| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Developmental Endocrinology Section,2 Laboratory of Molecular Toxicology, National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709
Department of Environmental and Molecular Toxicology,3 North Carolina State University, Raleigh, North Carolina 27605
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
developmental biology, female reproductive tract, mechanisms of hormone action, ovary
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Phytoestrogens are a group of naturally occurring compounds that have been reported to cause fertility problems in animals [9–13]. Of particular concern is genistein (Gen), the major phytoestrogen in soy products [14], which has potent estrogenic activity both in vitro and in vivo [15–18]. Human fetuses and neonates can be exposed to high levels of Gen if their mothers consume excessive amounts of soy [19] or if they are given soy-based formulas and other soy products marketed specifically for children [14, 20, 21]. The concentrations of Gen and other isoflavones found in some of these soy-based products far exceeds the amount found in an adult diet; one study estimates that infants fed soy-based formulas consume approximately 6–9 mg/kg per day of Gen compared with 1 mg kg day for an adult vegetarian [14]. Soybeans also have an extremely variable isoflavone content depending on variety and environmental conditions such as growing season and location [22], and the U.S. Department of Agriculture reports variable amounts of Gen in various soy products [23].
Over the last few years, public and scientific interest in phytoestrogens such as Gen has increased because of its proposed beneficial effects. Currently, there are mixed results on developmental exposure to Gen, suggesting some beneficial effects, but also adverse effects depending on the timing of exposure, dose level, and end points examined. For example, two studies report that prenatal exposure to Gen prevents carcinogen-induced mammary gland cancer in rats [24, 25], whereas another study shows an increase in mammary gland cancer if the developmental window of exposure is shifted to neonatal life [26]. Other investigations report improved cholesterol synthesis rates of human infants consuming soy-based formulas [27]. Vegetarian diets containing high levels of soy during pregnancy have been associated with increased incidence of hypospadias in the male offspring [28]. Further, an epidemiology of health outcomes in young adults who were fed soy-based infant formulas reported an increase in more frequent use of allergy medicines in both men and women, and longer menstrual bleeding and more discomfort during the menstrual cycle in women [29, 30]. So the adverse effects of developmental exposure to Gen remain of concern.
A recent study from our laboratory has shown that neonatal exposure to Gen at a dose of 50 mg kg day on Days 1–5 leads to an increased incidence of uterine adenocarcinoma in mice later in life; the incidence of uterine tumors in Gen-treated mice (35%) was similar to the incidence found in mice treated with an equal estrogenic dose of DES (0.001 mg/kg per day; 31%) [31]. Although Gen was administered as s.c. injections, the levels of Gen used in our studies produced circulating serum levels similar to the range of those found in infants consuming soy-based formulas [32]. Therefore, the dose of Gen to the target tissue was comparable between s.c. injections and oral exposures. Similar findings have also been reported by Lewis et al. [33] in neonatal rats using a dose of 40 mg/kg per day. Further studies have shown adverse effects on the developing rat following Gen exposure, including altered brain function, estrous cyclicity, and reproductive behavior [34, 35]. Studies using other phytoestrogens including coumestrol [9, 36], daidzein [13, 35], and red clover [10, 12] have also demonstrated disruptions in reproduction and/or reproductive end point, supporting the concept that phytoestrogens, although weaker than DES or 17ß-estradiol, can cause adverse effects on the developing reproductive tract [7]. Also, some of these effects may not be apparent until later in life, similar to those caused by DES [1].
To gain further understanding into the mechanism by which phytoestrogens can interfere with development and reproduction, a recent study in our laboratory examined the effects of Gen on the developing ovary following neonatal exposure to doses of 0.5, 5, and 50 mg/kg per day. Effects on the ovary included alterations in morphology with the presence of multioocyte follicles (MOFs) as well as alterations in function as determined by ovulatory capacity [37]. The formation of MOFs involves the actions of estrogen receptor (ER) ß, because Esr2-null (bERKO) mice do not exhibit Gen-induced MOFs [37].
The doses of Gen used in our studies were chosen to span the range of human exposure levels and estrogen activity. Previously, we have shown that Gen at a dose of 50 mg kg day is estrogenic in the mouse uterotropic bioassay in a 5-day-old neonate and in an immature mouse, and this dose is equal in estrogenic activity to DES at 0.001 mg/kg per day [7, 31]. Therefore, we chose Gen 50 mg/kg per day as our highest dose and used two lower doses of Gen to examine lower-level effects. These doses span the range to which vegetarian mothers are exposed during pregnancy and lactation, and to which infants are exposed on soy-based formulas [14, 19, 20, 28].
To further study the effects on the developing murine reproductive system, our current study examines the effects of neonatal Gen exposure on attainment of puberty and subsequent fertility, including ovarian function, estrous cyclicity, and pregnancy outcome over time.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Adult CD-1 (Crl:CD-1 [ICR] BR) mice were obtained from Charles River Breeding Laboratories (Raleigh, NC) and bred to male mice of the same strain in the breeding facility at the National Institute of Environmental Health Sciences (NIEHS; Research Triangle Park, NC). Vaginal plug detection was considered Day 0 of pregnancy. Pregnant mice were housed under controlled lighting (12L:12D) and temperature (21–22°C) conditions. Mice were housed in polysulfone, ventilated cages (Technoplast, Inc., Exton, PA) and provided with NIH-31 laboratory mouse chow and fresh water ad libitum. All animal procedures complied with an approved NIEHS/National Institutes of Health animal care protocol.
Pregnant mice delivered their young on Day 19 of gestation, pups were separated according to sex, and then randomly standardized to eight female pups per litter (a minimum of three litters are represented in each standardized litter); male pups were used in another experiment. Female pups were then assigned to a dose group and treated on Days 1–5 with Gen (Sigma, St. Louis, MO) by s.c. injection at doses of 0.5, 5, or 50 mg kg day dissolved in corn oil or left untreated as controls; these treatment groups are referred to throughout the remainder of the study as Gen-0.5, Gen-5, or Gen-50. (These doses are approximately 1, 10, or 100 µg per pup per day, respectively.) Mice were weaned at 22 days of age and housed four per cage.
Assessment of Puberty and Estrous Cyclicity
At weaning, mice (15–16 per group) treated with Gen (Gen-0.5, Gen-5, or Gen-50) and controls were checked daily until vaginal opening was observed. These mice were then allowed to age to 2 mo (after establishing regular cycles), and eight mice randomly selected from each group were monitored for 2 wk (14 consecutive days) for estrous cyclicity by taking daily vaginal smears. Smears were stained with hematoxylin-eosin (H&E) and evaluated for the stage of the estrous cycle. If the number of days in a particular stage exceeded 3 days, the mouse was considered to have an extended cycle. If the mouse never showed entry into or out of the estrous cycle and the vaginal smears contained cornified epithelium, the animal was considered in persistent estrus. Mice from all groups were allowed to age to 6 mo and daily vaginal smears were taken again for 2 wk, stained with H&E, and evaluated.
Fertility Assessments
Control female mice or those treated with Gen-0.5 or Gen-5 mg kg day were allowed to age to 2, 4, and 6 mo of age (eight mice per treatment group) and bred to proven control 2-mo-old male mice of the same strain. The females were housed with males overnight and checked the following morning for the presence of a vaginal plug. Females that were vaginal plug-positive (Day 0 of pregnancy) were removed, housed singly until delivery of pups, or until it was apparent they were not pregnant. Breeding continued for 2 consecutive wk at each time period (2, 4, and 6 mo of age) to allow for the maximal possibility of pregnancy. Females delivered their young and pups were sexed and counted. The same group of females was tested at each time point.
In a separate experiment, the fertility of Gen-50 female mice compared with controls was tested as described above. At 2–3 mo of age, control and Gen-50-treated females (eight per group) were bred to proven control males for a 2-wk period and allowed to deliver their pups. None of the mice treated with Gen-50 delivered live pups. A repeat of this study using another eight mice per group confirmed that Gen-50 treatment resulted in no live births, so this dose was not further tested at 4 and 6 mo of age.
Implantation and Pregnancy Assessment of the Gen-50 Mice
Because mice treated neonatally with Gen 50 mg/kg did not deliver live pups, implantation defects and pregnancy loss was further assessed. At 2 mo of age, control and Gen-50 female mice were bred to proven control males of the same strain (64 controls and 64 Gen-50 females). Female mice were housed overnight with males and checked for a vaginal plug the following morning. The day of vaginal plug-positive was recorded as Day 0 of gestation and females were then housed singly. The reproductive tract was collected on Pregnancy Days 6, 8, or 10, and the total number of visible implantation sites in the uterus was determined. Half of the plug-positive mice in each group were assessed for reabsorption sites on Pregnancy Days 6, 8, or 10 by collecting the uteri and soaking them in 2% NaOH in PBS for 1 h. Blood was collected from the remaining plug-positive mice in each group on Pregnancy Days 6, 8, or 10 to determine hormone levels of progesterone, estradiol, and testosterone. (The plug-positive mice were split into two groups, one for implantation site counts and one for blood collection, because the two procedures could not be performed in the same mouse due to inability to visualize implantation sites in mice that were bled [unpublished observations]). Ovaries were also collected from pregnant mice on Pregnancy Days 6, 8, or 10 and fixed in 10% neutral-buffered formalin, cut at 6 µm, and stained with H&E to determine the number of corpora lutea (CLs) per mouse. The number of CLs was determined by counting the number of CLs in both ovaries from three sections per ovary. The averages from each mouse were combined across the treatment group and the mean ± SEM was determined for the group.
Ovarian Function
Ovaries were collected from 6-wk-old or 4 mo-old control and Gen-treated mice (Gen-0.5, Gen-5, and Gen-50), fixed in 10% formalin, and processed for histological evaluation (eight mice per treatment group per age). The number of CLs per mouse was determined as described above and the data for the group is expressed as the mean ± SEM.
The ovulatory capacity of neonatally Gen-treated mice was determined using methods previously described [38]. At 4 mo of age, all mice received a single s.c. injection of 2.2 IU eCG (Sigma) followed 48–52 h later with 3.2 IU hCG (Sigma). The animals were then killed by CO2 16–20 h after hCG injection and oviducts were placed in M-2 medium (Specialty Media, Lavallette, NJ) supplemented with 0.3% hyaluronidase (Sigma). The oocyte/cumulus mass was surgically extracted from the oviduct and oocytes were counted after enzymatic disassociation from the surrounding cumulus. This experiment was carried out twice with each set including eight mice per treatment group. Data from the two experiments were averaged together and are presented as mean ± SEM.
Serum Hormone Levels
To determine the baseline of circulating serum levels of progesterone and estradiol, serum measurements were taken before puberty, before the onset of estrus cyclicity to avoid the variations of the cycle. Mice treated neonatally with Gen-0.5, Gen-5, Gen-50, or controls were killed by CO2 at 19 days of age (eight mice per group), and blood was collected from the caudal vena cava using a hypodermic needle and a 1-ml syringe. Because the amount of blood obtained was limited due to the size of the animals, some samples were pooled to perform the assay. Blood was also collected during pregnancy from control and Gen-50 mice as described for prepubertal mice and assayed to determine circulating levels of progesterone, estradiol, and testosterone.
All blood samples were centrifuged at 8000 x g at 4°C for 10 min; serum was isolated and frozen at –70°C until further analysis. Serum levels of progesterone, estradiol, and testosterone were measured using respective kits (Diagnostic Systems Laboratory, Webster, TX) according to the manufacturer's instructions.
Statistical Analysis
When examining continuous outcomes (day of vaginal opening, litter size, number of implantation sites, CLs, and serum hormone measurements), an initial analysis of variance was performed to determine differences, and then each treatment group was compared with controls by the Dunnett test using Statistical Analysis Systems software (Cary, NC). For serum hormone measurements at 19 days of age, where pooled samples were required, weighted analyses were performed with the weights reflecting the pool size. For litter sizes at 2, 4, and 6 mo of age, where the same animals were used at all three time points, a simultaneous analysis at all ages was also performed; effects of age and dose were included in the analysis. For hormones during pregnancy, because variance increased with the mean, analysis was performed on a log scale.
When examining categorical outcomes (estrous cycle and pregnancy rates), Fisher exact tests were used to compare across treatment groups. For the pregnancy rate at 2, 4, and 6 mo of age an analysis of trend was also performed using the Cochran-Armitage test. Statistical significance was determined at P < 0.05 for all tests.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Vaginal opening, used as an indicator of puberty, was determined following neonatal treatment with Gen (Table 1). Mice were observed from 22 days of age at weaning until vaginal opening occurred. Three out of 16 (19%) mice treated with Gen-0.5 had vaginal opening before their age-matched controls with the earliest vaginal opening occurring at Day 26 compared with Day 29 for the controls. Two out of 16 (13%) mice from the Gen-5 group, and 1 out of 16 (6%) mice treated with Gen-50 had vaginal opening on Day 28, 1 day earlier than controls. Mice from the Gen-50 group showed a general delay in vaginal opening when compared with controls with only 1 out of 16 (6%) of the Gen-50 mice exhibiting vaginal opening at 30 days of age compared with 6 out of 15 (40%) of the controls. All mice in all treatment groups exhibited vaginal opening by 37 days of age. The average day of vaginal opening was calculated for each group and compared with controls using the Dunnett test; there were no statistically significant differences between control and Gen-treated mice (Table 1).
|
Mice treated with Gen-50 had an intense reddening of the area surrounding the vagina. This reddened area was present at weaning and before vaginal opening and remained throughout puberty and into adulthood. Mice from the other two doses of Gen did not exhibit this abnormality.
Although some mice treated with Gen-0.5 and Gen-5 had early vaginal opening, and some Gen-50 mice had delayed vaginal opening, circulating serum levels of estradiol and progesterone at 19 days of age just before puberty were not statistically different from controls by the Dunnett test (Table 2).
|
Genistein Effects on Estrous Cyclicity
At 2 mo of age, all control mice (8/8; 100%) exhibited regular estrous cycles but Gen-treated mice had altered cycles (Table 3; Fig. 1A). Two out of eight (25%) mice treated with Gen-0.5 had extended diestrus and one of eight (13%) mice had extended estrus. Four out of eight (50%) mice treated with Gen-5 had extended diestrus, and three out of eight (38%) had extended estrus. Mice treated with Gen-50 exhibited extended estrus (6/8; 75%); in fact, one mouse (1/8; 13%) was in persistent estrus having cornified epithelium in all of its vaginal smears.
|
|
At 6 mo of age, similar patterns of abnormal estrous cyclicity were observed in the Gen-treated mice as at 2 mo, but there was an increase in severity of altered cycles (Table 3; Fig. 1B). In Gen-0.5 treated mice, one of eight (13%) had extended estrus. Mice in the Gen-5 treatment group had extended diestrus (4/8; 50%), extended estrus (2/8; 25%), and one demonstrated persistent estrus (1/8; 13%). More than half of the mice treated with Gen-50 were in persistent estrus (5/8; 63%), indicating increased severity and higher incidence over time; in addition, one of eight (13%) was in extended diestrus, and two of eight (25%) were in extended estrus. The estrous cycle patterns for each mouse are illustrated in Figure 1, A and B. The time spent in a stimulated (estrogenized) portion of the cycle is shown in black; the Gen-50 group shows increased black areas compared with controls. Differences among the doses in the distribution across categories are highly significant at 2 and 6 mo using the Fisher exact test (P < 0.01).
Fertility Assessments
No difference was observed in the number of plug-positive mice in any of the Gen-treated groups compared with controls at 2, 4, and 6 mo of age (Table 4). There were significantly fewer pregnant mice in the Gen-treated groups at 2 mo of age using the Fisher exact test (P < 0.05) when the Gen-50 group was included; this group did not have any litters. In addition, a trend test showed a statistically significant decrease in the number of mice with litters at 2 and 6 mo of age with increasing dose as determined by the Cochran-Armitage test (P < 0.05; Table 4); the significance of reduced numbers of mice with litters remains even when the Gen-50 group was excluded from the analysis. This effect is most pronounced at 6 mo of age; fewer plug-positive mice had litters in the Gen-0.5 group (60%) and even less in the Gen-5 group (40%) than their age matched controls (100%). This is consistent with early reproductive senescence.
|
The number of live pups in Gen-treated groups did not differ from the number of live pups in control groups at any age using the Dunnett test, examining each time separately. However, looking at all ages simultaneously, the mice in the Gen-5 treatment group had significantly fewer pups than controls, whereas the Gen-0.5 treatment group was not statistically different (Table 4).
Implantation and Pregnancy Loss Assessment of the Gen-50 Mice
Because mice treated neonatally with the highest dose of Gen did not deliver live pups, pregnancy loss was further studied in this treatment group. There was no difference between control and Gen-50 mice in the number of mice that were plug-positive following mating (control, 43/64, 67%; Gen, 43/64, 67%). Examination of uterine contents during the course of pregnancy showed that on Pregnancy Day 6, 8/13 (62%) of the Gen-treated plug-positive mice had visible implantation sites compared with 16/18 (89%) of the controls (Fig. 2A). In addition, the total number of implantation sites was less in the Gen-treated group (Fig. 2B) and the ones that were present appeared to be smaller in size compared with the controls; there were no apparent reabsorptions at this stage of pregnancy. On Pregnancy Day 8, the outcome was similar to Day 6 but with a lower number of Gen-treated plug-positive mice showing signs of implantation sites (7/19; 37%) compared with controls (18/19; 95%) (Fig. 2A). In addition to fewer and smaller visible implantation sites (Fig. 2B) compared with controls, three mice had apparent reabsorptions (13 sites total). A representative example of a control uterus with implantation sites on Day 8 of pregnancy can be seen in Figure 2C(a). Also seen in Figure 2C are three examples of uteri from plug-positive Gen-treated mice on Pregnancy Day 8 (Fig. 2C, b, c, and d). Note the smaller size of the implantation sites in two of the Gen-treated examples (b and c) compared with control (a) as well as reabsorption sites in one example (c); the third uterus is from a mouse that had no visible implantation sites (d). By Pregnancy Day 10, there were still fewer Gen-treated mice with visible implantation sites (5/11; 45%) compared with controls (6/6; 100%) as well as fewer and smaller implantation sites (Fig. 2, A and B).
|
Statistical analysis was performed on the pregnancy rate using Fisher exact tests and the number of implantation sites using the Dunnett test (Fig. 2, A and B). Gen treatment significantly reduced the number of mice that were pregnant on Day 8 and Day 10 (P < 0.05). Gen treatment also significantly reduced the number of implantation sites compared with controls on all days of pregnancy examined (P < 0.05). (Nonpregnant mice were not included in the analysis of the average number of implantation sites per mouse.)
CLs were counted in ovaries from control and Gen-50 mice on Pregnancy Days 6, 8, and 10 (Fig. 3). Of mice that were pregnant, Gen-treated mice had far fewer CLs than their control counterparts at all stages of pregnancy examined, reaching statistical significance on Days 6 and 8 using the Dunnett test (P < 0.05). As expected, Gen-treated mice with no implantation sites had even fewer CLs than their pregnant counterparts (statistically significant at all three time points; Fig. 3).
|
Serum hormone levels during pregnancy are shown in Figure 4. There were no statistically significant differences between the control and Gen-treated pregnant mice in the levels of progesterone at any stage of pregnancy examined despite fewer CLs in ovaries of the Gen-treated mice; however, the Gen-treated nonpregnant mice demonstrated significantly less progesterone compared with control pregnant mice on Day 6 and Day 8 (Fig. 4A). There were also no significant differences in estradiol levels between treatment groups (Fig. 4B); however, there was a large range in estradiol levels in the Gen-treated mice (pregnant and nonpregnant) compared with control mice on Pregnancy Day 6 (control pregnant, 61.5–106.2; Gen pregnant, 62.8–353.5; Gen nonpregnant, 41.0–464.2 ng/ml), suggesting increased serum estradiol levels in some Gen-treated mice; however, this does not appear to depend on pregnancy outcome. There was also no significant difference in testosterone levels between control pregnant and Gen-treated pregnant mice; Gen-treated nonpregnant mice had significantly less testosterone than control pregnant mice on Day 8 (Fig. 4C).
|
Ovarian Function
To examine the early reproductive senescence observed at Gen-0.5 and Gen-5, ovarian function was studied in all treatment groups. At 6 wk of age, there were no statistically significant differences in the numbers of CLs in any of the Gen treatment groups compared with controls using the Dunnett test at P < 0.5; however, two of eight (25%) mice in the Gen-50 group did not have any CLs (Table 5), indicating that some mice were more severely affected than others.
|
At 4 mo of age, there were more CLs in the Gen-0.5 and Gen-5 mice compared with controls (significant in the Gen-5 group using the Dunnett test at P < 0.05), suggesting that under their own hormonal cues, more oocytes are ovulated (Table 5). In sharp contrast, the Gen-50 mice did not have any CLs (also significant at P < 0.05), indicating they do not ovulate under their own hormonal cues at this age, suggesting an increase in severity over time.
To determine whether these mice were capable of ovulation, oocytes were collected and counted from mice that were superovulated with eCG and hCG (Table 5). All Gen-treated mice had similar numbers of ovulated oocytes compared with controls, suggesting that these mice are capable of ovulation following exogenous stimulation even though mice treated with the high dose of Gen do not appear to ovulate under their own hormonal cues.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Although humans are exposed to Gen primarily by an oral route of exposure, data from our laboratory have demonstrated that serum levels of Gen in female mice following neonatal treatment with Gen-50 average 6.8 ± 1.4 µM [29] compared to human infants on soy-based infant formulas who have circulating levels ranging from 1 to 5 µM [14, 32]. This study also showed high circulating levels of the aglycone form of Gen (
2 µM), which is 10-fold higher than an adult rat following similar exposure [41]; the aglycone form has been previously shown to exhibit estrogen receptor-binding activity [42]. The higher fraction of the aglycone form of Gen has also been shown in the perinatal rat [43], supporting the idea that Gen glucuronidation is lower during the neonatal period compared to adulthood, most likely due to lower uridine diphosphate glucuronosyltransferase (UGT) activities in neonatal mice [44, 45]. The ontogeny of many human UGT isoforms is similar to the pattern of rodent development with lower activity in the neonatal period [46, 47]. Although the serum circulating levels of the aglycone form of Gen in the neonatal infant is not known, an elevated fraction compared to that in adults and similar to what is observed in neonatal rodents, seems likely. Data from another laboratory using orally dosed neonatal rat pups with Gen at similar doses used in our study showed similar adverse effects, including MOFs in the ovary and reduced female fertility [48]. Another study exposing Sprague-Dawley rats during pregnancy and lactation to Gen in the diet showed a reduction in the percentage of female mice delivering live pups (controls, 9/ 10; Gen 1250 ppm, 5/10) [49]. All of these data taken together strongly suggest that Gen exhibits similar results on the female reproductive system regardless of the route of exposure (oral or s.c.) or the species examined (rat or mouse).
We have previously addressed the role of diet used in our studies, because the NIH-31 lab chow contains low levels of phytoestrogens. This diet contains approximately 98 µg/g of Gen and daidzein, which is about 16.7 mg kg day for a 30-g mouse [50]. It has been shown that mice exposed to Gen at a dose of 16 mg/kg orally during lactation have a serum circulating level of genistein of 1.8 µg/ ml, but the level found in the milk was only 0.04 µg/ml, which is 45 times less. Therefore, the amount of Gen that is consumed by the mother from the diet would result in very low exposure to the pups [33] and is far below the treatment levels used in this study. Further, control and Gen-treated mice were all fed the same diet. Therefore, any contribution of phytoestrogens from the diet was minimal in causing the effects observed in this study.
The current study indicates that there are problems with female reproductive development and function in mice exposed neonatally to Gen. There were no apparent differences in serum hormone levels in immature Gen-treated mice compared with controls, indicating no inherent differences in circulating levels of progesterone or estradiol between the treatment groups before puberty. While there were slight alterations in the onset of puberty, with lower doses advancing vaginal opening and higher doses delaying it, the mean day of vaginal opening was not statistically different between the groups. Although we did not show a significant difference in the mean age of vaginal opening in the current study, others have shown differences, particularly in the rat model, supporting the idea that puberty may be altered following developmental exposure to Gen depending on the time of exposure and the species examined. For example, Nikaido et al. [51] showed earlier onset of puberty in mice following prenatal exposure to several environmental estrogens including Gen at a low dose of 0.5 mg/kg [51]. Levy et al. [52] showed a delay in vaginal opening following prenatal exposure of rats to Gen at a dose of 5 mg/kg, and Kouki et al. [35] showed an advanced time of vaginal opening in rats treated neonatally with Gen at a dose of 1 mg/kg.
Data from the current study also showed alterations in the estrous cycle of mice following neonatal exposure to Gen at all doses examined with extended cycle length being the most common finding. Others have shown similar estrous cycle alterations in other model systems, including the study by Nikaido et al. showing several environmental estrogens including Gen, resveratrol, zearalenone, and bisphenol A given during pregnancy caused extended estrous cycles [51]. Kouki et al. also showed irregular estrous cycles following neonatal exposure of rats to Gen with prolonged periods in estrus [51]. Alterations in estrous cyclicity were exacerbated over time, with more mice exhibiting persistent estrus at 6 mo compared with 2 mo of age; interestingly, not only at the high dose, but at the lower doses as well. This altered estrous cyclicity at the lower doses may in part explain the early reproductive senescence observed in these mice, particularly at 6 mo of age.
Our current study also clearly demonstrates that the ovary is adversely affected by neonatal exposure to Gen. Although all Gen-treated mice ovulate under exogenous hormonal influence, the ovulation rate was much lower in the highest dose of Gen under their own hormonal cues as evidenced by fewer CLs at 2 mo and during pregnancy, and by the absence of CLs at 4 mo of age. In addition, the lower doses of Gen treatment also resulted in alterations in ovarian function with more CLs than their control counterparts at 4 mo of age. This enhanced ovulation rate is similar to what was observed at 26 days of age following superovulation in a previous study in our laboratory [37]. This may also in part explain the early reproductive senescence observed in the two lower doses of Gen treatment at 6 mo of age. Because more oocytes are ovulated earlier, there may be a decrease in the number of oocytes available for fertilization at later time points. This effect is not unique to Gen, because earlier studies in our laboratory using DES showed similar effects; low doses of prenatal exposure, neonatal exposure, or both to DES causes enhanced ovulation rates and early reproductive senescence, further suggesting that ovulation of too many oocytes early in life may lead to lower fertility rates later in life [2]. It has been reported that aged mice do not typically exhaust their total complement of oocytes; however, they exhibit characteristics of reproductive senescence, including lowered responsiveness of the pituitary to estradiol, gradual loss of ovulatory function, decreased fertility, and smaller litter sizes [53–57], but this occurs much later in life than 6 mo of age, which was observed following neonatal Gen exposure. Early reproductive senescence could be important for human reproductive health because more and more women are waiting longer to become pregnant [58].
One possible explanation of enhanced ovulation rates observed in lower doses was proposed by Faber and Hughes [59]. That study showed that neonatal exposure of rats to low doses of Gen (0.01 mg/kg) was associated with an increased pituitary response to GnRH, producing higher levels of LH [59]. Mice treated with lower doses of Gen used in this study may be hyperresponsive to GnRH stimulation leading to enhanced ovulation rates, which we have shown previously in younger mice [37] and again in older mice in this study. In addition, Faber and Hughes showed that higher doses of Gen were associated with decreased pituitary responsiveness [59], which may explain the lower number of CLs at 2 mo of age and the lack of CLs at 4 mo of age observed herein. Altered pituitary responsiveness later in life could also account for the early reproductive senescence observed in mice treated with lower doses used in our study. Further investigation of the effects of Gen on the hypothalamic-gonadal axis is currently underway in our laboratory.
The current study also shows that Gen adversely affects pregnancy outcome, as mice exposed to the highest dose of Gen did not deliver live pups, and mice exposed to lower doses of Gen showed signs of reduced fertility with age. While some of the mice treated with Gen-50 were able to become pregnant at 2 mo of age, they were unable to carry these litters to term. The Gen-50-treated mice that became pregnant had fewer implantation sites, which may be explained, in part, by the lower ovulation rate exhibited by these mice under their own hormonal cues. In addition to having fewer implantation sites, the implantation sites were much smaller than the implantation sites from controls at the same gestational age. There are several possibilities why this may have occurred. One explanation is that the lower number of CLs in the Gen-50-treated mice could have led to inability to support pregnancy due to lower levels of circulating progesterone, because progesterone is essential for maintenance of pregnancy. However, we have shown that serum circulating levels of progesterone, as well as estradiol and testosterone, were similar in Gen-treated and control pregnant mice. Another possibility is that the uterus was unable to support pregnancy. The observation that the embryos implanted suggests that there was some capacity of the uterus to function properly. Another possibility is that the oocyte itself is of poor quality. We have shown previously that the development of the ovary and ovarian follicle were altered following neonatal Gen treatment [37]. Ovaries of Gen-treated mice contained MOFs at 19 days of age, a phenotype not often observed in control CD-1 mice. This phenotype may be a marker for altered development of the ovary, leading to oocytes of poor quality. In fact, a paper by Iguchi et al. using neonatal DES treatment showed that oocytes derived from single oocyte follicles were far more likely to be fertilized in vitro than oocytes derived from MOFs, suggesting that these oocytes are less competent [60]. Because neonatal Gen treatment causes an increase in MOFs [37], perhaps fewer ovulated oocytes are capable of being fertilized. We are currently investigating specific uterine and ovarian defects using embryo transplantation experiments.
In summary, our data demonstrate that neonatal exposure to Gen has deleterious effects on the developing murine reproductive system and can have long-term consequences on fertility at environmentally relevant doses. While there are certainly adverse effects on the reproductive system, this appears to be a multifaceted problem, because these mice have altered estrous cycles, altered ovarian function, and lower pregnancy rates. While the most severe effects were observed at a dose of 50 mg/kg with lack of ovarian function and inability to carry pups to term, there were also adverse consequences to reproduction observed at the Gen 0.5 and 5 mg/kg treatment groups with altered ovarian function, extended estrous cycles, and early reproductive senescence. Additional studies are warranted in human infants who are exposed to high levels of Gen during development before concluding that such exposure is safe.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received: 23 February 2005.
First decision: 13 April 2005.
Accepted: 26 May 2005.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Related articles in Biol Reprod:
This article has been cited by other articles:
|
|
 |
|
  J. P. Konhilas and L. A. Leinwand The Effects of Biological Sex and Diet on the Development of Heart Failure Circulation, December 4, 2007; 116(23): 2747 - 2759. [Full Text] [PDF]
|
|
|
|
 |
|
 E. Padilla-Banks, W. N. Jefferson, and R. R. Newbold Neonatal Exposure to the Phytoestrogen Genistein Alters Mammary Gland Growth and Developmental Programming of Hormone Receptor Levels Endocrinology, October 1, 2006; 147(10): 4871 - 4882. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. A.L. Tan, M. Walker, K. Morris, I. Greig, J. I. Mason, and R. M. Sharpe Infant feeding with soy formula milk: effects on puberty progression, reproductive function and testicular cell numbers in marmoset monkeys in adulthood Hum. Reprod., April 1, 2006; 21(4): 896 - 904. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 W. Jefferson, R. Newbold, E. Padilla-Banks, and M. Pepling Neonatal Genistein Treatment Alters Ovarian Differentiation in the Mouse: Inhibition of Oocyte Nest Breakdown and Increased Oocyte Survival Biol Reprod, January 1, 2006; 74(1): 161 - 168. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||
