a process of necrosis and not of apoptosis. Necrosis, which is a result of environmental injuries, changes the cellular appearance by increased permeability of membranes, destruction of cytoplasmic structures, and finally nuclear degeneration. Apoptosis, or programmed cell death, starts with cellular and chromatin condensation (pycnosis), which in our material, was observed in only 1 out of the 182 examined follicles. In apoptosis, pycnosis is followed by nuclear and plasma membranes becoming irregular and cell organelles remaining intact until the final stages of cell death [26, 27]. Cytogenetic studies also failed to demonstrate apoptosis in resting follicles. In contrast, the role of apoptosis in atresia of growing follicles is firmly established [28, 29]. However, it is possible that apoptosis could not be demonstrated in resting follicles because of the high pace of the apoptotic process and the supposed low rate of atresia in these follicles [2, 7, 24]. Furthermore, considering the low rate of atresia in our sample of resting follicles, we cannot draw final conclusions from the present study on the characteristics of the atretic process.
Because healthy granulosa cells were frequently present in follicles with degenerative oocytes, it seems that in resting follicles atresia starts in the oocyte, which is in line with the findings of other researchers [8, 30]. In contrast, atresia in growing follicles has been shown to start in granulosa cell layers [8, 24, 31]. In these follicles the granulosa cells probably shield the oocyte from environmental insults.
The early involvement of mitochondria in the atretic process suggests a role for oxygen radical-induced damage. Oxygen radicals induce cellular damage as they accumulate with age. The mitochondria would be the first organelles to show degenerative signs because they are the site of oxygen radical production, and accumulation of mitochondrial DNA deletions with age has been shown [32–34].
Scoring all variables to determine follicle quality is laborious and not practical if large numbers of follicles are to be assessed. Several variables make only a small contribution to the total variance of the data. Therefore, we reduced the number of variables by stepwise forward selection to a much smaller number that would still closely approximate the original PCA quality score. From our selection, seven variables resulted with an explained variance of the original score of 93%. Five of these variables reflected atretic signs in the oocyte, which is in line with the assumption that the oocyte mainly determines the quality and fate of a resting follicle. The algorithm to calculate the score from these variables is presented. With this algorithm follicle scores can be obtained for those aiming to compare follicle populations on atresia (examples in Fig. 3 and Table 4). Of course, the biological and clinical relevance of this general follicle quality score needs further evaluation and validation. We hope future studies will be undertaken to test the value of this score for more specific applications, such as investigations on the effects of tissue culture, cryopreservation, or aging.
|
FOOTNOTES |
|---|
1 Financial support was received from Serono Benelux, Abbott. 
2 Correspondence: J.P. de Bruin, Department of Obstetrics and Gynaecology, Diakonessenhuis Utrecht, Bosboomstraat 1, 3582 KE Utrecht, The Netherlands. FAX: 31 30 2505433; jp{at}debruin-kers.demon.nl
Accepted: October 25, 2001.
Received: January 20, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  C. Tatone, F. Amicarelli, M. C. Carbone, P. Monteleone, D. Caserta, R. Marci, P. G. Artini, P. Piomboni, and R. Focarelli Cellular and molecular aspects of ovarian follicle ageing Hum. Reprod. Update, March 1, 2008; 14(2): 131 - 142. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 C. G. Westergaard, A. G. Byskov, and C. Y. Andersen Morphometric characteristics of the primordial to primary follicle transition in the human ovary in relation to age Hum. Reprod., August 1, 2007; 22(8): 2225 - 2231. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. R. Hurst, J. M. Mora, and M. A. Fenwick Caspase-3, TUNEL and ultrastructural studies of small follicles in adult human ovarian biopsies Hum. Reprod., August 1, 2006; 21(8): 1974 - 1980. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M.-M. Dolmans, N. Michaux, A. Camboni, B. Martinez-Madrid, A. Van Langendonckt, S. Annarita Nottola, and J. Donnez Evaluation of Liberase, a purified enzyme blend, for the isolation of human primordial and primary ovarian follicles Hum. Reprod., February 1, 2006; 21(2): 413 - 420. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. G. Gosden Germline stem cells in the postnatal ovary: is the ovary more like a testis? Hum. Reprod. Update, May 1, 2004; 10(3): 193 - 195. [Full Text] [PDF]
|
|
|
|
 |
|
 J.P. de Bruin, M. Dorland, E.R. Spek, G. Posthuma, M. van Haaften, C.W.N. Looman, and E.R. te Velde Age-Related Changes in the Ultrastructure of the Resting Follicle Pool in Human Ovaries Biol Reprod, February 1, 2004; 70(2): 419 - 424. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. R. Baerwald, G. P. Adams, and R. A. Pierson Characterization of Ovarian Follicular Wave Dynamics in Women Biol Reprod, September 1, 2003; 69(3): 1023 - 1031. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K.L.T. Schmidt, A.G. Byskov, A. Nyboe Andersen, J. Muller, and C. Yding Andersen Density and distribution of primordial follicles in single pieces of cortex from 21 patients and in individual pieces of cortex from three entire human ovaries Hum. Reprod., June 1, 2003; 18(6): 1158 - 1164. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
