Toxicological Sciences 68, 121-146 (2002)
Copyright © 2002 by the Society of Toxicology
REPRODUCTIVE AND DEVELOPMENTAL TOXICOLOGY |
Three-Generation Reproductive Toxicity Study of Dietary Bisphenol A in CD Sprague-Dawley Rats
* RTI, Research Triangle Park, North Carolina;
EPL, Inc., Research Triangle Park, North Carolina;
GE Plastics, Pittsfield, Massachusetts;
Aristech Chemical Corp., Pittsburgh, Pennsylvania;
¶ Shell Chemical Co., Houston, Texas;
|| Bayer Corp., Stilwell, Kansas;
|| Bayer AG, Wuppertal, Germany; and
|||| The Dow Chemical Co., Midland, Michigan
Received March 20, 2001; accepted February 11, 2002
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bisphenol A (BPA) was evaluated at concentrations of 0, 0.015, 0.3, 4.5, 75, 750, and 7500 ppm (
0.001, 0.02, 0.3, 5, 50, and 500 mg/kg/day of BPA) administered in the diet ad libitum to 30 CD® Sprague-Dawley rats/sex/dose for 3 offspring generations, 1 litter/generation, through F3 adults. Adult systemic toxicity at 750 and 7500 ppm in all generations included: reduced body weights and body weight gains, reduced absolute and increased relative weanling and adult organ weights (liver, kidneys, adrenals, spleen, pituitary, and brain), and female slight/mild renal and hepatic pathology at 7500 ppm. Reproductive organ histopathology and function were unaffected. Ovarian weights as well as total pups and live pups/litter on postnatal day (PND) 0 were decreased at 7500 ppm, which exceeded the adult maximum tolerated dose (MTD). Mating, fertility, gestational indices; ovarian primordial follicle counts; estrous cyclicity; precoital interval; gestational length; offspring sex ratios; postnatal survival; nipple/areolae retention in preweanling males; epididymal sperm number, motility, morphology; daily sperm production (DSP), and efficiency of DSP were all unaffected. At 7500 ppm, vaginal patency (VP) and preputial separation (PPS) were delayed in F1, F2, and F3 offspring, associated with reduced body weights. Anogenital distance (AGD) on PND 0 was unaffected for F2 and F3 males and F3 females (F2 female AGD was increased at some doses, not at 7500 ppm, and was considered not biologically or toxicologically relevant). Adult systemic no observed adverse effect level (NOAEL) = 75 ppm (5 mg/kg/day); reproductive and postnatal NOAELs = 750 ppm (50 mg/kg/day). There were no treatment-related effects in the low-dose region (0.001–5 mg/kg/day) on any parameters and no evidence of nonmonotonic dose-response curves across generations for either sex. BPA should not be considered a selective reproductive toxicant, based on the results of this study. Key Words: Bisphenol A; CAS No. 80-05-7; dietary administration; systemic toxicity; reproductive toxicity; postnatal toxicity; OPPTS 837.3800 guidelines.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Bisphenol A (BPA) is a high production volume chemical used principally as a monomer in the manufacture of numerous chemical products, including polycarbonate plastics and epoxy resins. Industrial exposure during the manufacture and use of the monomer is variable, depending on the use and duration of exposure. Consumer exposure to BPA may be possible from migration of BPA from dental sealants or from polycarbonate or epoxy-lined food and drink containers. Although it had been known for decades that BPA has weak estrogen-like activity in vivo by sc injection (Dodds and Lawson, 1936
), there has been recent, renewed interest in its potential for estrogen-like activity. Krishnan et al. (1993) discovered that BPA leaching from polycarbonate flasks during autoclaving induced an estrogen-like response in yeast cultures. Gaido et al. (1997) confirmed its estrogen-like activity in vitro, calculating BPA as approximately 15,000-fold less potent than 17ß-estradiol (E2). Kuiper et al. (1997) demonstrated that BPA could interact with both the 
- and ß-estrogen receptors. Kuiper et al. (1998) also showed that the in vitro binding affinity of BPA was approximately 10,000-fold less potent than that of E2 and 20,000-fold less potent than diethylstilbestrol (DES) for both ER and ERß. Maruyama et al. (1999) also reported BPA to be 10,000-fold less potent than E2 in vitro using an E2 responsive rat pituitary cell line. In vivo, Milligan et al. (1998) showed BPA to be 10,000-fold less potent in producing a uterotrophic effect than estradiol following sc injections into ovariectomized mice. Ashby and Tinwell (1998), Jekat et al. (2000), Kim et al. (2001), Laws et al. (2000), Matthews et al. (2001), and Yamasaki et al. (2000) also reported uterotrophic effects in rats following high oral and/or sc dosing, and Goloubkova et al. (2000) reported stimulatory effects on the growth of the pituitary gland following high sc doses of BPA.
Research conducted in the 1970s and 1980s, using 1-generation (CD rats) or 2-generation continuous breeding (CD-1 mice) designs, indicated that BPA was not a selective reproductive toxicant with high dietary BPA concentrations (Morrissey et al., 1989; Wazeter and Goldenthal, 1984a,b). Standard Segment II developmental toxicity studies in CD rats and CD-1 mice administered BPA at high doses by gavage on gestational day (GD) 6–15 indicated that BPA was not a selective developmental toxicant (Morrissey et al., 1987). More recently, Liaw et al. (1998) showed that exposure of pregnant SD (Sprague-Dawley) rats to BPA in drinking water from GD 2 through lactation (until PND 21) at 0, 0.005, 0.05, 0.5, 5, or 50 mg/l (ppm) and DES at 0.05 mg/l (ppm) did not affect the age or body weight at acquisition of VP and had no significant effects on reproductive organ development. DES accelerated acquisition of VP (with reduced body weights).
In contrast to the "high" dose studies above, oral administration (presentation to the dam's buccal cavity) of BPA at 2 and 20 µg/kg/day in corn oil to pregnant CF-1 mice on GD 11–17 was reported to increase prostate gland weight at both doses and decrease DSP per gram testis (efficiency of DSP) at 20 µg/kg/day in offspring males (Nagel et al., 1997; vom Saal et al., 1998). However, these reported low-dose effects of BPA could not be reproduced in more robust studies designed with larger numbers of animals and the same (Ashby et al., 1999) and additional lower and higher doses (Cagen et al., 1999a), and the NTP Low-Dose Peer Review's Statistical Subpanel could not confirm the statistical significance of the decreased DSP per gram testis at 20 µg/kg/day (NTP, 2001).
In another study examining low-dose exposure, adult male offspring of female Wistar rats exposed to 1 ppm BPA (corresponding to approximately 0.1–0.4 mg/kg/day) in their drinking water for 8 to 9 weeks (during prebreed, mating, gestation, and lactation) were reported to exhibit significantly reduced testes weights (Sharpe et al., 1995). The results of this study were brought into question when the original authors could not reproduce their initial findings or other studies that had produced the same results in different chemicals (Sharpe et al., 1998). The results of the initial study with BPA could also not be reproduced in another study using the same exposure route, timing, and strain of rat, but with a larger number of dose groups, more animals per dose, and more reproductive parameters (Cagen et al., 1999b). The studies previously reported as positive usually had smaller numbers of animals, fewer doses, and/or parenteral routes of administration. Therefore, the present study was designed and performed to definitively evaluate the concerns for possible low-dose effects, for possible nonmonotonic ("inverted-u") dose-response curves, and for possible effects of exposure to BPA by a relevant route of administration during sensitive life stages (pre- and early postnatal), as well as postweaning peripubertal maturational stages, over 3 generations of offspring using an internationally accepted reproductive toxicity protocol under Good Laboratory Practice (GLP) regulations (U.S. EPA, 1989).
Specifically, this study evaluated exposure of CD® (SD) rats (30/sex/group) to BPA administered in the diet ad libitum at 0, 0.015, 0.3, 4.5, 75, 750, and 7500 ppm (resulting in BPA intakes of 0.001, 0.02, 0.3, 5, 50, and 500 mg/kg/day) for 3 generations, 1 litter per generation, using the U.S. EPA OPPTS test guidelines (U.S. EPA OPPTS 837.3800, 1998). Additional assessments beyond the guideline requirements included a third offspring generation, 1 control and 6 treatment groups, examination for retained nipples and areolae in male F1, F2, and F3 preweanlings, and retention of F3 offspring until adulthood with continuing exposure, with histopathologic and andrological assessments at their termination.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Test material and dietary formulations.
BPA (4,4'-isopropylidene-2'-diphenol; CAS No. 80-05-7) was obtained in one shipment and lot number from Acros Organics NV (Fairlawn, NJ) as a 99.5% pure white crystalline solid and purity was reconfirmed throughout the study.
The basic diet was Purina Certified Rodent Chow® (No. 5002, PMI Feeds, Inc., St. Louis, MO). Dosed diets were formulated by dissolving BPA in a fixed volume of acetone as separate stock solutions, 1 for each dietary dose. Each BPA-acetone stock solution was added to a premix feed aliquot. After evaporating the acetone, each premix was blended with additional feed to make the prescribed concentrations for each of the 17 formulation dates. Control diets were formulated as described above.
Stability of formulations at 15 ppb and 7500 and 10,000 ppm was confirmed at approximately -20°C for 50 days, and at room temperature in open containers to simulate cageside conditions for at least 9 days. Homogeneity was also confirmed by assaying 1 sample each at 15 ppb and 7500 and 10,000 ppm in triplicate from 3 locations within the blender. Aliquots from all dosed feed preparations were analyzed for BPA concentration, and the diet was used only if within the acceptable range (± 15% of the nominal). All analyses of the feed were performed using negative ion CI (chemical ionization) gas chromatography-mass spectrometry (GC-MS) analysis. The estimated limit of detection was 0.0008–0.0018 ppm.
Formulated diets were stored at -20°C for up to 50 days in sealed containers. Feed was changed at least every 7 days.
Animals and husbandry.
The SD rat is recommended for use in reproductive and developmental toxicity testing by worldwide regulatory agencies such as the U.S. EPA, OECD, and Japanese MAFF. It was also chosen for this study because of the extensive historical database with this strain at RTI. Two hundred forty virgin female and male rats were ordered for the study. Ten/sex were used as quality controls for assessment of viral antibody status within 1 day after receipt, 8/sex were used as sentinels for monitoring of health status of study animals (with 2/sex each evaluated for viral antibody titers at the necropsy of F0, F1, F2, and F3 adults); 12/sex were available to replace any animals inappropriate for use, and 210/sex went on study. All viral antibody titer assessments for quality control and sentinel animals were negative.
At the end of the approximately 1-week quarantine period, all animals were in good health and were randomly distributed into 7 strata by sex and body weight. The rats within each stratum were then randomly assigned, 1 to each treatment group, using a random number table, and uniquely identified by eartag and animal study numbers. All selected weanling offspring were also identified by eartag and animal study numbers.
The animals were individually housed in stainless-steel hanging cages upon arrival, during the acclimation period, and upon the initiation of the treatment period. An automatic watering system was used for all animals during prebreed and for the males after mating during the holding period. Mating pairs (1 male:1 female) and sperm/plug positive females, from GD 0 until weaning of their litters on PND 21, were housed in solid-bottom polypropylene cages (Laboratory Products, Rochelle Park, NJ), with Sani-Chip® cage bedding (P. J. Murphy Forest Products, Inc., Montville, NJ) with glass water bottles. The caging, water bottles and sipper tubes, and storage containers for feed were made from materials that did not contain BPA to prevent any extraneous exposure of animals. Temperature (64–79°F), 12-h light/dark cycle, and relative humidity (30–70%) in the animal rooms were continuously monitored, controlled, and automatically recorded.
Purina Certified Ground Rodent Chow (No. 5002, PMI Feeds, Inc., St. Louis, MO) was available ad libitum, 7 days per week, 24 hours per day, throughout the study. The analyses of each feed batch for nutrient levels and possible contaminants were performed by the supplier. The supplier reported total isoflavone content (as aglycone equivalents) of 309.2 µg/g feed (range 290.0–358.0 µg/g), of which genistein was 127.6 µg/g (113.0–139.0 µg/g) and daidzein was 131.3 µg/g (114.0–167.0 µg/g). Total protein content was 20.1%. For all feed batches, nutrient levels were at or above, and contaminant levels were below the certified levels, and therefore judged suitable for use. Water was available ad libitum by an automatic watering system during the time the animals were in hanging cages, and by water bottles during the time the animals were in solid-bottom cages. At all times, the regular analyses of the water showed that contaminants were below the maximum levels defined for drinking water.
Study design.
A graphic representation of the study design is presented in Figure 1. The study began with 30 males/group and 30 females/group (designated the F0 generation), to yield at least 20 pregnant females/group at or near term, and 7 groups (see Table 1). The target dietary concentrations (0, 0.015, 0.3, 4.5, 75, 750, and 7500 ppm) were selected to provide BPA intakes of approximately 0.001, 0.02, 0.3, 5, 50, and 500 mg/kg/day, respectively, to encompass the ranges of low oral BPA doses (0.002 and 0.02 mg/kg/day) at which male mouse offspring prostate weights were reported to be significantly increased (Nagel et al., 1997; vom Saal et al., 1998), and of doses at which testis weight and efficiency of DSP were reported to be significantly reduced in rat offspring (Sharpe et al., 1995). The dietary concentrations were also chosen to provide an MTD that is expected to exceed the metabolic capability of the adult liver and to produce reductions in body weight or other indications of systemic toxicity. The target dietary concentrations were based on the chosen BPA intakes in mg/kg/day for the female rats (Table 1).
|
|
Dietary BPA was available ad libitum to the F0 animals for a 10-week prebreed exposure period, during mating, during gestation, and females through lactation until weaning. Body weights and feed consumption were recorded weekly, and clinical signs were recorded at least once daily. Vaginal cytology was evaluated for the last 3 weeks of the prebreed period. Animals were randomly mated within treatment groups for a 2-week period to produce the F1 generation. All F0 males were necropsied after F1 delivery, with histopathologic evaluation of reproductive and other organs (all controls and 10 per dose, plus any gross lesions and reproductive tissues from unsuccessful breeders or animals suspected of reduced fertility) and andrological assessments (reproductive organ weights; epididymal sperm number, motility, and morphology; testicular homogenization-resistant spermatid head counts; DSP; and efficiency of DSP in all males in all groups; Robb et al., 1978
; Sharpe et al., 1995). F1 litters were culled to 10 pups (with equal sex ratio, if possible) on PND 4, and F1 males were examined on PND 11–13 for retained areolae and/or nipples. At weaning (PND 21), 30/sex/dose were then randomly selected as F1 parents of the F2 generation, and up to 3 remaining weanlings/sex/litter were randomly selected, necropsied, and selected organs weighed. All F0 females were necropsied and selected organs were weighed. The stage of estrus at necropsy and enumeration of ovarian primordial follicles (from step sections of both ovaries of ten females each at high dose and control) were determined, and histopathological examinations of reproductive and other selected organs (same as F0 males above) were performed.
Selected F1 weanlings were administered BPA in the diet for the exposure period as described above for the F0 generation. Acquisition of VP in F1 females and PPS in F1 males was determined during prebreed. Vaginal cytology for estrous cyclicity was evaluated during the last 3 weeks of prebreed. Since acquisition of puberty was delayed in F1 offspring at 7500 ppm, measurement of AGD was performed on all F2 and F3 offspring at birth (PND 0) using an ocular micrometer and eyepiece grid (precision = 0.2 mm). At weaning of F2 litters, the same procedure as described above was used to select the F2 parents of the F3 generation. All F1 males and females were necropsied, with histopathology as described above.
Randomly selected F2 weanlings were administered BPA in the diet for the exposure period as described above for the F0 and F1 generations. Acquisition of VP and PPS and evaluation of estrous cyclicity were performed as above for the F1 generation. They were then mated as described above to generate F3 litters. F2 parental animals were necropsied with histopathology as described above. At weaning of F3 litters, up to 3 weanlings/sex/litter were randomly selected and necropsied, and 30/sex/dose were randomly selected and retained until adulthood (up to 17 weeks), with exposures continuing, with acquisition of VP, PPS, and estrous cyclicity evaluated. At necropsy of these retained adult F3 offspring, they were evaluated as described above for F0, F1, and F2 parental animals.
Statistical analyses.
The unit of comparison was the individual animal or the litter, as appropriate. Data from the cohorts were combined for summarization and statistical analyses. See Figure 2 for a graphical representation and reference citations of the decision trees employed for the statistical analyses. Quantitative continuous data (e.g., parental and pup body weights, organ weights, feed consumption, AGD, etc.) were compared among the 6 treatment groups and the vehicle control group. For the litter-derived percentage data (e.g., periodic pup survival indices), the ANOVA was weighted according to litter size. General Linear Models (GLM) analysis was used to determine the significance of the dose-response relationship and to determine whether significant dosage effects had occurred for selected measures. A one-tailed test was used for all pairwise comparisons to the vehicle control group, except that a two-tailed test was used for parental and pup body weight and organ weight parameters, feed consumption, percent males per litter, and AGD per sex per litter (Figure 2A).
|
Nonparametric tests for continuous data were used to determine if significant differences were present among the groups or to identify significant dose-response trends (Figure 2A
). Frequency data, such as reproductive indices (e.g., mating and fertility indices), were analyzed for differences among treatment groups and for pairwise comparisons (Figure 2B).
For acquisition of developmental landmarks (e.g., VP and PPS) and AGD, ANOVA and analysis of covariance (ANCOVA), with body weight (at birth, PND 0, for AGD; at acquisition of puberty and on study day [SD] 7 for females [VP] and SD 14 for males [PPS]; see Discussion) as the covariate, were used for pairwise comparisons (Figure 2C). For correlated data (e.g., body and organ weights at necropsy of weanlings, with more than 1 pup/sex/litter), SUDAAN® software was used for analysis of overall significance, presence of trend, and pairwise comparisons to the control group values (Figure 2D). For all statistical tests, the significance limit of 0.05 (one- or two-tailed) was used as the criterion for significance.
A test for statistical outliers (SAS, 1990b) was performed on parental body weights and feed consumption (in g/day) and parental and weanling offspring organ weights at necropsy. If examination of pertinent study data did not provide a plausible biologically sound reason (i.e., a reason that could not be ruled out as being within the possible range for the organ or measurement being made) for inclusion of the data flagged as "outlier," the data were excluded from summarization and analysis and were designated as outliers.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Parental Systemic Parameters 2
Body weights.
Adult systemic toxicity was evident for F0, F1, and F2 parental animals and F3 retained animals at 750 and 7500 ppm (50 and 500 mg/kg/day, respectively), including consistent and persistent reductions in body weights and weight gains in both sexes and in F0, F1, F2, and F3 generations. Body weights for F1 males and females, during the prebreed and mating periods, are presented in Figure 3
; these data are representative of all the generations evaluated. Body weights during gestation and lactation were significantly reduced in F0, F1, and F2 females at 7500 ppm, and during gestation and lactation at 750 ppm for F0 and F2 females and for F1 females during lactation (Fig. 4 for F1 females; these data are representative of all generations evaluated). Body weights at terminal sacrifice were significantly decreased in all generations at 7500 ppm, in F1 females at 750 ppm, and in F1 and F2 males at 750 ppm (Table 2). There were no toxicologically significant, consistent, or persistent effects on these parameters at 75 ppm (5 mg/kg/day) or below.
|
|
|
Feed consumption and BPA intake.
Feed consumption in g/day and g/kg/day was variable and showed no clear treatment-related effects (data not shown). The actual intake of BPA for both males and females throughout the study was 0, 0.0007–0.003, 0.015–0.062, 0.22–0.73, 4.1–15.4, 37.6–167.2, and 434–1823 mg/kg/day for the 0, 0.015, 0.3, 4.5, 75, 750, and 7500 ppm groups, respectively (Table 1
). The BPA intake was highest in the prebreed (in both sexes, Fig. 3) and lactation periods (in the females, Fig. 4).
Clinical observations.
There were no treatment- or dose-related clinical observations in either sex in any of the generations, except for transient evidence of dehydration at the start of the F0 prebreed in all groups, since some animals had difficulty adjusting to the automatic watering system, and at the start of the F1, F2, and F3 postwean (prebreed) exposure period at 7500 ppm, due to the small pups at this dose adjusting to the "nipples" of the automatic watering system, which was quickly resolved (data not presented).
Organ weights.
At necropsy, F0, F1, and F2 parental and F3 retained adult absolute nonreproductive organ weights were almost uniformly reduced for liver, kidneys, adrenal glands, spleen, pituitary, and brain at 7500 ppm (Table 2). Relative organ weights at 7500 ppm were typically significantly increased (or unaffected), with these effects most likely caused by reduced terminal body weights at this dietary dose. Changes in absolute and relative organ weights did occur rarely in other groups, but they were not consistent across generations and did not exhibit a dose-response pattern (Table 2).
Histopathology.
There were no treatment- or dose-related gross or microscopic findings for the examined organs for F0, F1, and F2 parental animals, and for F3 retained adults at any concentration for either sex, except for slight to mild renal tubular degeneration and chronic hepatic inflammation observed at a higher incidence in F0, F1, and F2 (but not F3) females at 7500 ppm (Tables 3 and 4).
|
|
Parental Reproductive Parameters
Absence of effects.
For absolute or relative reproductive organ weights, there were no treatment-related effects at any concentration for either sex for any generation, except for significantly reduced absolute and relative paired ovary weights as discussed below. There were no treatment- or dose-related direct effects in F0, F1, F2, and retained F3 males on absolute or relative weights of the testes, epididymides, prostate, or seminal vesicles plus coagulating glands (Table 2
).
There were no effects of treatment in F0, F1, or F2 females on mating, fertility, pregnancy, or gestational indices, dead pups per litter, or percent postimplantation loss (prenatal mortality index; Table 5). Estrous cycle length in days was equivalent across all groups for F0, F1, F2, and F3 females. Paired ovarian primordial follicle counts were similar between the high dose and control F1, F2, and F3 females (and increased at 7500 ppm for F0 females). Precoital interval in days and gestational length in days were equivalent across all groups for all generations.
|
There were no effects of treatment in F0, F1, or F2 males on mating or fertility indices (data not shown). Also, there were no effects on epididymal sperm concentration (except for a significant reduction in epididymal sperm concentration in F1 males, but not F0, F2, or F3 males, at 7500 ppm), percent motile or progressively motile sperm, testicular homogenization-resistant spermatid head counts, DSP (except for a significant reduction in DSP at 7500 ppm for F3 males only), or efficiency of DSP in any generation of males (Table 5
). Percent abnormal sperm was also unaffected for all F0, F1, F2, and F3 males in all groups. Occasionally and sporadically, 1 male (or rarely 2) in some groups, in all generations, exhibited low motility and high incidence of abnormal sperm. In every case for F0, F1, and F2 males (F3 males were not bred), the male sired a live litter.
There were no treatment-related gross or microscopic findings in reproductive organs for F0, F1, F2, or F3 adult males or females in any group (Tables 3 and 4).
Presence of effects.
The only significant effects were seen primarily in the 7500 ppm group in both sexes. There were significantly reduced absolute paired ovary weights in F0, F1, F2, and F3 females and relative paired ovary weights in F0, F1, and F2 (but not F3) females at 7500 ppm, in the presence of significant systemic maternal toxicity (Table 2). The only observed effects in F0, F1, F2, and retained F3 males were consistently reduced absolute organ weights and increased (or unaffected) relative organ weights, caused by the reduced terminal body weights of the males at 750 and 7500 ppm. The number of implants, total pups, and live pups per litter at birth and on PND 4 precull were significantly reduced at 7500 ppm (500 mg/kg/day) for F1, F2, and F3 offspring (Tables 5 and 6).
|
Offspring Parameters
Absence of effects.
There were no differences among groups for F1, F2, or F3 stillbirth index (data not shown), prenatal (postimplantation) loss per litter (Table 5
), sex ratio (% males) per litter at birth and throughout lactation (data not shown) or early postnatal (PND 0–4 precull) and lactational survival (PND 4 postcull–21) indices (Table 6). Interim offspring survival indices (PND 4–7, 7–14, and 14–21) were also unaffected (data not shown). In male offspring, there were no statistically significant effects on AGD, the number of nipples per pup, the number of areolae per pup, or the percent of pups with 1 or more nipples/areolae (Table 6).
Presence of effects.
Pup body weights per litter were reduced at 7500 ppm for F1, F2, and F3 offspring for the lactational period, measured on PND 7, 14, and 21 (Fig. 5). For F1 litters, pup body weights per litter were also significantly reduced in the high dose group on PND 4 for all pups analyzed together, but not for sexes analyzed separately.
|
In female offspring, AGD was significantly increased in the F2 generation at all dietary doses, with the exception of the 75 and 7500 ppm groups (Table 6
, Fig. 8). The absolute age at VP (days) was significantly delayed in the F1, F2, and F3 generations at 7500 ppm (and at 75 ppm only for the F2 generation). When the age at acquisition was adjusted for the body weight at acquisition, VP was delayed only at 7500 ppm for all 3 offspring generations. When the age at acquisition was adjusted for the body weight on SD 7, VP was delayed at 7500 ppm for the F1 and F3 generations and unaffected in the F2 generations (see Discussion; Table 6, Fig. 7).
|
|
In male offspring, the absolute age at PPS (days) was significantly delayed in the F1 generation at 750 and 7500 ppm, in the F2 generation at 0.3, 75, 750, and 7500 ppm, and in the F3 generation at 7500 ppm. When the age at acquisition was adjusted for the body weight at acquisition, PPS was delayed in the F1 generation at 750 and 7500 ppm and in the F2 and F3 generations at 7500 ppm. When the age at acquisition was adjusted for the body weight on SD 14, nothing changed (PPS was delayed at 750 and 7500 ppm for the F1 generation and at 7500 ppm for the F2 and F3 generations; see Discussion; Table 6
, Fig. 7). For F1, F2, and F3 weanling males and females sacrificed on PND 21, the absolute organ weights were decreased at 7500 ppm (the dietary concentration at which the terminal body weights were also decreased; data not shown). There were reductions in absolute organ weights at lower doses, but they were not consistently affected in F1, F2, and F3 weanlings or reproducible in specific dose groups. Relative organ weights were increased (or unaffected) at 7500 ppm (again, caused by reduced body weights at this dietary dose).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This study evaluated exposure of CD (SD) rats to BPA administered in the diet ad libitum at 0, 0.015, 0.3, 4.5, 75, 750, and 7500 ppm (approximate BPA intakes of 0, 0.001 [1 µg/kg/day], 0.02 [20 µg/kg/day], 0.3 [300 µg/kg/day], 5, 50, and 500 mg/kg/day, respectively) for 3 offspring generations, 1 litter per generation, through F3 adulthood. In the more than one million measurements made in this study, the number of significant, treatment-related effects was extremely small. The vast majority of parameters measured was within the concurrent and historical control ranges and, thus, are only briefly discussed. The focus of this discussion is to address those few endpoints that appeared to be either statistically significant and/or of possible biological significance.
Parental Systemic Parameters
Systemic toxicity effects in adult animals were limited to reductions in body weight, weight gain, and feed consumption in the top 2 doses (750 and 7500 ppm).
At 7500 ppm, there were consistent and persistent reductions in body weights and weight gains in both sexes and in F0, F1, F2, and F3 generations. Feed consumption in g/day and g/kg/day was variable and showed no clear treatment-related effects, nor were there treatment- or dose-related clinical observations in either sex in any generation. Body weights during gestation and lactation were significantly reduced in F0, F1, and F2 females at 7500 ppm, in F0 and F2 females at 750 ppm, and at 750 ppm in F1 females during lactation.
At necropsy, F0, F1, and F2 parental and F3 retained adult absolute organ weights were almost uniformly reduced for liver, kidneys, adrenal glands, spleen, pituitary, and brain at 7500 ppm. Relative organ weights at 7500 ppm were typically significantly increased, with both effects most likely caused by the reduced terminal body weights at this dietary dose. There were no treatment- or dose-related gross or microscopic findings for the examined organs in any parental animal, except for renal tubular degeneration and chronic hepatic inflammation observed at a higher incidence in F0, F1, and F2 (but not F3) females at 7500 ppm. There were no toxicologically significant effects on these parameters at 75 ppm or below.
Parental Reproductive Parameters
There were no effects of treatment in F0, F1, or F2 females on mating, fertility, pregnancy or gestational indices, dead pups per litter, or of percent postimplantation loss (prenatal mortality index). There were no treatment-related effects on absolute or relative reproductive organ weights, except for significantly reduced paired ovary weights (see below). Estrous cycle length in days was equivalent across all groups for F0, F1, F2, and F3 females. Paired ovarian primordial follicle counts were similar between the high dose and control F1, F2, and F3 females (but increased at 7500 ppm for F0 females). Precoital interval in days and gestational length in days were equivalent across all groups for all generations.
There were no effects of treatment in F0, F1, or F2 males on mating or fertility indices, or treatment- or dose-related direct effects in F0, F1, F2 and retained F3 males on absolute or relative weights of the testes, epididymides, prostate, or seminal vesicles plus coagulating glands. Also, there were no effects on epididymal sperm concentration (except for a significant reduction in epididymal sperm concentration in F1 males, but not F0, F2, and F3 males, at 7500 ppm), percent motile or progressively motile sperm, testicular homogenization-resistant spermatid head counts, DSP (except for a significant reduction in DSP at 7500 ppm for F3 males, but not F0, F1, or F2 males, with no effect on efficiency of DSP), or efficiency of DSP. Percent abnormal sperm was also unaffected for all F0, F1, F2, and F3 males in all groups. The slightly higher (but not statistically significant) values for F2 males at 0.015, 0.3, 4.5, and 75 ppm and for F3 males at 0.015 and 75 ppm were due to 1 or 2 males per group with few or no motile sperm and most or all abnormal sperm. In all cases for the F2 males, the affected males sired live litters (F3 males were not bred). There were no treatment-related gross or microscopic findings on reproductive organs for F0, F1, F2, or F3 adult males or females.
The vast majority of the relatively few effects observed for parental reproductive parameters occurred only at the highest dose of 7500 ppm. The number of implants, total pups, and live pups per litter at birth and on PND 4 precull were significantly reduced at 7500 ppm for F1, F2, and F3 offspring.
The explanation for the reduced live litter size at birth at 7500 ppm for F1, F2, and F3 offspring is not known. It is not due to the male since there is no evidence of reproductive effects on the males at 7500 ppm (or any other dietary dose), nor is it due to prenatal postimplantation loss of conceptuses, since postimplantation loss was unaffected at any dose for F0, F1, and F2 dams carrying F1, F2, and F3 litters. Preimplantation loss cannot be determined from this study design since, by the time the parental females are sacrificed, the ovarian corpora lutea of pregnancy (which form after ovulation) have involuted to corpora albicans, indistinguishable from corpora albicans from previous ovulation cycles. Although the absolute and relative paired ovarian weights were reduced in F0, F1, F2, and F3 (absolute only) females in the present study, there was no evidence of reduced ovarian primordial follicle counts at 7500 ppm in any generation, even in the presence of significant systemic maternal toxicity.
There were no significant histopathological findings for any reproductive organ in either sex at any dose in any generation.
Offspring Parameters
As in the parental animals, the vast majority of the relatively few effects observed for offspring parameters occurred only at the highest dose of 7500 ppm.
Body weights.
Effects on body weights were also observed in offspring only at 7500 ppm, beginning on PND 7 and continuing through lactation, weaning, and the postweaning period to adulthood in all 3 generations (F1, F2, and F3). The reduced body weight in periweanlings at 7500 ppm and in older animals at 750 and 7500 ppm, in all generations, was most likely the cause of the reduced absolute organ weights in F1, F2, and F3 weanlings; F1, F2, and F3 adults; and consistent with the increased (or absence of an effect on) relative organ weights at these dietary doses.
Organ weights.
Absolute and relative organ weight data for F1, F2, and F3 weanling (PND 21) pups indicate that for all but the ovaries, the absolute organ weights were reduced, and the relative organ weights were increased or unaffected in all groups, including 750 and 7500 ppm, at which postweaning body weights were significantly reduced. For paired ovary weights, the effects in the F1, F2, and F3 female weanlings at 7500 ppm paralleled effects observed in the F0, F1, and F2 adult females (both absolute and relative weights were reduced) and in the F3 adult females (only absolute ovary weights were reduced). For F1, F2, and F3 males and females, the absolute organ weight changes were decreased at 7500 ppm (the dietary concentration at which the terminal body weights were also decreased). At 7500 ppm, there were reductions in absolute and relative paired ovarian weights (absolute in F1, F2, and F3 females; relative in F1 and F2, but not F3, females). Statistical analysis of ovary weight, covaried by body weight at necropsy (Fig. 6A), indicated consistent effects only at 7500 ppm. Similarly, testis weight covaried by body weight at necropsy (Fig. 6B) indicated effects only in the F3 generation (the generation not mated), with no effects in the F0, F1, or F2 generation males.
|
Some of the effects on absolute or relative organ weights in the F1, F2, and F3 weanlings were also present in the corresponding adults, including absolute and relative paired ovary weights that were significantly reduced in the weanling and adult F1, F2, and F3 females at 7500 ppm. The patterns of absolute and relative weights show that test material-related effects exist only in the highest dose group and only for paired ovaries.
In accordance with current thinking on absolute and relative organ weights, when terminal body weights are reduced, only those organ weight parameters that exhibit statistically significant differences in the same direction for both absolute and relative values are considered biologically important and directly treatment related. Therefore, the changes in relative F1, F2, and F3 male and female weanling organ weights were not considered to be biologically significant and were most likely secondary to the decreased body weights. The difference in effects on weanling versus adult animals is likely the result of the very high dietary intakes (greater than 750 [786–1205] mg/kg/day) of the test material being consumed by the weanling animals in the high dose group (in their first postwean exposure week). These intakes are approximately 1.5 to 2 times the daily intakes in the adults at the same dose (less than 575 [534–570] mg/kg/day, during the last prebreed exposure week immediately prior to mating). BPA intakes in Table 1 and for F1 postwean males and females in Figures 3B, 3D, and 4 indicate that the maximum exposure to BPA, in mg/kg/day for all animals during all phases of the study, was during the first week of prebreed for the pups (when they were small and in a growth spurt) and the last week of lactation for the dams (where there could have been a substantial contribution from pups self feeding).
Total and live pups per litter.
Also, only at 7500 ppm were there reduced total and live litter sizes on PND 0. The explanation for the reduced number of total and live pups per litter at birth at 7500 ppm for F1, F2, and F3 offspring is not known. It is not due to effects on males, since there is no evidence of reproductive effects on the males at 7500 ppm (or any other dietary dose). It is not due to prenatal postimplantation loss of conceptuses, since postimplantation loss was not affected at any dose for F0, F1, and F2 dams carrying F1, F2, and F3 litters. Preimplantation loss cannot be determined from this study design, since by the time the parental females are sacrificed (at the weaning of their litters), the ovarian corpora lutea of pregnancy (which form after ovulation) have involuted to corpora albicans, indistinguishable from corpora albicans from previous ovulation cycles. The possibilities, therefore, exist that there were increased preimplantation loss and/or fewer eggs ovulated at 7500 ppm. Although the absolute and relative paired ovarian weights were reduced in F0, F1, F2, and F3 (absolute only) females in this study, there was no evidence of reduced ovarian primordial follicle counts at 7500 ppm in any generation.
Biegel et al. (1998a) also reported reduced live litter sizes associated with reduced number of implantations per litter (the latter was not observed in the present study) at 2.5 ppm dietary E2, but did not offer an explanation. In the present study, dams at 7500 ppm exhibited profound reductions in body weight and weight gain, which is at least consistent with effects of profound maternal systemic toxicity as causative per se.
Acquisition of Pubertal Characteristics
VP and PPS.
Reduced body weights are also most likely the cause of the significant delay in acquisition of puberty in both sexes (age at acquisition of VP in females and PPS in males), observed in all offspring generations at 7500 ppm, using ANCOVA with body weight at acquisition as the covariate. Analysis of the ages at acquisition alone (by nonparametric Kruskal-Wallis and Mann-Whitney U tests) did result in significant delays at lower doses (rarely and not consistently), but the values were not significant with ANCOVA (Fig. 7).
Body weight at acquisition was significantly reduced at 7500 ppm for F1 males and females and for F2 and F3 males with ANCOVA. However, acquisition of developmental landmarks is dependent on both age and weights, i.e., heavier animals acquire the landmark earlier, while lighter animals acquire the landmark later. However, lighter animals do eventually acquire the landmark (unless there is another cause for the delay) and in many cases acquire the landmark at the same or lighter weight than the heavier animals, but at an older age (e.g., Carney et al., 1998; Kennedy and Mitra, 1963). All animals in this study acquired puberty. The lighter animals acquired puberty at a later time (older age). Most of them were comparable in weight at the time of acquisition to the control (and lower dose groups) animals that acquired puberty at an earlier time (and thus, a younger age).
There is much discussion among reproductive toxicologists as to the most appropriate body weight to use as a covariate for ANCOVA other than that at acquisition. We covaried age at acquisition of puberty both by the body weight at acquisition (to standardize pup weights to the same physiologic state; i.e., puberty regardless of age) and by the body weight on a prebreed study day, encompassing the time of acquisition (i.e., SD 7 for females and SD 14 for males) to standardize pup weights to the same age, regardless of physiologic state.
To use the body weights on SD 7 and 14 as the covariate for age at acquisition of puberty, we established that the ages for each group on the chosen study day, within each generation, were equivalent (by ANOVA), and that the variances (i.e., the distributions) also did not differ (by Levene's test). The results for the covariate analyses by body weight on SD 7 (females) or SD 14 (males) are presented in Table 6. The ANCOVA analyses resulted in essentially the same findings, regardless of which body weight was employed as the covariate, which is consistent with the reduced body weights in both sexes in all 3 offspring generations throughout their respective prebreed periods at 7500 ppm.
Since acquisition of both landmarks in both sexes of both generations was delayed, these results are probably not caused by estrogen receptor-mediated events or other endocrine-related toxicity. The only endocrine-mediated mechanism currently known to result in delays in puberty in both sexes would be interference with steroidogenesis, thereby reducing testosterone (and DHT) levels in males and estrogen levels in females, and there is no evidence that BPA interferes with steroidogenesis in rats. It is most likely that the delays in puberty in both sexes at 7500 ppm were caused by reduced body weights prior to and at acquisition in all offspring generations.
This interpretation is consistent with the recognition by the U.S. EPA (1996, p. 56295) that "body weight at puberty may provide a means to separate specific delays in puberty from those that are related to general delays in development." The delays in VP in females and in PPS in males at 7500 ppm in this study were relatively minor: 2.5 (F1, F3) and 3.5 (F2) days for females and 3.1 (F3), 3.9 (F1), and 5.8 (F2) days for males (the delay in acquisition in F1 males at 750 ppm was 1.7 days). Biegel et al. (1998a,b) have shown that dietary administration of E2 at 0.05 and 2.5 ppm resulted in accelerated VP (by 7 days) in CD (SD) rats.
AGD.
The significant effect on acquisition of reproductive landmarks in F1 and F2 offspring required a measurement of AGD in newborn F2 and F3 offspring, as specified in the guidelines (U.S. EPA, 1998). AGDs in newborn F2 and F3 males were statistically equivalent across all groups at PND 0. In the newborn F2 females, AGD was statistically significantly longer at 0.015, 0.3, 4.5 (not 75), and 750 (not 7500) ppm, with mean values of 0.98, 0.98, 0.98, and 0.99 mm, respectively, relative to the control group mean value of 0.95 mm (and 0.97 mm at 75 ppm and 0.96 mm at 7500 ppm; Figs. 8A and 8B). These effects were increases in AGD of only 0.03–0.04 mm, equivalent to increases of only 3.16–4.21%. They were also present only at doses where the mean F2 female body weights per litter were slightly, but not statistically significantly, higher than in the control group and in the groups with unaffected AGDs; body weights, per se, are known to affect AGD (Gallavan et al., 1999). These small differences (0.03–0.04 mm), especially since the AGDs for F3 female pups in all groups were statistically equivalent, are considered of no biological significance because the magnitude of the differences is minimal, all mean values round to 1.0 mm, and these changes, along with the similarly minor delays in acquisition of PPS and VP, are not associated with any alterations in reproductive organ structures or function in the animals exhibiting them. In addition, AGD is under androgenic control, specifically dihydrotestosterone (Gray et al., 1998; Gray and Ostby, 1998) and is not affected by estrogens (Biegel et al., 1998a). BPA was shown to be neither an androgen nor antiandrogen in vivo (Laudenbach et al., 2001). Therefore, the effects reported on F2 female AGD are considered of no biological significance and not due to BPA exposure.
Comparisons across Generations
One of the possible analyses that can be done with a multigeneration dataset is to characterize an effect (or lack thereof) across generations. This is permitted if 2 important statistical criteria are met: the control groups are not statistically different and there is no interaction between dose and generation, i.e., a dose x generation interaction.
To determine if it was appropriate for data across generations to be pooled, a two-way ANOVA was performed for organ weights (such as epididymides, testis, and ovaries), developmental landmarks (VP and PPS), and AGD. The results showed that several parameters could be pooled and several could not be pooled. For testis and epididymides weights, there were significant differences between F1 and F3 controls (p = 0.0004) and between F2 and F3 controls (p = 0.0001), respectively. This is understandable since the F3 animals were younger, had less total exposure duration to BPA, and had never been mated.
For daily sperm production, there were significant differences between the F1 controls and both F2 (p = 0.0048) and F3 (p = 0.0005) controls. This was understandable since the F1 generation controls had a lower epididymal sperm concentration but a higher spermatid head concentration than the other generations, which caused the DSP and efficiency of daily sperm production to be higher than the other generations. For ovary weight, the results (p = 0.0007) of the ANOVA for interaction showed that there was a significant dose x generation effect. Based on these results, the generations could not be pooled for epididymides, testis, or ovary weights or for DSP. Thus, the only statistically valid comparisons for these parameters were between doses for each individual generation and its concomitant control.
For PPS, VP, and AGD for both males and females, there were no statistical differences among control groups (for PPS, p = 0.8797; for VP, p = 0.1848; for AGD male, p = 0.7262; for AGD female, p = 0.3181). This is also understandable since the animals were all covaried with body weight at the same study day (SD 7 for females and SD 14 for males), were statistically the same age on that date (no statistical differences in the distribution of ages among groups), and had the same range of total exposure durations to BPA across groups. Thus, the data for these 3 parameters could be pooled.
The pooled data (n = 385 litters; dose and dose x generation df = 6) from both F2 and F3 generations for male AGD showed the same results as did the individual generations when compared to their concomitant controls (i.e., there were no effects of BPA at any dose in any generation or across generations). For pooled female AGD (n = 385 litters; dose and dose x generation df = 6), none of the values differed more than 0.04 mm from the control value, although 3 were statistically significant (at 0.3, 750, and 7500 ppm). None of the individual values for female AGD for the F3 generation differed more than 0.04 mm from the control value (as with the pooled values), and none of these 6 dose group values were statistically different from its concomitant control value. None of the F2 values differed more than 0.04 mm from the control values, yet 4 of the 6 doses were statistically significant. Biologically, the difference of 0.04 mm is insignificant and most likely due to the exceptionally well-controlled micrometric measurement techniques for AGD (all standard errors were within 0.02 mm of the mean).
The pooling (n = 626; dose and dose x generation df = 6) of the PPS data across all 3 offspring generations created a statistically significant difference at 0.3 ppm, which was not present in any of the individual generations. No other dose below 750 ppm in any generation or in the pooled data was significant. Thus, this finding was considered to be an anomaly and not biologically meaningful.
The pooled data (n = 627; dose and dose x generation df = 6) from all 3 offspring generations for VP showed the same results as the data from the individual generations when compared to their concomitant controls, i.e., there were no effects of BPA at any dose in any generation or across generations other than at the highest dose of 7500 ppm.
Based on the results of the statistical tests for the pooled data, pooling the data did not lend any more insight into interpretation of the data than did just comparing individual generations with their concomitant controls.
Other Research and Routes of Exposure in Rat Reproductive Toxicity Evaluations
Absence of effects.
Welsch and colleagues (Elswick et al., 2000; Welsch et al., 2000, 2001) reported that exposure of CD (SD) female rats (13–16 pregnancies/group) to BPA in drinking water at 0, 0.005, 0.5, 5, or 50 mg/l from GD 2 through PND 21 (with intakes of 0.001 to 10 mg/kg/day) resulted in no effects on differentiation and function of the reproductive system in female (Welsch et al., 2000) or male (Elswick et al., 2000) F1 offspring when evaluated through 10 months of age. In F1 females, there were no effects of BPA on fertility, fecundity, organ weights, AGD, VP, age at first estrus, estrous cyclicity, ovarian follicle counts, or lordosis. The positive control DES (at 0.05 mg/l) did cause accelerated VP and age at first estrus in females. In F1 males, there were no effects of BPA on AGD, PPS, organ weights, hormone levels, sperm counts, fertility, immunohistochemically measured ventral prostate AR levels, and no treatment-related histopathological changes.
Other researchers have also reported no effect of exposure to BPA at low doses. Kwon et al. (2000) administered BPA by gavage to pregnant CD (SD) rats at 0, 3.2, 32, or 320 mg/kg/day from GD 11 through PND 20. DES at 15 µg/kg/day was employed as a positive control. Offspring female pubertal development was unaffected by indirect BPA exposure at any dose. There were also no effects on the volume of the sexually dimorphic nucleus of the preoptic area (SDN-POA) of the brain in 10-day-old offspring females, on estrous cyclicity, on sexual behavior of the offspring females at 4 months of age, or on offspring male reproductive organ weights at 6 months of age (including testis, epididymis, seminal vesicle, and ventral and dorsolateral prostate lobes). DES increased the volume of the SDN-POA in offspring females and caused irregular estrous cyclicity.
Ema et al. (2001) from the Chemical Compound Safety Research Institute in Hokkaido, Japan, administered BPA in distilled water by gavage (stomach tube) to Crj:CD (SD) rats, 25/sex/dose at 0, 0.2, 2, 20, and 200 µg/kg/day. This study, like the present study, was conducted under GLP regulations and was compliant with the U.S. EPA testing guidelines (U.S. EPA, OPPTS, 837.3800, 1998). The Ema study also included endocrine-sensitive measurements and neurobehavioral endpoints, evaluating functional development in F1 and F2 offspring (open field motor activity and Morris water maze learning and memory tests), and various serum hormone concentrations in F0 and F1 parental animals and retention of F2 weanlings. Thirty-seven animals per sex per group (25 from the main study and 12 from "satellite groups") were evaluated until adulthood, including gross necropsy, organ weights, and histopathology for F2 males and estrous cyclicity and gross necropsy for F2 females. The authors concluded that "oral doses of BPA of between 0.2 and 200 µg/kg administered over two generations did not cause significant compound-related changes in reproductive or developmental parameters in rats" (Ema et al., 2001, p. 522). These conclusions are supported by our findings of no biologically relevant effects of BPA below 5 mg/kg/day in any generation in either sex.
Nagao et al. (1999) administered BPA by sc injection to rat pups on PND 1–5 at 300 µg/g (300 mg/kg/day) and reported no effects on male or female reproductive development or on adult offspring reproductive structures or functions. Estradiol benzoate was also administered by the same route and timing to a separate group and caused clear effects in male and female reproductive development and in reproductive structures and functions.
Atanassova et al. (2000) and Williams et al. (2001) showed that Wistar rats treated neonatally with a range of doses (0.01–10 µg; equivalent to 1.0 µg–1.0 mg/kg in a 10-g neonatal rat pup) of DES on alternate days from PND 2 to 12 developed a dose-dependent retardation of pubertal spermatogenesis on day 18, as evidenced by decreases in testis weight, seminiferous tubule lumen formation, spermatocyte nuclear volume per unit Sertoli cell, and elevation of the germ cell apoptotic index. The 2 lowest doses of DES (0.1 and 0.01 µg) significantly increased spermatocyte nuclear volume per unit Sertoli cell. Similarly, daily treatment on PND 2–12 with BPA (0.5 mg; equivalent to 50 mg/kg in a 10 g neonatal rat pup) significantly advanced this and some of the other aspects of pubertal spermatogenesis. In adulthood, testis weight was decreased dose dependently in rats treated neonatally with DES, but only the lowest dose group (0.01 µg) showed evidence of mating (3 of 6) and normal fertility (3 litters). Animals treated neonatally with BPA had increased testis weights and exhibited "reasonably normal" mating/fertility (Attanassova et al., 2000, pp. 3898 and 3904). The authors concluded that the effect of high doses of BPA on the first wave of spermatogenesis at puberty was "essentially benign" (Attanassova et al., 2000, pp. 3898 and 3908). Furthermore, this group concluded that weak environmental estrogens in general are "unlikely to pose a significant risk to the reproductive system of the developing human male unless the compound in question also possesses some other biological activity of relevance" (Williams et al., 2001, p. 245).
Rubin et al. (2001) reported no effects on the number of pups per litter, sex ratio, day of VP, AGD, and no significant histopathological findings in offspring of rats (SD) exposed to BPA in drinking water to approximately 0.1 and 1.2 mg/kg/day from GD 6 through lactation. Kubo et al. (2001) showed that a BPA dose of 1.5 mg/kg/day in drinking water to 10 female Wistar rats during pregnancy and lactation produced no differences between organ weights (testis, epididymis, ventral prostate, ovaries, uterus) and serum hormone levels (LH, FSH, testosterone, or 17ß-estradiol) in offspring at 12 weeks of age, when compared to the control group values. Ramos et al. (2001) reported no effects on litter size, male or female pup body weight, sex ratios, or AGD following exposure to 25 µg/kg/day and 250 µg/kg/day of BPA dissolved in DMSO administered by continuous sc infusion via osmotic pump from GD 8 to PND 23 to pregnant Wistar rats (4 dams/group).
The data presented above from other laboratories for BPA administered by various routes of exposure are consistent with the findings from the present study, which indicated no effects of BPA at 0.001–5 mg/kg/day (i.e., at low doses) when administered in the feed.
Presence of effects.
Still others have reported effects of BPA exposure in rats, usually by some nonstandard means of dose delivery, such as via continuous sc infusion by implanted osmotic minipumps or at relatively high doses. Steinmetz et al. (1997) exposed Fischer (F344) and SD rats to BPA (approximate dose of 220–225 µg/kg/day) or E2 (approximate dose of 6–7.5 µg/kg/day), using silastic implants for 3 days. With BPA, F344 rats showed an increase in serum prolactin levels and hyperprolactinemia but showed no effect on anterior pituitary weight. There were no effects on either endpoint with the SD rat. E2 produced hyperprolactinemia in both strains of rat, but produced an increase in anterior pituitary weight in only the F344 rat. Stoker et al. (1999) reported that BPA, given to male (prepubertal) Wistar rat pups on PND 22–32 by sc injections of 0 or 50 mg/kg once daily, stimulated increased secretion of prolactin during the dosing period and increased mean lateral prostate weight and inflammation of the lateral lobes of the prostate at 120 days of age. Tohei et al. (2001) also reported increased serum prolactin and increased plasma concentrations of luteinizing hormone in male Wistar rats exposed to 1 mg/kg BPA via sc injection for 2 weeks.
Chahoud and his colleagues (Fialkowski and Chahoud, 2000; Schönfelder et al., 2001; Talsness and Chahoud, 2000; Wu and Chahoud, 2000) exposed pregnant SD rats to BPA by gavage on GD 6–21 at doses of 0, 0.02, 0.1, and 50 mg/kg (11–20 litters/group) and reported various effects of sexual development in the offspring. However, the NTP Environmental Disrupters Low-Dose Peer Review Statistics Subpanel indicated that a "severe design deficiency" of absence of a concurrent control group precluded "statistical reanalysis of the data" [and] "any reliable assessment of the effects" reported for BPA by this group (NTP, 2001, Appendix A, p. A-58).
Rubin et al. (2001) reported increased body weight gain in offspring of SD rats exposed to BPA in drinking water to approximately 0.1 and 1.2 mg/kg/day from GD 6 through lactation. They also reported altered patterns of estrous cyclicity and lowered plasma LH levels in the high-dose BPA group. Kubo et al. (2001) showed that a BPA dose of 1.5 mg/kg/day in drinking water to 10 female Wistar rats during pregnancy and lactation produced similar results at weeks 6 and 7 in offspring (5/sex/litter) for movement, passive avoidance patterns, and size of the locus coeruleus (7 rats total/sex for BPA, 6 rats/sex for control at week 20). In the control group, females showed a higher activity, lower avoidance memory, and larger locus coeruleus than the males.
Ramos et al. (2001) reported that both 25 µg/kg/day and 250 µg/kg/day of BPA, dissolved in DMSO administered by continuous sc infusion via osmotic pump from GD 8 to PND 23 to pregnant Wistar rats (4 dams/group), produced an effect on the proliferation and differentiation of epithelial and stromal cells in the ventral prostate (up to 4 rats/litter). This was expressed as an increase in the fibroblast:smooth muscle cell ratios and a decrease in the AR-positive cells of the periductal stroma.
Takahashi and Oishi (2001) reported that young (4 weeks of age) F344 rats (8/group), given dietary concentrations of 0, 234, 466, and 950 mg/kg/day of BPA for 44 days, had decreased body weight, food consumption, and liver weight at 466 and 950 mg/kg/day and increased kidney weight at all 3 doses. Seminal vesicle and dorsolateral prostate gland weights were decreased at only 950 mg/kg/day, and seminal vesicle weight was decreased at all BPA doses. Although there were no effects on testis, epididymides, or ventral prostate gland weight at any BPA dose, histopathological examination of the testes revealed seminiferous tubule degeneration and loss of elongated spermatid in a dose-dependent fashion at all doses.
Many of the effects observed by the above authors were from BPA administered by various parenteral (non-oral) routes,