Oocyte number in newborn mice after prenatal octylphenol exposure

Reproductive Toxicology 17 (2003) 59–66

Oocyte number in newborn mice after prenatal octylphenol exposure Katrine Sonne-Hansen a , Majken Nielsen a,b , Anne Grete Byskov a,∗ a

Laboratory of Reproductive Biology, Juliane Marie Center for Children, Women and Reproduction, Rigshospitalet Section 5712, University Hospital of Copenhagen, Blegdamsvej 9, DK-2100 Copenhagen, Denmark b Department of Endocrinology, Faculty of Biology, Utrecht University, Utrecht, The Netherlands Received 21 March 2002; received in revised form 20 June 2002; accepted 11 August 2002

Abstract The aim of the present study was to investigate whether prenatal exposure to the suspected estrogenic compound 4-tert-octylphenol (OP), influences oocyte number in newborn female mice. In addition, effects on the percentage distribution of prefollicular, follicular, and atretic oocytes were investigated. Pregnant mice were subcutaneously injected with OP (1 or 250 mg/kg) or vehicle alone on embryonic day 11.5–16.5 (plug = embryonic day 0.5). As a positive control for estrogenic effects, a group of animals was injected with the synthetic estrogen diethylstilbestrol (DES, 100 ␮g/kg). Ovaries from the offspring were collected on the day of birth and a stereologic method, the optical fractionator, was used to estimate the number of prefollicular, follicular, and atretic oocytes in the ovaries. The total number of oocytes was calculated as the sum of the three subpopulation estimates. Neither OP nor DES exposure could be observed to affect the total number of oocytes or the percentage distribution of atretic, prefollicular, and follicular oocytes. Thus, prenatal OP exposure does not appear to cause a serious threat to fetal female germ cell proliferation and survival, or early follicle formation. © 2002 Elsevier Science Inc. All rights reserved. Keywords: Octylphenol; Diethylstilbestrol; Endocrine disrupters; Prenatal; Oocytes; Germ cell number; Follicle formation; Stereology

1. Introduction In recent decades, concern has arisen that synthetic chemicals contaminating the environment are compromising wildlife and human reproduction. Increased incidences of reproductive anomalies such as morphologically abnormal reproductive organs and reproductive dysfunction have been reported in various species. It has been hypothesized that compounds capable of disrupting normal endocrine function are responsible for the anomalies and such compounds are often referred to as endocrine disrupters. Evidence suggests that individuals are most sensitive to endocrine disrupters during fetal life. During this period, exposure appears to lead to permanent effects, in contrast to exposure during adulthood where the effects are often absent or reversible [1–3]. Endocrine disrupters are thought to act as hormone receptor agonists or antagonists, they may alter synthesis, metabolism, and/or transportation of endogenous hormones, and/or modify hormone receptor levels [4]. Metabolites of alkylphenol polyethoxylates are suspected to possess endocrine disrupting properties. Alkylphenol polyethoxylates are non-ionic surfactants widely used in in∗ Corresponding author. Tel.: +45-354-55820; fax: +45-354-55824. E-mail address: [email protected] (A.G. Byskov).

dustrial, agricultural and household applications, including detergents, paint, pesticides, plastic for food packaging, and toiletries [5,6]. During treatment in sewage plants, or upon discharge into the environment, alkylphenol polyethoxylates become microbially degraded into rather persistent and lipophilic metabolites [6,7]. One of the alkylphenol polyethoxylate metabolites, 4-tert-octylphenol (OP), has been observed to be estrogenic in both in vitro [8,9] and in vivo [10–12] screening assays. Indeed, OP has been suggested to modulate the expression of estrogen-sensitive genes in a pattern similar to ethinylestradiol [12]. However, as OP does not only show binding affinity for the estrogen receptors alpha (ER␣) and beta (ER␤) [13], but also to the androgen receptor [14], the endocrine disrupting potential of OP may not be limited to estrogenic properties. Studies have been published indicating that developmental exposure to OP may affect germ cell dynamics and/or function, for example, by inducing ovarian follicular cysts, absence of normal corpora lutea, and lack of ovulation in rats [15,16], and accelerating ovarian follicle turnover in sheep [17]. However, other studies have failed to detect effects on ovarian morphology and fertility in rats after OP exposure [5,18]. An in vitro study has shown that the number of germ cells was markedly decreased in human fetal testes cultured in presence of OP, whereas the number

0890-6238/02/$ – see front matter © 2002 Elsevier Science Inc. All rights reserved. PII: S 0 8 9 0 - 6 2 3 8 ( 0 2 ) 0 0 0 7 9 - 5

60

K. Sonne-Hansen et al. / Reproductive Toxicology 17 (2003) 59–66

of germ cells in fetal ovaries was unaffected [19]. Thus, studies on the effects of developmental OP exposure are contradictory. In mammals, the early female germ cells, the oogonia, proliferate during ovarian development. As they enter meiosis in fetal or early neonatal life, they develop into oocytes. The oocytes pass through the stages of the first meiotic prophase, leptotene, zygotene, pachytene, and diplotene. When the oocytes reach the diplotene stage they are enclosed in follicles, a process necessary for oocyte survival [20]. By entering meiosis, the cells loose their ability to proliferate, preventing renewal of the naturally decreasing pool of oocytes, or replacement of accidentally lost oocytes. Exposure to a compound that affects oogonia proliferation, survival of oogonia and oocytes, or follicle formation may therefore be critical for female reproduction. The present study investigated the effects of prenatal OP exposure on the total number of oocytes and the percentage distribution of prefollicular, follicular, and atretic oocytes in newborn mice. To investigate whether an effect of OP could be due to estrogenic properties, diethylstilbestrol (DES), a potent synthetic estrogen with well-documented effects on human and rodent reproduction, was used as a positive control. If OP mimicked the effect of DES, the effect on the oocytes might be attributable to estrogenic effects of OP. The optical fractionator [21–23] was used to estimate the number of prefollicular, follicular, and atretic oocytes, and the total number of oocytes was estimated from the sum of these subpopulations.

2. Materials and methods 2.1. Chemicals 4-tert-Octylphenol (OP, 97% purity) was purchased from Aldrich Chemical Company, Inc. (Sigma–Aldrich, Vallensbæk Strand, Denmark). Diethylstilbestrol was purchased from Sigma Chemical Co. (Sigma–Aldrich, Vallensbæk Strand, Denmark). Both compounds were dissolved in sterile peanut oil (H:S pharmacy, Frederiksberg, Denmark). 2.2. Animals and exposure protocol Adult mice, strain NMRI (Bomholtgaard, Ry, Denmark), were kept under controlled conditions with a photoperiod of 12 h light and 12 h dark. The animals had free access to food and water. Male and female mice were caged together overnight and the females were considered pregnant if a vaginal plug was observed the following morning. The day of appearance of the plug was defined as embryonic day (E) 0.5. Term for delivery was E19.5. Pregnant mice were randomized into four different exposure groups: a control group exposed to vehicle (peanut oil), a group exposed to 250 mg OP/kg maternal body weight (OP250 group), a

group exposed to 1 mg OP/kg maternal body weight (OP1 group) and a group exposed to 100 ␮g DES/kg maternal body weight (DES group). The pregnant mice were treated once daily with a subcutaneous injection (s.c.) from E11.5 to E16.5. The DES dose was selected based on a mouse model of developmental DES exposure [24] and the mouse uterothrophic assay [25]. To assure effects, if any, of prenatal DES exposure in the present study, the highest dose used in the two models was selected (100 ␮g/kg). The low OP dose was selected based on a study of prenatal s.c. exposure of female pigs, where 1 mg/kg OP was observed to affect the endocrine system [26]. The high OP dose was included based on a study of direct exposure of newborn rats, where 100 mg/kg was the lowest dose causing persistent estrous and effects on ovarian follicle dynamics and function [15]. Because the animals in the present study were exposed indirectly, a higher dose was selected (250 mg/kg) than in the neonatal study. 2.3. Tissue preparation On the day of birth, the gonads and adherent sex ducts were removed from the newborn female mice and immediately fixed in Bouins solution for 1.5 h at room temperature. The tissues were rinsed in 70% ethanol, dehydrated in increasing concentrations of ethanol and xylene and embedded in paraffin. Serial sections of 35 ␮m were cut on a rotary microtome (RM2155, Leica, Glostrup, Denmark). Before cutting of each section, the surface of the paraffin block was moistened with wet filter paper to avoid crackling of the section. The sections were placed on silica coated glass slides (SuperFrost Plus, Prohosp, Værløse, Denmark), dried overnight at 60 ◦ C, and processed for routine hematoxylin staining. Following staining, the tissues were dehydrated in increasing concentrations of ethanol and xylene and cover slips were mounted. One ovary from each of nine newborn pubs in the control group and one ovary from each of eight newborn pubs in each of the OP1, OP250, and DES group were prepared for examination. The individuals came from eight different litters in the control group, four litters in the OP250 and OP1 group, and six litters in the DES group. 2.4. Classification of oocyte subpopulations selected for stereological counting All germ cells had entered meiosis and were thus classified as oocytes. For counting, the oocytes were divided into three subpopulations: prefollicular oocytes, follicular oocytes, and atretic oocytes. As follicle formation had just begun, no distinction was made between early or late follicle stages. All oocytes were distinguishable from somatic cells by a characteristically spherical nucleus and a considerably larger nucleus and cell size. The following classification was used for recognition of each oocyte subpopulation during stereologic counting.

K. Sonne-Hansen et al. / Reproductive Toxicology 17 (2003) 59–66

2.4.1. Prefollicular oocytes Oocytes in the first meiotic prophase that had not yet become enclosed in follicles. Prophase oocytes could be recognized by a stage specific chromatin pattern. 2.4.2. Follicular oocytes Oocytes in the diplotene stage of the first meiotic prophase, surrounded by at least two flattened granulosa cells enclosing the oocyte in a follicle. The diplotene oocytes were characterized by diffuse nuclear chromatin condensation, a smooth nuclear membrane and one or more visible nucleoli. 2.4.3. Atretic oocytes An oocyte containing a pycnotic nucleus and/or a highly irregularly shaped nucleus. Atretic oocytes were only counted if they could be distinguished from atretic somatic cells, based on their relative larger size and the presence of remnants of condensed meiotic chromosomes. No distinction was made between atretic prefollicular oocytes and atretic follicular oocytes. 2.5. Stereologic procedures The optical fractionator technique [21–23] was used to estimate the number of each of the classified oocyte subpopulations per ovary. Briefly, the principle of the optical fractionator technique is to count cells in a known fraction of the volume of an organ. Knowing the fraction of the organ in which the cells have been counted, the total number of cells for the whole organ can be estimated. The technique is based on mathematical and statistical principles and facilitates achievement of unbiased estimates of the total number of selected cell types in an organ [21,23]. Counting was performed using a computer assisted microscope (Olympus BH-2, Olympus, Albertslund, Denmark) equipped with a motorized specimen stage for stepwise displacement in the x- and y-axis, an electronic microcator for measuring movement in the z-axis of the thick sections, and a video camera connected to a computer monitor. The system was controlled by CAST-GRID software, version 1.60 (Olympus, Albertslund, Denmark). The sections were examined using a 100× oil-immersion objective with a numerical aperture of 1.4. A sampling scheme was determined to facilitate counting of approximately 100–200 cells of each oocyte subpopulation in each ovary. Due to the small size of the ovaries every section was examined. Optical dissector counting rules were used for counting the cells [27,28]. The counting criterion was when the first chromatin of the oocyte nucleus came into focus within the counting frame. Number estimates for each oocyte subpopulation were calculated according to West et al. [21,23]. The total number of oocytes in an ovary was calculated as the sum of the number estimates obtained for each oocyte subpopulation in the ovary. The percentage of oocytes in each subpopulation was calculated from the estimated num-

61

ber of oocytes in a given subpopulation and the estimated total number of oocytes. The same person examined all tissues. The identification of the ovaries was blinded by a colleague prior to counting to ensure that assumptions regarding exposure effects did not affect counting. 2.6. Precision of the number estimates and group variance Using the optical fractionator makes it possible to calculate the precision of the obtained number estimate and to conclude whether the observed variance in a group is due to biologic variance or variation in the individual number estimates. The precision of the individual number estimates is given by the coefficient of error CE(N), which was calculated for all counted oocyte subpopulations in each individual according to West et al. [21,23] and Gundersen et al. [29]. CE(N) for the total number of oocytes (prefollicular, follicular, and atretic) in each individual was calculated from the CE(N) of each oocyte subpopulation, taking into account the number of cells each subpopulation contributed with to the total number (Nyengaard, personal communication). Subsequently, in each exposure group, the group mean CE(N) was calculated for all oocyte populations. The group mean CE(N) expresses the variation due to sampling. The coefficient of variation between individuals within an exposure group, CV, was calculated for all oocyte populations according to West et al. [21,23]. The CV is a product of the true biologic variance, ICV, and the variance due to sampling (i.e. group mean CE(N)) according to the following equation: CV2 = ICV2 + group mean CE(N)2 [23]. Therefore, when the group mean CE(N)2 is less than half of CV2 , the true biologic variance (i.e. ICV2 ), makes the major contribution to CV2 and the variance within a group can mainly be ascribed to biologic variance [23]. The relative source of variation to CV2 can therefore be examined by calculating the ratio between the group mean CE(N)2 and CV2 . When the ratio is less than 0.5, the group mean CE(N)2 is less than half of CV2 . The ratio between the group mean CE(N)2 and CV2 were calculated for all number estimates. 2.7. Statistical analyses The total number of oocytes in the four exposure groups was compared using a one-way analysis of variance (ANOVA). The percentages of prefollicular, follicular, and atretic oocytes were transformed (arcsine(fraction)1/2 ) and a one-way ANOVA was used to compare the fraction of oocytes in each subpopulation among the four exposure groups. Following the ANOVA, a multiple comparison test using the Bonferroni Correction was performed. The power of the ANOVA test to detect differences was calculated. The between-litter variation was compared to the within-litter variation using ANOVA. The SAS® software package (version 8.2; SAS Institute, Cary, NC) was used to perform the

62

K. Sonne-Hansen et al. / Reproductive Toxicology 17 (2003) 59–66

Table 1 Estimated total number of oocytes in ovaries from newborn mice prenatally exposed to vehicle alone, OP, or DES Treatment

Mean ± S.D. Range

Estimated total number of oocytes Control

OP1

OP250

DES

6812 ± 1900 3066–9632

6009 ± 1897 4164–9713

6061 ± 1614 4181–8682

6393 ± 1361 4476–8131

The control group was exposed to vehicle alone (peanut oil), the OP1 group to 1 mg OP/kg maternal body weight, the OP250 group to 250 mg OP/kg maternal body weight, and the DES group to 100 ␮g DES/kg maternal body weight. For the control group, n = 9; for the OP1, OP250, and DES group, n = 8.

statistical analyses. The level of significance was set to 0.05 for all tests. 3. Results 3.1. Ovarian morphology The general morphology was similar for all examined ovaries and it was not possible to recognize to which exposure group an ovary belonged. All germ cells had reached the oocyte stage. The majority of the oocytes were in the pachytene or diplotene stage. Leptotene oocytes were not present, whereas zygotene oocytes were observed on very

few occasions. Enclosure of diplotene oocytes in follicles had begun. Oocytes were present throughout the ovaries, with the highest density in the outer cortex. Developing follicles were primarily seen in the medulla and the inner cortex. Fully developed primordial follicles and small growing follicles were observed occasionally. 3.2. Number of oocytes and distribution of oocytes in subpopulations As some of the examined individuals in each exposure group were littermates, within- versus between-litter variation in germ cell number and percentage distribution of

Fig. 1. Percentage distribution of prefollicular, follicular, and atretic oocytes in ovaries from newborn mice exposed prenatally to vehicle alone (control group), 1 or 250 mg octylphenol/kg maternal body weight (OP1 and OP250 groups), or 100 ␮g diethylstilbestrol/kg maternal body weight (DES group). Values from all examined individuals are presented (n = 9 for the control group, n = 8 for the OP1, OP250, and DES group). Where fewer points than n are observable, identical values resulted in point overlap. The horizontal bars represent the group means and the vertical bars the standard deviation.

K. Sonne-Hansen et al. / Reproductive Toxicology 17 (2003) 59–66

oocyte populations was analyzed. Generally, a large individual variation was observed for all examined parameters in all exposure groups. The analysis showed that the individual variation between littermates was as great as between-litters (total number of oocytes: P = 0.26; percentage of prefollicular oocytes: P = 0.33; follicular oocytes: P = 0.54; atretic oocytes: P = 0.21; data not shown). Correction for within- versus between-litter variation was therefore not performed. The power of the ANOVA test to detect differences with the present study design was calculated. The change that could be detected with 80% power was a decrease or increase of approximately one-third in the total number of oocytes. In regard of the percentage distribution of oocyte subpopulations, the change that could be detected was approximately 16% points for the prefollicular oocytes and 10-percentage points for the follicular and atretic oocytes. The estimated total number of oocytes is presented in Table 1. Individual variation with up to two- to three-fold difference within an exposure group was observed. The variation was comparable for all exposure groups and no significant difference could be observed in the total number of oocytes between groups (P = 0.75). The percentage distribution of prefollicular, follicular, and atretic oocytes is presented in Fig. 1. Although follicle formation had begun in all ovaries, most oocytes were not yet enclosed in follicles. A large individual variance also existed in the percentage distribution of prefollicular and follicular oocytes, particularly in the OP250 group. Two animals in this group had a relative high percentage of follicular oocytes. These two animals came from two different litters and their examined littermates had percentages of follicular oocytes equal to or lower than the group average. Here, a large individual variance was clearly evident both between- and within-litters. Comparison of the distribution of prefollicular and follicular oocytes in the four exposure groups showed no statistical significant difference in percentage of prefollicular oocytes (P = 0.07) or percentage of follicular oocytes (P = 0.08). The relatively low P-values could be interpreted as possibly indicating some effect of OP or DES exposure. However, the multiple comparison tests showed that the low P-values were due to fewer prefollicular oocytes and more follicular oocytes in the OP250 group compared to the DES group. No difference could be detected between the control animals and the OP or DES exposed animals. The fraction of atretic oocytes was constant in all exposure groups and no statistical significant difference was observed among the groups (P = 0.81). 3.3. Precision of the number estimates The ratio group mean CE(N)2 /CV2 was calculated for all oocyte subpopulations, as well as for the total number of oocytes in all exposure groups (Table 2). The ratio was well below 0.5 in all instances, implying that the variation due to sampling was low and that the main variation within

63

Table 2 Precision of the oocyte number estimates obtained with the optical fractionator Oocyte population

Prefollicular Follicular Atretic Total

Group mean CE(N)2 /CV2 Control

OP1

OP250

DES

0.024 0.091 0.075 0.011

0.013 0.073 0.107 0.009

0.015 0.029 0.103 0.009

0.042 0.080 0.083 0.019

When the ratio group mean CE(N)2 /CV2 is below 0.5, the variation within a group can mainly be ascribed to biologic variation. The control group was exposed to vehicle alone (peanut oil), the OP1 group to 1 mg OP/kg maternal body weight, the OP250 group to 250 mg OP/kg maternal body weight, and the DES group to 100 ␮g DES/kg maternal body weight. For the control group, n = 9; for the OP1, OP250, and DES group, n = 8.

a treatment group could be ascribed to biological variation (see Section 2).

4. Discussion Large variation exists in the reported total number of oocytes in newborn mice, ranging from 3000 to 60,000 oocytes per ovary [30–33]. In the present study, the mean number of oocytes in the control group was 6812 per ovary. Differences among mouse strains may to some degree explain the conflicting results among published studies, but also variation in the methods used to obtain the estimates may contribute to the variation. In the present study, the optical fractionator was used to obtain the number estimates because this method leads to unbiased estimates, that is, estimates that approach the true number of cells in an organ with increased sample size [21]. In addition, the precision of the number estimates can be calculated within this model. In the present study, the variation due to sampling was negligible, which means that the precision of the estimate was high. Thus, the large individual variation within an exposure group was mainly due to biologic variation and the estimates are considered close to the true number of oocytes in the ovaries. The only way to decrease the large variation within an exposure group would be to include more individuals in each group. In general, stereologic investigations have shown that a large individual variation in oocyte number exists in various species, including humans [34], primates [35], sheep [36], opossums [37], and pigs (Nielsen, in preparation). No effect could be observed in the total oocyte number or percentage distribution of prefollicular, follicular, and atretic oocytes after prenatal exposure to OP or the positive control DES. The lack of effect was not due to inactivity of the OP and DES batches, as in a parallel study, they were found to be biologically active in an in vitro estrogenicity assay [38]. Neither does failure of placenta transfer explain the absence of effects since in the same parallel study, which

64

K. Sonne-Hansen et al. / Reproductive Toxicology 17 (2003) 59–66

used tissues from the same litters as in the present study, uterine ER␣ expression was changed after DES exposure [38]. However, OP exposure failed to induce any changes in the uterine ER␣ expression. Nevertheless, other studies have shown effects on the reproductive system after prenatal OP exposure in rat [39–41], sheep [17,42], and pig [26], indicating that maternal OP exposure during gestation does result in fetal exposure in various species. The large individual variance has to be taken into consideration when evaluating the results. The power of the present study design allowed for detection of a difference of one-third in the total number of oocytes and for a difference in percentage distribution of oocyte subpopulations close to the difference observed between the OP250 and DES group with regard to prefollicular and follicular oocytes. Whether differences smaller than those detectable in the present study are of biologic interest can only be speculated, as the level of change in germ cell number and follicle dynamics that would lead to an effect on reproductive capacity is unknown. However, the biologic variance in the control group is large, indicating that normal fertility can be sustained despite large variation in both oocyte number and time window for follicle formation. This finding may indicate that only dramatic changes in these parameters affect fertility of the individual animal, and that the present study is sufficiently sensitive. Still, a more extensive study of prenatally exposed adult animals is needed to evaluate whether fertility is compromised. Indeed, it has been suggested that some effects of prenatal exposure may only become evident at onset of puberty or in adult reproductive life [1–3], possibly due to amplification of the prenatal exposure by secondary exposure to hormone compounds, either endogenous hormones as their production increases at puberty, or additional exposure to endocrine disrupters [1]. The lack of effect on the total number of oocytes and the percentage of atretic oocytes is in agreement with an in vitro study in which fetal human ovaries were cultured in the presence of OP, showing no effect on the number of germ cells [19]. In addition, in a comprehensive study of adult rats exposed to OP throughout life through their mother’s diet and their own diet, no change in the number of follicle-enclosed germ cells was observed [18]. In the present study, the total number of oocytes and percentage of atretic oocytes was also unaffected by exposure to the positive control DES. The present results could imply that the proliferation and survival of fetal female germ cells are independent of the endocrine milieu, or possibly that the endocrine milieu was not disturbed to a degree to which proliferation or survival of the germ cells was affected. The lack of effect on the percentage distribution of prefollicular and follicular oocytes after prenatal OP or DES exposure could indicate that the early follicle formation was undisturbed. However, as mentioned, effects smaller than 10–16% points cannot be detected with the present study design. Indeed, it is possible that relatively more follicles

and fewer prefollicular oocytes occurred in the OP250 group than in the DES group, which may indicate that OP and DES have small, opposite directed effects on follicle formation. This finding may indicate that the mechanisms of action of the two compounds are different. However, as both OP [14] and DES [43] may have other actions in addition to being estrogenic, the mechanisms underlying the possible effect of the compounds are not straightforward and cannot be determined from the present study. Only a few other studies have dealt with follicle formation or function after developmental exposure to OP. In a study of fetal sheep exposed intravenously or through s.c. injections to OP or DES, accelerated growth of resting primordial follicles was observed [17]. This finding was speculated to lead to accelerated follicle turnover and precocious depletion of the ovarian follicle pool. In a study of rats exposed neonatally to OP through s.c. injections, ovarian follicular cysts and lack of ovulation were observed [15], indicating a critical effect on follicle development and function. However, in another study of neonatally s.c. injected rats [5], as well as the mentioned comprehensive study of rats exposed continuously through the diet [18], no changes on ovarian morphology or fertility were observed. Thus, considerable discrepancies in the observed ovarian effects of OP exposure exist. These contradictory findings could be results of methodologic differences in administration route, period or length, in species or animal strain examined, or in time of examination. Regarding exposure route, persistent estrous and increased uterine weight has been reported in rats after s.c. injections of OP, whereas no or low-order effects was observed with concomitant oral exposure of another group of rats [16,44]. This discrepancy may be due to differences in the toxicokinetics, as variation has been reported according to exposure route of OP [45,46]. Regarding effects of DES exposure, a study of prenatally s.c. injected mice has shown that follicle formation and development was highly accelerated in newborn animals [47]. Pregnant mice were only exposed to a single injection on E15 with a DES dose 10 times lower than in the present study, making the discrepancy with the present study difficult to explain. However, difference in mouse strain sensitivity may play a role. Further, the classification of oocytes differed markedly from that in the present study, making direct comparison difficult. Exposure to both OP and DES has been observed to result in formation of polyovular follicles [5,48]. In the case of OP exposure, the polyovular follicles could be observed only at the time of immaturity, and fertility of the adults was unaffected. Thus, the polyovular follicles must have degenerated before maturity and despite this decrease in the oocyte pool, a sufficient number of healthy follicles were present to sustain normal fertility. In the present study, follicle formation had just begun and whether polyovular follicles would form is unknown. Culturing of human fetal gonads has indicated that sensitivity to OP may be different for male and female germ cells. It was found that the number of female germ cells was

K. Sonne-Hansen et al. / Reproductive Toxicology 17 (2003) 59–66

unaffected by OP exposure, whereas the number of male germ cells was dramatically decreased [19]. However, no effect was observed on germ cell number in newborn male mice exposed prenatally to OP (Sonne-Hansen, unpublished results) or in adult male rats after neonatal OP exposure [49]. The reason for the discrepancies between studies is unclear, but in vivo and in vitro exposure could well result in different effects. Regarding DES, the number of male germ cells has been shown to be significantly decreased in adult rats after neonatal exposure [50,51]. Thus, an indication exists that, in contrast to the female, the male endocrine system may in some way, directly or indirectly, influence early germ cell proliferation and/or survival, or possibly, the male endocrine system may be more sensitive to endocrine disrupters, resulting in an imbalance capable of affecting the male germ cells. In conclusion, under the present experimental conditions, prenatal OP exposure was not observed to affect the total number of oocytes or the percentage distribution of prefollicular, follicular, or atretic oocytes. These findings may indicate that prenatal OP exposure does not constitute a threat to early female germ cell proliferation and survival, or the processes important for early follicle formation. However, it should be kept in mind that the study does not simulate the complexity that may influence susceptibility to environmental OP exposure in humans or wildlife. Nevertheless, even though the level of environmental exposure has not been established, doses are probably not as high or consistent as in the present study. In addition, route of environmental exposure may primarily be oral, which has been indicated to posses less risk than exposure by s.c. injection [16,44]. In the present study, neither did DES exposure affect any of the investigated parameters. The lack of effects after both OP and DES exposure could indicate that mammalian female germ cells are insensitive to estrogen exposure during fetal life. However, as different estrogenic compounds have been shown to induce different changes in gene expression [12], a general insensitivity of female germ cells to estrogenic compounds cannot be predicted from the present study. Further, it cannot be ruled out that effects could be identified in older animals.

Acknowledgments The study was supported by the Danish Environmental Research Programme: Danish Center for Environmental Estrogen Research, the Danish Medical Research Council No. 9700832, and the University of Copenhagen (scholarship to K. Sonne-Hansen). Ib Christensen, the Finsen Laboratory, Copenhagen, is gratefully acknowledged for performance of the statistical analyses and the graphics. Jens Nyengaard, Stereological Research Laboratory, Århus, and Hans Peter Larsen, Laboratory of Reproductive Biology, Copenhagen, is thanked for indispensable help with application of the stereological method.

65

References [1] Bern HA. The fragile fetus. In: Colborn T, Clement C, editors. Chemically-induced alterations in sexual and functional development: the wildlife/human connection. Princeton, NJ: Princeton Scientific Publishing Co.; 1992, p. 9–15. [2] Guillette Jr LJ, Crain DA, Rooney AA, Pickford DB. Organization versus activation: the role of endocrine-disrupting contaminants (EDCs) during embryonic development in wildlife. Environ Health Perspect 1995;103(Suppl 7):157–64. [3] Dencker L, Eriksson P. Susceptibility in utero and upon neonatal exposure. Food Addit Contam 1998;15(Suppl):37–43. [4] Sonnenschein C, Soto AM. An updated review of environmental estrogen and androgen mimics and antagonists. J Steroid Biochem Mol Biol 1998;65(1–6):143–50. [5] Nagao T, Yoshimura S, Saito Y, Nakagomi M, Usumi K, Ono H. Reproductive effects in male and female rats from neonatal exposure to p-octylphenol. Reprod Toxicol 2001;15(6):683–92. [6] Nimrod AC, Benson WH. Environmental estrogenic effects of alkylphenol ethoxylates. Crit Rev Toxicol 1996;26(3):335–64. [7] Isobe T, Nishiyama H, Nakashima A, Takada H. Distribution and behavior of nonylphenol, octylphenol, and nonylphenol monoethoxylate in Tokyo metropolitan area: their association with aquatic particles and sedimentary distributions. Environ Sci Technol 2001;35(6):1041–9. [8] White R, Jobling S, Hoare SA, Sumpter JP, Parker MG. Environmentally persistent alkylphenolic compounds are estrogenic. Endocrinology 1994;135(1):175–82. [9] Abraham EJ, Frawley LS. Octylphenol (OP), an environmental estrogen, stimulates prolactin (PRL) gene expression. Life Sci 1997;60(17):1457–65. [10] Bicknell RJ, Herbison AE, Sumpter JP. Oestrogenic activity of an environmentally persistent alkylphenol in the reproductive tract but not the brain of rodents. J Steroid Biochem Mol Biol 1995;54(1/2): 7–9. [11] Cummings AM, Laws SC. Assessment of estrogenicity by using the delayed implanting rat model and examples. Reprod Toxicol 2000;14(2):111–7. [12] Diel P, Schulz T, Smolnikar K, Strunck E, Vollmer G, Michna H. Ability of xeno- and phytoestrogens to modulate expression of estrogen-sensitive genes in rat uterus: estrogenicity profiles and uterotropic activity. J Steroid Biochem Mol Biol 2000;73(1/2):1–10. [13] Kuiper GG, Lemmen JG, Carlsson B, et al. Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor beta. Endocrinology 1998;139(10):4252–63. [14] Gray Jr LE, Kelce WR. Latent effects of pesticides and toxic substances on sexual differentiation of rodents. Toxicol Ind Health 1996;12(3/4):515–31. [15] Katsuda S, Yoshida M, Watanabe G, Taya K, Maekawa A. Irreversible effects of neonatal exposure to p-tert-octylphenol on the reproductive tract in female rats. Toxicol Appl Pharmacol 2000;165(3):217–26. [16] Yoshida M, Katsuda S, Ando J, Kuroda H, Takahashi M, Maekawa A. Subcutaneous treatment of p-tert-octylphenol exerts estrogenic activity on the female reproductive tract in normal cycling rats of two different strains. Toxicol Lett 2000;116(1/2):89–101. [17] Picton H. Effects of endocrine disrupting compounds on follicular growth and development in sheep. In: Proceedings of the conference on the effect of endocrine disrupting compounds in animal feed on reproduction and health in farm animals. Euroconference 1998, Wageningen, The Netherlands. [18] Tyl RW, Myers CB, Marr MC, et al. Two-generation reproduction study with para-tert-octylphenol in rats. Regul Toxicol Pharmacol 1999;30(2 Pt 1):81–95. [19] Bendsen E, Laursen S, Olesen C, Westergaard L, Andersen C, Byskov AG. Effect of 4-octylphenol on germ cell number in cultured human fetal gonads. Hum Reprod 2001;16(2):236–43.

66

K. Sonne-Hansen et al. / Reproductive Toxicology 17 (2003) 59–66

[20] Byskov AG, Høyer PE. Embryology of mammalian gonads and ducts. 1994;2(9):487–540. [21] West MJ, Slomianka L, Gundersen HJ. Unbiased stereological estimation of the total number of neurons in thesubdivisions of the rat hippocampus using the optical fractionator. Anat Rec 1991;231(4):482–97. [22] Wreford NG. Theory and practice of stereological techniques applied to the estimation of cell number and nuclear volume in the testis. Microsc Res Tech 1995;32(5):423–36. [23] West MJ, Ostergaard K, Andreassen OA, Finsen B. Estimation of the number of somatostatin neurons in the striatum: an in situ hybridization study using the optical fractionator method. J Comp Neurol 1996;370(1):11–22. [24] Newbold R. Cellular and molecular effects of developmental exposure to diethylstilbestrol: implications for other environmental estrogens. Environ Health Perspect 1995;103(Suppl 7):83–7. [25] Shelby MD, Newbold R, Tully DB, Chae K, Davis VL. Assessing environmental chemicals for estrogenicity using a combination of in vitro and in vivo assays. Environ Health Perspect 1996;104:1296– 300. [26] Bogh IB, Christensen P, Dantzer V, et al. Endocrine disrupting compounds: effect of octylphenol on reproduction over three generations. Theriogenology 2001;55(1):131–50. [27] Gundersen HJ. Notes on the estimation of the numerical density of arbitrary profiles: the edge effect. J Microsc 1977;111:219–23. [28] Gundersen HJ, Bagger P, Bendtsen TF, et al. The new stereological tools: disector, fractionator, nucleator and point sampled intercepts and their use in pathological research and diagnosis. APMIS 1988;96(10):857–81. [29] Gundersen HJ, Jensen EB, Kieu K, Nielsen J. The efficiency of systematic sampling in stereology-reconsidered. J Microsc 1999;193(Pt 3):199–211. [30] Borum K. Oogenesis in the mouse. Exp Cell Res 1961;24:495–507. [31] Burgoyne PS, Baker TG. Perinatal oocyte loss in XO mice and its implications for the aetiology of gonadal dysgenesis in XO women. J Reprod Fertil 1985;75(2):633–45. [32] Morita Y, Tilly JL. Segregation of retinoic acid effects on fetal ovarian germ cell mitosis versus apoptosis by requirement for new macromolecular synthesis. Endocrinology 1999;140(6):2696–703. [33] Pepling ME, Spradling AC. Mouse ovarian germ cell cysts undergo programmed breakdown to form primordial follicles. Dev Biol 2001;234(2):339–51. [34] Forabosco A, Sforza C, De Pol A, Vizzotto L, Marzona L, Ferrario VF. Morphometric study of the human neonatal ovary. Anat Rec 1991;231(2):201–8. [35] Miller PB, Charleston JS, Battaglia DE, Klein NA, Soules MR. An accurate, simple method for unbiased determination of primordial follicle number in the primate ovary. Biol Reprod 1997;56(4): 909–15. [36] Smith P, WS O, Hudson NL, et al. Effects of the Booroola gene (FecB) on body weight, ovarian development and hormone concentrations during fetal life. J Reprod Fertil 1993;98(1):41–54.

[37] Shackell GH, Norman NG, McLeod BJ, Hurst PR. A morphometric study of early ovarian development in pouch young of the brushtail possum (Trichosurus vulpecula). Anat Rec 1996;246(2):224–30. [38] Nielsen M, Hoyer PE, Lemmen JG, van der BB, Byskov AG. Octylphenol does not mimic diethylstilbestrol-induced oestrogen receptor-alpha expression in the newborn mouse uterine epithelium after prenatal exposure. J Endocrinol 2000;167(1):29–37. [39] Majdic G, Sharpe RM, O’Shaughnessy PJ, Saunders PT. Expression of cytochrome P450 17alpha-hydroxylase/C17-20 lyase in the fetal rat testis is reduced by maternal exposure to exogenous estrogens. Endocrinology 1996;137(3):1063–70. [40] Saunders PT, Majdic G, Parte P, et al. Fetal and perinatal influence of xenoestrogens on testis gene expression. Adv Exp Med Biol 1997;424:99–110. [41] Majdic G, Sharpe RM, Saunders PT. Maternal oestrogen/xenoestrogen exposure alters expression of steroidogenic factor-1 (SF-1/ Ad4BP) in the fetal rat testis. Mol Cell Endocrinol 1997;127(1): 91–8. [42] Sweeney T, Nicol L, Roche JF, Brooks AN. Maternal exposure to octylphenol suppresses ovine fetal follicle-stimulating hormone secretion, testis size, and sertoli cell number. Endocrinology 2000;141(7):2667–73. [43] Sohoni P, Sumpter JP. Several environmental oestrogens are also anti-androgens. J Endocrinol 1998;158(3):327–39. [44] Laws SC, Carey SA, Ferrell JM, Bodman GJ, Cooper RL. Estrogenic activity of octylphenol, nonylphenol, bisphenol A and methoxychlor in rats. Toxicol Sci 2000;54(1):154–67. [45] Certa H, Fedtke N, Wiegand HJ, Muller AM, Bolt HM. Toxicokinetics of p-tert-octylphenol in male Wistar rats. Arch Toxicol 1996;71(1/2):112–22. [46] Upmeier A, Degen GH, Schuhmacher US, Certa H, Bolt HM. Toxicokinetics of p-tert-octylphenol in female DA/Han rats after single i.v. and oral application. Arch Toxicol 1999;73(4/5):217–22. [47] Wordinger RJ, Derrenbacker J. In utero exposure of mice to diethylstilbestrol alters neonatal ovarian follicle growth and development. Acta Anat (Basel) 1989;134(4):312–8. [48] Iguchi T, Takasugi N. Polyovular follicles in the ovary of immature mice exposed prenatally to diethylstilbestrol. Anat Embryol (Berl) 1986;175(1):53–5. [49] Atanassova N, McKinnell C, Turner KJ, et al. Comparative effects of neonatal exposure of male rats to potent and weak (environmental) estrogens on spermatogenesis at puberty and the relationship to adult testis size and fertility: evidence for stimulatory effects of low estrogen levels. Endocrinology 2000;141(10):3898–907. [50] Sharpe RM, Atanassova N, McKinnell C, et al. Abnormalities in functional development of the Sertoli cells in rats treated neonatally with diethylstilbestrol: a possible role for estrogens in Sertoli cell development. Biol Reprod 1998;59(5):1084–94. [51] Atanassova N, McKinnell C, Walker M, et al. Permanent effects of neonatal estrogen exposure in rats on reproductive hormone levels, Sertoli cell number, and the efficiency of spermatogenesis in adulthood. Endocrinology 1999;140(11):5364–73.