| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ARTICLE |
-Irradiation
Laboratoire de Physiologie et Physiopathologie, Centre National de la Recherche Scientifique-UMR 7079, Université Paris VI (S.M., C.J.G., A.L., N.C., S.M.), Paris, France; Institut National de la Santé et de la Recherche Médicale, U-329, Hôpital Debrousse (M.G.F.), 69322 Lyon, France; and Département de Radiobiologie et Radiopathologie, Commissariat à l’Energie Atomique (H.C.), 92265 Fontenay-aux-Roses, France
Address all correspondence and requests for reprints to: Dr. S. Magre, Laboratoire de Physiologie et Physiopathologie, UMR 7079, Université Paris VI, 7 quai Saint-Bernard, 75005 Paris, France. E-mail: solange.magre{at}snv.jussieu.fr
 |
Abstract |
|---|
 Top Abstract IntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-irradiation. As a consequence of the severe depletion in oocytes, females evidenced premature ovarian failure from 6 months on. Nevertheless, puberty onset and fertility at the beginning of reproductive life were similar to those of controls. The differentiation and evolution of the entire follicular population were followed during the immature period, using follicle counts, in situ hybridization of follicular maturation markers, and analysis of atresia. Primordial follicles were much more affected by irradiation (1.4–1.9% of controls) than growing follicles (30–45% of controls). As the very low number of primordial follicles remained constant throughout this period, it may be considered that the growing follicle pool plays the role of follicular reserve, permitting the transient normal fertility of irradiated females. Within the neonatal period, primary and secondary follicles, as revealed by proliferating cell nuclear antigen immunostaining, remain quiescent longer in irradiated than in control ovaries. Consequently, the majority of the most mature follicles (i.e. the first follicular wave) characterized by a high expression of aromatase transcripts during the infantile period, are missing in irradiated ovaries. Concomitantly, the 17ß-estradiol plasma peak is absent, and plasma FSH levels are higher than those in control females.
In conclusion, these observations emphasize that the female reproductive life span depends not merely on the size of the primordial follicle stock, but also on the entire follicle complement as well as follicular dynamics during the immature period.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
The primordial follicle stock is not renewable and serves as a reserve for the entire reproductive life span of the adult. The number of primordial follicles, therefore, limits the fertility of the female. A decrease in the initial complement of oocytes provokes anomalies ranging from reduced fertility to premature ovarian failure or even complete sterility. In the most severe cases of germ cell deficiency seen in genetic disorders such as atm, Sl, or W mice, sterile females have dysgenesic ovaries that are incapable of forming follicles after birth and become streak gonads (4, 5). In humans, 45,XO Turner’s syndrome is associated with a high rate of oocyte attrition during fetal life, leading in almost every case to ovarian dysgenesy and sterility. In extremely rare cases, spontaneous pregnancies have been described (6). In mice, as in humans, X chromosome monosomy induces a severe reduction in the number of oocytes. Nevertheless, XO mice are initially fertile; the consequence of the oocyte deficiency is a shortened reproductive life span (7, 8, 8).
Until now, neither the mechanisms regulating the size of the pool at birth nor those controlling the release of follicles from their resting status to initiate growth have been clearly understood. Recent studies of transgenic mice have shown the possible involvement of pro- and antiapoptotic factors, Bax and Bcl-2, respectively, in the control of the size of the primordial follicle stock (10, 11). The growth factors nerve growth factor (NGF) and anti-Mullerian hormone (AMH) have also recently been considered as playing a role in this control. Using a gene disruption strategy, NGF has been shown to be a proliferative signal inducing the exit of follicles from the stock, and inversely, AMH has been shown to be an inhibiting factor in the initiation of primordial follicle growth (12, 13, 14). In rat experimental models a controlled reduction of the size of the oocyte pool was obtained by inducing the death of mitotic oogonia by exposure to either x-irradiation or antimitotic drugs such as busulfan, with a maximal toxic effect on d 15 post conception (pc) (15, 16). In irradiated females the ovary, although severely depleted in primordial follicles, nevertheless contains a relatively high number of oocytes, corpus lutei, and remaining antral follicles at 3 months (17). These observations were interpreted as indicative of a compensatory hypertrophy, which would allow the ovary to function at least at the beginning of the reproductive life span. Using various doses of busulfan, it was shown that in the rats treated with the highest dose, the remaining primordial follicles had begun to grow very early in life, and the stockpile was rapidly exhausted. There was an inverse correlation between the size of the primordial follicle stock at birth and the rate at which the surviving follicles began to grow (18).
To further investigate the relationship between the size of the follicular reserve and the functionality of the ovary, we have undertaken analysis of the morphological and functional differentiation of the ovary in females -irradiated on d 15 pc. Parallel to the study of reproductive performance of irradiated females, particular emphasis has been placed on the analysis of ovarian differentiation in immature females from birth to d 28 pn. With this objective, in addition to determining hormonal balance, we have followed in situ the evolution of the follicular population by morphological approaches, studied the expression of follicular maturation markers, and investigated follicular atresia.
|
Materials and Methods |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-irradiation using a 60Co source with a total dose of 1.5 Gy (dose rate, 0.25 Gy/min). Animals were killed by decapitation on d 3, 6, 9, 12, 15, 21, and 28 pn and at 8 months. Trunk blood was collected into heparinized tubes and centrifuged, and the plasma was kept frozen until assayed for FSH, inhibin B, estradiol, and testosterone.
Fourteen irradiated and 15 control females were used to analyze fertility. After weaning, females were weighed daily and checked for vaginal opening. Cycles were followed for 3 wk. Then they were mated with males during 5 consecutive nights, once every 5 wk. The length of mating was chosen to cover the estrous cycle. Litters were sexed, weighed, and killed 2 d after birth.
Ovarian histology and quantification of follicle number
On d 9, 15, 21, and 28 pn, ovaries were fixed in Bouin’s liquid and embedded in paraffin. Sections of 5 µm thickness were stained with hematoxylin-eosin or Tuchmann’s blue. The follicles were counted in every fifth section and classified using the oocyte nucleus as a marker, according to the stages of follicular development previously described (15, 19) with some modifications (Table 1
).
|
In situ hybridization
The cDNA used for synthesis of the different riboprobes was obtained by RT-PCR and subcloned into pGEMTeasy or pBluescript vectors, and the nucleotide sequence was verified. GenBank accession numbers and positions are as follows: AMH, S98336, nucleotides 5388–6049; aromatase, M33986-1, nucleotides 797-1487; activin ßA-subunit, M37482-1 nucleotides 1040–1483; steroidogenic factor-1 (SF-1), D42156-1, nucleotides 463–968. Riboprobes were generated by transcription with digoxigenin-labeled deoxy-UTP and the appropriate SP6 or T7 RNA polymerase.
Ovaries were fixed for 1 h at 4 C with 2% (wt/vol) paraformaldehyde in PBS, rinsed with 12%, 15%, and 18% (wt/vol) sucrose in PBS for 30 min each, embedded in Tissue-Tek OCT compound (Miles, Inc., Elkhart, IN), and frozen at –80 C. Sections (5–7 µm thick) were obtained and stored at –20 C. After thawing, frozen sections were delipidized in chloroform for 1 min, rehydrated in PBS, postfixed for 20 min with 2% paraformaldehyde in PBS, pH 9, and treated for 10 min in 0.25% acetic anhydride in 0.1 M triethanolamine, pH 8. After 2 h of prehybridization in 50% formamide, 2x sodium saline citrate, 5x Denhardt’s solution, 50 µg/ml yeast tRNA, 250 µg/ml salmon sperm DNA, 4 mM EDTA, and 2.5% dextran sulfate at 55 C, hybridization was carried out at 55 C in a moist chamber with riboprobes diluted in the same buffer without salmon sperm and EDTA. Subsequent washes and detection of the riboprobes using alkaline phosphatase antidigoxigenin were performed as previously described (22).
Terminal deoxynucleotidyltransferase-mediated deoxy-uridine 5'-triphosphate-fluorescein nick end labeling (TUNEL)
Detection of apoptotic cells was performed on sections previously treated for AMH in situ hybridization using the in situ cell death detection kit, fluorescein (Roche). After PBS washing, sections were incubated 1 h at room temperature with the TUNEL reaction mixture containing terminal transferase to label free 3'-hydroxy ends of genomic DNA with fluorescein-labeled deoxy-UTP. TUNEL labeling was then observed with an epifluorescence microscope (Carl Zeiss New York, NY).
Immunocytochemistry
Double labeling was performed against proliferating cell nuclear antigen (PCNA; monoclonal antibody sc-56, Santa Cruz Biotechnologies, Inc. Santa Cruz, CA) and laminin (L-9393, Sigma St. Louis, MO). Frozen tissue sections were delipidized in chloroform, rehydrated in PBS, and boiled twice in 10 mM sodium citrate for antigen retrieval. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide for 5 min. Slides were incubated with anti-PCNA antibody (diluted 1:200) at room temperature for 2 h. After washing in PBS, slides were incubated for 1 h at room temperature with a secondary antimouse Ig biotin-conjugated antibody (diluted 1:200). The avidin-biotin complex reaction was performed with streptavidin-peroxidase from the L-SAB+ staining kit (DAKO Corp. Carpenteria, CA). After washing in PBS, subsequent immunofluorescence was determined with a primary anti-laminin antibody incubated overnight at 4 C (diluted 1:500). After washing, slides were incubated for 1 h at room temperature subsequently with biotinylated antirabbit Ig antibody (diluted 1:500; RPN 1004, Amersham Pharmacia Biotech) and with fluorescein-conjugated streptavidin (diluted 1:100; RPN 1232, Amersham Pharmacia Biotech). Finally, slides were counterstained with hematoxylin.
Histoenzymology: 3ß-hydroxysteroid dehydrogenase (3ß-HSD) activity
3ß-HSD activity was revealed by deposits of formazan in delipidized and rehydrated sections incubated with 5ß-androstan-ß-ol-17one (3ß-etiocholanolone; Sigma) as substrate in the presence of nitro blue tetrazolium and NAD, according to the method described by Bara and Anderson (23).
Hormone assays
Plasma LH and FSH were assayed as previously described (24) using RIA kits provided by Dr. A. F. Parlow and the NIDDK (Baltimore, MD) with highly purified rat LH (NIDDK I-9) and FSH (NIDDK I-8) for iodination, reference preparations (rLH-RP3 and rFSH-RP2), and appropriate antisera (anti-rLH-S11 and anti-rFSH-S11). Inhibin B was assayed with double antibody ELISA (Argene-Biosoft, Varlhes, France) (25). Plasma 17ß-estradiol (E2) and testosterone (T) were assayed by specific RIA after ethyl ether extraction, followed by chromatographic purification on Celite columns, as previously described (26, 27). RIA for LH and FSH as well as ELISA for inhibin B were performed in duplicate for individuals. RIAs for E2 and T were performed on pooled samples. Intraassay coefficients of variation were 5.1%, 7.0%, 4.9%, and 5.0%, and interassay coefficients of variation were 11.2%, 12.5%, 12%, and 9–10% for FSH, LH, inhibin B, and steroids, respectively. The sensitivities of the assays were 0.1 ng/ml, 3.0 ng/ml, 15 pg/ml, 3 pg/tube, and 5 pg/tube for FSH, LH, inhibin B, E2, and T, respectively.
Statistical analysis
Follicle numbers and hormone concentrations were analyzed using two-ways ANOVA and expressed as the mean ± SEM. The differences were considered significant when P < 0.05. Proportions of pregnant females were compared using the two-sided test of exact probabilities in 2 x 2 contingency tables (28). Densitometric analysis of signals from the Southern blots of DNA fragmentation gel analysis were analyzed using the NIH Image program, and mean values were compared using two-way ANOVA.
|
Results |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
).
|
). Uteri were similar to those of controls until d 21 pn. On d 28 pn, they were heavier in irradiated females (data not shown).
|
, A and B, arrows) were classified as primary follicles from the beginning of their formation. As folliculogenesis progressed in a centrifugal fashion, the stock of primordial follicles became established at the periphery of the ovary (Fig. 2A). In irradiated ovaries, fragmentation of the ovigerous cords immediately gave rise to primary follicles (Fig. 2, C and D, arrow). Very few primordial follicles were detected, and no primordial follicle stock was observed at the periphery of the ovary (Fig. 2C). In regions devoid of germ cells, epithelial sterile cords reminiscent of fetal ovigerous cords never split and disappeared progressively within the first 2 wk of life (data not shown).
|
, A–D, arrows). Preantral follicles, present in a very limited number in the central part of controls, were absent from irradiated ovaries. Immunocytochemical detection of the proliferation marker PCNA demonstrated that follicles at the same stage of development were indeed at different states of maturation in control and irradiated ovaries. In control ovaries about half of the granulosa cells of growing follicles located in the center of the ovary were PCNA positive (Fig. 2, E–G, arrows). In irradiated ovaries, the granulosa cells of all follicles, even the most developed, were mainly PCNA negative (Fig. 2, H–J, arrows).
On d 6 pn, granulosa cells of primary, secondary, and preantral follicles were highly proliferating in control ovaries (Fig. 2, K–M, arrowheads). In irradiated ovaries, most follicles still presented resting granulosa cells (Fig. 2, N–P, asterisks). Only a few follicles in the center of the ovary had all of their granulosa cells positive for PCNA (Fig. 2, N–P, arrowheads). These observations suggest that the first follicles formed in irradiated ovaries were resting on d 3 pn and began to grow around d 6 pn, whereas the first follicles formed in control ovaries had begun to grow as early as d 3 pn.
The prolonged resting status of follicles in irradiated ovaries taken together with the absence of preantral follicles on d 3 pn may well indicate a time lag in the follicular development in irradiated ovaries compared with that in controls.
Follicular counts in immature females
To define more specifically follicular dynamics in immature animals, we analyzed the evolution of follicle populations by follicle counts using classical histology, during the infantile (d 9 and 15 pn) and juvenile (d 21 and 28 pn) periods. Follicle classification was established according to Table 1, and the results concerning healthy follicles are shown in Table 3 and Fig. 3. The massive depletion of germ cells and specifically the marked decrease in the number of primordial follicles reported by Beaumont (15) on d 25 pn were observed throughout the immature period in the present work (Table 3). The number of primordial follicles in irradiated ovaries, expressed as the percentage of the number in controls (P i/c), was 1.4–1.9% throughout the immature period. Depletion was less severe regarding the population of growing follicles, i.e. primary, secondary, preantral, and antral follicles (Fig. 3). The P i/c for growing follicles varied between 30 and 45%.
|
|
). It also resulted from a normal recruitment of primary follicles to enter the pool of secondary follicles. Indeed, the number of secondary follicles in irradiated ovaries was 50% of that in controls from d 9 and 21 pn, underlying the parallel time course of changes in this population between irradiated and control ovaries.
In contrast, changes in the number of preantral and antral follicles were not parallel between irradiated and control ovaries from d 9 to 21 pn. Although the number of follicles of both categories was maximum on d 15 pn in controls, it was maximum on d 21 pn in irradiated ovaries. This delay taken together with the absence of preantral follicles on d 3 pn (see above) and the extremely few antral follicles on d 9 pn (P i/c, 1.1%; Fig. 3) confirms the time lag in follicular development in irradiated ovaries compared with controls. Moreover, it is noteworthy that on d 21 pn the number of preantral and antral follicles was remarkably elevated in irradiated ovaries. The P i/c were 63.0% and 73.2%, respectively, and the number of preantral and antral follicles in irradiated ovaries was not statistically different from that in controls.
Together these results show that the severe deficit in the number of the most mature follicles on d 3 and 9 pn was not observed later and specifically not on d 21 pn. As only healthy follicles have been considered, follicular atresia must be taken into account to clarify follicular dynamics during the immature period.
Follicular atresia in immature females
As the first follicles that enter the growing pool at birth are not destined to ovulate and undergo atresia before puberty during the third pn week (1, 3), we investigated their evolution throughout the immature period (Fig. 4).
|
). From d 21 pn, atretic follicles were essentially antral follicles. In irradiated as in control ovaries, their number increased until d 28 pn (Fig. 4B). At all stages, the number in irradiated ovaries was below that in controls. It was, however, significantly lower on d 15 and 21 pn than on d 28 pn (Fig. 4B).
To identify the early signs of follicular atresia, we analyzed more precisely cellular apoptosis by studying DNA fragmentation either in situ with the TUNEL method (Fig. 4A) or with a global gel electrophoresis approach (Fig. 4, C and D). TUNEL was systematically performed on sections treated by in situ hybridization for the expression of AMH (Fig. 4A), which is known to be present in granulosa cells of healthy preantral and antral follicles and absent from atretic follicles (30, 31). On d 9 pn, in irradiated as well as in control ovaries no follicle displayed TUNEL-positive granulosa cells (at least two granulosa cells by follicle). The follicles localized in the center of the ovary expressed AMH more weakly than those on the periphery. They represented nearly half of the growing follicles in control ovaries and only one third in irradiated ones (Fig. 4A, a–d). On d 15 pn, atretic follicles located essentially in the center of the ovary were significantly fewer in irradiated than in control ovaries (Fig. 4A, e–h). Counts on every tenth section of four ovaries in both cases revealed that 13.1 ± 4.3% of growing follicles displayed at least two TUNEL-positive granulosa cells in irradiated ovaries vs. 40.2 ± 4.7% in controls (P = 0.005). Electrophoresis of digoxigenin-dideoxy-UTP-labeled fragmented DNA confirmed that cellular apoptosis was reduced in irradiated ovaries compared with controls on d 15 pn (data not shown) and 18 pn (Fig. 4, C and D). For this latter stage, apoptosis in irradiated ovaries represented half of that in controls. In contrast, on d 21 and 28 pn, no significant difference was noted [Fig. 4, A (i–p), C, and D].
The reduction in follicular atresia in irradiated vs. control ovaries on d 15 and 18 pn indicates that many of the most mature follicles destined to degenerate before ovulation in control ovaries are absent in irradiated ovaries. With the development of subsequent follicular waves, the proportion of healthy and atretic follicles changes, as illustrated by the identical proportion of fragmented DNA in irradiated and control ovaries on d 21 and 28 pn.
In situ analysis of follicular cytodifferentiation in immature females
The gene encoding aromatase, the enzyme implicated in the synthesis of estrogens by preovulatory follicles of cyclic females, has been shown to be transcribed at high levels in the infantile mouse ovary (32). Using in situ hybridization, we observed that transcripts for aromatase were present as soon as d 6 pn in control, but not in irradiated, ovaries (data not shown). Transcripts were detected in granulosa cells of the follicles located in the center of the control ovaries. On d 9 and 12 pn in both irradiated and control ovaries, all growing follicles expressed aromatase (Fig. 5, A and E, and data not shown), then the number of positive follicles progressively decreased to about half on d 15 pn (Fig. 5, B and F). From d 21 pn, rare antral follicles expressing aromatase were observed in both control and irradiated ovaries (Fig. 5, C and G). Double labeling with TUNEL showed that on d 21 pn, aromatase-positive follicles were healthy follicles (data not shown).
|
, D and H).
The first evidence for the differentiation of thecal cells was obtained on d 6 pn in control ovaries and on d 9 pn in irradiated ones. By histoenzymology, the presence of a few dispersed 3ß-HSD-positive cells was detected in the vicinity of growing follicles. From d 9 pn, 3ß-HSD positive cells progressively organized to form a true theca in control and irradiated ovaries (on d 12 pn; Fig. 5, I and M). These observations were completed by in situ hybridization of SF-1, the transcriptional regulator of genes encoding steroidogenic enzymes of the cytochrome P450 family (34), expressed by follicular cells in neonatal ovaries (35). SF-1 mRNAs were detected on d 6 pn in both irradiated and control ovaries (data not shown). On d 9 pn, its expression pattern became modified in control ovaries in which the theca of the most mature follicles started to highly express SF-1 mRNA (Fig. 5, J and K). In irradiated ovaries, SF-1 remained at similar levels in follicular and thecal cells (Fig. 5, N and O). On d 15 pn, patterns were similar between irradiated and control ovaries (Fig. 5, L and P).
Together the data indicate that expression profiles of all of the maturation markers were identical between irradiated and control ovaries, the only differences being the time of their appearance and the number of positive follicles. These observations were consistent with the occurrence of a time lag in follicular differentiation in irradiated ovaries and evidence that the differentiation of granulosa and thecal cells is not affected in growing follicles of irradiated ovaries.
Hormonal environment in immature females
Plasma levels of E2, T, inhibin B, and the gonadotropin FSH were assayed during the prepubertal period (Fig. 6). In control females, plasma E2 levels increased neonatally to reach a peak on d 12 pn and then decreased to low levels (Fig. 6A). This peak was absent in irradiated females (Fig. 6A). In contrast, no difference was observed in plasma levels of T and inhibin B between irradiated and controls during infantile and juvenile periods (Fig. 6, B and C). Plasma FSH levels were significantly higher in irradiated than in control females during the infantile period on d 9, 12, and 15 pn (Fig. 6D). On d 21 pn, the difference was no longer observed.
|
Fertility
The reproductive capacity in terms of fecundability and fecundity was examined (Fig. 7). The fertility of irradiated females was not different from that of controls at the beginning of the reproductive life span; at 2 months, 57% of irradiated females were pregnant vs. 60% for controls (Fig. 7A), and litter size was 8 ± 0.65 vs. 10 ± 1.09 pups/female (Fig. 7B). Fertility decreased progressively, leading to precocious ovarian failure in irradiated females. At 4 months, the number of pups per litter was 7 ± 1.09 in irradiated females vs. 11.17 ± 0.99 in controls (P < 0.05; Fig. 7B). At 6 months, the percentage of pregnant females fell drastically (30.7%) compared with controls (80%; Fig. 7A). The fecundity index, i.e. the number of pups per mated female, declined to 0.92 ± 0.63 at 6 months, whereas it remained high (9.2 ± 1.41) in the controls.
|
). The sterile ovaries were smaller (Fig. 8, A–D) and weighed less than controls [4.75 ± 0.34 (n = 32) vs. 14.41 ± 0.5 mg/100 g body weight (n = 25); P < 0.0001]. They displayed degenerated follicles (Fig. 8G, inset), rare healthy follicles (Fig. 8F, inset), and sometimes cysts (Fig. 8, C and F). The ovaries were full of fibrous tissue; small corpora lutea were still observed in some of them (compare Fig. 8, E–G). Sterile fibrous tissue was steroidogenic, as assessed by 3ß-HSD activity histochemistry detection (Fig. 8, H and I) in small clusters of cells (Fig. 8I). The hormonal environment was consistent with the ovarian senescence; the plasma level of inhibin B was depressed [31.94 ± 6.7 (n = 16) vs. 67.75 ± 6.49 (n = 8) pg/ml; P = 0.0027], whereas FSH levels were increased [145.39 ± 11.02 (n = 30) vs. 68.36 ± 10.41 ng/ml (n = 23); P < 0.0001] as were LH levels [5.48 ± 0.76 (n = 16) vs. 2.81 ± 0.7 ng/ml (n = 9); P < 0.001].
|
|
Discussion |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-irradiation of rat females modifies neither puberty onset nor fertility at the beginning of reproductive life, but provokes premature ovarian failure. In mice with a severely reduced complement of oocytes, such as in XO or Zfx mice, the shortened reproductive life span has been attributed to a deficiency in oocytes taking place before sexual maturation rather than to an accelerated loss of oocytes in adult cyclic females (7, 8, 36). Our analysis of the complete follicle population before puberty shows that all classes of follicles are not affected to the same extent. As previously reported by Beaumont (15, 37) in her study of rat ovaries on d 25 and 100 pn, the reduction in the number is considerably greater for primordial than for growing follicles. The present work points out that the massive depletion of primordial follicles exists from the beginning of follicular histogenesis in the neonatal period and that the very low number of primordial follicles remains statistically constant throughout the immature period. Thus, during this time period, primordial follicles are not recruited to enter the growing follicle pool, and, consequently, they do not play the role of follicular reserve. The number of primordial follicles may have reached such a critical value that the mechanisms controlling the recruitment of primordial follicles into the growing pool (1, 38) could not be effective in irradiated ovaries. Nevertheless, despite the absence of the primordial follicle stock, the number of ovulations is normal in irradiated females at 2 months and only slightly reduced at 4 months. As already advanced in cases of unilateral ovariectomy performed in rat on d 1 or 10 pn (39), the growing follicle pool, before its progressive exhaustion, can constitute the follicular reserve maintaining normal ovulation.
Our observations during immature period clearly show that the development of growing follicles is delayed in infantile irradiated ovaries. For example, the peak numbers of preantral and antral follicles were observed on d 21 pn in irradiated ovaries and on d 15 pn in controls. Such a delay in the differentiation of growing follicles was already reported in dysgenesic ovaries of XO mice on d 12 pn (7) as well as in dysgenesic ovaries of AMH transgenic mice on d 10 pn (40). In irradiated ovaries, this delay is observed from the very beginning of follicular differentiation, since the first preantral follicles present on d 3 pn in controls (41) are missing in irradiated ovaries. In the subsequent development, whatever the stage from d 3 pn until d 15 pn, the most mature follicles are absent or considerably fewer than in controls. This became quite evident when the expression of molecular markers of cellular differentiation was considered. For all of the markers examined (aromatase, activin ßAsubunit, 3ß-HSD, and SF-1), the onset of expression was delayed in irradiated ovaries.
In XO mice as well as in AMH transgenic mice the retardation in follicular maturation was attributed to a delay in the progression of the meiotic prophase during fetal life (7, 8, 40). In x-irradiated ovaries on d 15.5 pc, in addition to the depletion of germ cells, Beaumont (42) observed an accumulation of cells in the prophase of mitosis for the 24 h after irradiation as well as a great variety in meiotic cell stages in the following days. The high mitotic activity of germ cells was assumed to participate in a regeneration process that would disturb their later synchronization in the meiotic process (42). Our observations within the neonatal period have revealed another possible reason for the retardation of follicular development, namely, the prolonged resting status of primary follicles in irradiated ovaries. Although in control ovaries primary and secondary follicles were growing as early as d 3 pn, in irradiated ovaries they remained quiescent and only a few of them displayed granulosa cells resuming mitosis on d 6 pn. The reasons for this extended quiescent status are as yet a matter of conjecture. Regarding the consequences, could the delay in the development of follicles be the only effect of this extended quiescent status? It may also be possible that because of the delay in their growth, the first maturing follicles in irradiated ovaries are, indeed, not equivalent to the first growing follicles in control ovaries but, rather, to those growing in the subsequent waves.
In normal female rat, the first follicles, i.e. the first follicular wave, which differentiate in the core of the rat ovary (1) are destined to be depleted via atresia before fertile reproductive cycles are established (3). If this first follicular wave is missing in irradiated ovaries, the number of atretic follicles should be lower than that in control ovaries during immature period. This, indeed, was observed on d 15 and 18 pn using the TUNEL method and up to d 21 pn with follicular counts on histological sections. With the growth of the subsequent follicular waves occurring in irradiated ovaries as well as in controls, the proportion of atretic follicles returned progressively to an average range, as evidenced from d 21 pn with DNA fragmentation and on d 28 pn on histological sections. Taking into account the significant reduction of atresia during the infantile period, it may be concluded that the retardation in follicular maturation has indeed resulted in saving oocytes. Corroborating this interpretation is the fact that the number of healthy preantral and antral follicles is not significantly different from that in controls on d 21 pn despite the global depletion in oocytes.
The absence of the first maturing follicles in irradiated ovaries was further confirmed by in situ hybridization analysis of AMH expression that was used as a marker of healthy follicles (30, 31). On d 15 pn the most developed follicles in the inner part of the control ovaries, which are TUNEL positive, are negative or weakly stained for AMH mRNA, whereas in irradiated ovaries, the majority of follicles are TUNEL negative and strongly stained for AMH. This difference in the AMH mRNA staining intensity between irradiated and control ovaries was observed on d 9 pn, whereas no significant TUNEL labeling was detected. The decrease in AMH gene expression can thus be considered as early evidence for the commitment of follicles into the atretic process, taking place before DNA fragmentation. The present results show that in the normal rat ovary the atresia of the first follicular wave is initiated very early during the infantile period as soon as d 9 pn in preantral follicles.
Concomitantly with the rapid development of the first growing follicles during the infantile period, the levels of plasma FSH and E2 increase to peak between the second and third weeks of pn life (43, 44). In irradiated females, the levels of plasma FSH were higher than those in control females from d 9–15 pn, whereas the peak of E2 observed in controls on d 12 pn did not occur. In line with previous data reporting aromatase mRNA expression or enzymatic activity in the neonatal period (32, 45), we observed, by in situ hybridization the presence of aromatase transcripts in neonatal growing follicles. They were detected as early as d 6 pn in control ovaries and on d 9 pn in irradiated ones. From d 15 pn in both cases, the number of positive secondary and preantral follicles decreases, and from d 21 pn only healthy antral follicles continue to express the aromatase gene. This observation which is, to our knowledge, the first reported in situ localization of aromatase gene expression in follicles before the antral stage during the immature period emphasizes the parallelism between the time course of the expression pattern of aromatase and that of plasma E2 concentrations. Thus, in the control females, circulating E2 peaks on d 12 when the number of growing follicles expressing the aromatase gene is maximum, whereas in irradiated females the absence of the E2 peak can be correlated with the absence of the first follicular wave.
A similar reduced production of estrogen as well as a higher plasma FSH peak during the immature period were reported in rat females with dysgenesic ovaries resulting from fetal exposure to busulfan (46). The FSH rise was attributed to the abnormal steroid secretion of the ovaries. Nevertheless, as mentioned by the researchers (47), the role of gonadal factors in the control of gonadotropin release during the infantile period is controversial. It is usually admitted that functional negative feedbacks depending on E2 and inhibin are not operative before d 15 and 20 pn, respectively (48, 49, 50, 51), whereas androgen negative feedback would be efficient during the neonatal and infantile periods (48). On the other hand, considering the coincidence between the high concentrations of E2 and FSH on d 15 pn (44, 52), a positive feedback of estrogens on FSH secretion was suggested to occur during the infantile period (53). In irradiated females, whatever the stage from d 9–28 pn, T and inhibin levels are identical to those in controls, and plasma E2 does not display the positive correlation with plasma FSH observed in control females (52). Further investigation would be required to elucidate the mechanism(s) responsible for the higher secretion of FSH in irradiated females during the infantile period.
The decline in FSH concentrations occurring during the juvenile period was observed in irradiated as well as control females. It is likely that regulating mechanisms occurring throughout the juvenile period in control females take place similarly in irradiated females. In agreement with this assumption is the fact that puberty onset and fertility at the beginning of reproductive life are identical in both cases. Already in the experiments in which important modifications of follicular dynamics were induced in immature rats by the absence of gonadotropic stimulation, sexual maturation and cyclic function took place normally (54). A tight control of follicular dynamics is established shortly before the time of the first ovulation, independent of previous hormonal and follicular modifications (54). Indeed, in irradiated females the absence of the first follicular wave together with the changes in the levels of circulating FSH and E2 do not disturb the onset of sexual maturity. The consequences of the severe depletion of germ cells and the absence of the primordial follicle stock are visible later in the reproductive life span when the progressive exhaustion of follicles provokes infertility. Taken together, the present results emphasize the fact that the size of the primordial follicle stock is not the only parameter determining the reproductive life span. The size of the whole oocyte complement has to be taken into consideration as well as follicular dynamic modifications during the immature period.
|
Acknowledgments |
|---|
|
Footnotes |
|---|
1 S.M. is the recipient of a fellowship from Ministère de l’Education Nationale et de la Recherche Scientifique et Technique. 
Abbreviations: AMH, Anti-Mullerian hormone; E2, 17ß-estradiol; 3ß-HSD, 3ß-hydroxysteroid dehydrogenase; NGF, nerve growth factor; pc, postconception; PCNA, proliferating cell nuclear antigen; P i/c, percentage of the number in controls; pn, postnatal; SF-1, steroidogenic factor 1; T, testosterone; TUNEL, terminal deoxynucleotidyltransferase-mediated deoxyuridine 5'-triphospahte-fluorescein nick end labeling.
Received May 1, 2002.
Accepted for publication August 6, 2002.
|
References |
|---|
|
TopAbstractIntroductionMaterials and MethodsResultsDiscussionReferences |
|---|
-foetoprotein. J Endocrinol 81:199–208[Abstract]
This article has been cited by other articles:
 |
|
 |
|
  S. Mazaud Guittot, C. J. Guigon, N. Coudouel, and S. Magre Consequences of Fetal Irradiation on Follicle Histogenesis and Early Follicle Development in Rat Ovaries Biol Reprod, November 1, 2006; 75(5): 749 - 759. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Guigon and S. Magre |
) and irradiated (Irr; 
) female rats were weighed weekly from d 6 pn on, and values (in grams) are expressed as the mean ± SEM (the error bar in most cases is included in the symbol). The mean body weights of irradiated females were significantly different from controls at all observed ages, except on d 15 pn. n, Number of females at each time point.
) and irradiated (
) ovaries. The number of studied ovaries was eight, six, seven, and six for controls and eight, six, six, and four for irradiated ovaries on d 9, 15, 21, and 28 pn, respectively. The mean number of follicles per follicle category and developmental stage was compared by two-way ANOVA. Except for preantral and antral follicles on d 21 pn, in all other cases the number of follicles in irradiated ovaries was significantly different from that in controls (P < 0.05). Different letters (a–c for controls and x–z for irradiated ovaries) are used to indicate statistically significant differences between developmental stages within a follicle category (P < 0.05). The number of follicles in irradiated ovaries, expressed as the P i/c, is shown for each follicle category and each developmental stage. The values are joined by connecting lines.
