ToxSci Advance Access originally published online on April 29, 2008
Toxicological Sciences 2008 104(2):362-384; doi:10.1093/toxsci/kfn084
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Two-Generation Reproductive Toxicity Study of Dietary Bisphenol A in CD-1 (Swiss) Mice
* Health Sciences Unit, RTI International, Research Triangle Park, North Carolina 27709
Experimental Pathology Laboratories, Inc., Research Triangle Park, North Carolina 27709
SABIC Innovative Plastics, Pittsfield, Massachusetts 02101
Toxicology/Regulatory Services, Inc., Charlottesville, Virginia 22911
¶ Bayer Material Science, Pittsburgh, Pennsylvania 15205
|| Bayer Healthcare AG, Wuppertal, Germany D-42096
||| American Chemistry Council, Arlington, Virginia 22209
|||| The Dow Chemical Co., Midland, Michigan 48674
1 To whom correspondence should be addressed at RTI International, 3040 Cornwallis Road, P.O. Box 12194, Hermann Laboratory Bldg., Research Triangle Park, NC 27709-2194. Fax: (919) 541-5956. E-mail: rwt{at}rti.org.
Received January 21, 2008; accepted April 15, 2008
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Dietary bisphenol A (BPA) was evaluated in a mouse two-generation study at 0, 0.018, 0.18, 1.8, 30, 300, or 3500 ppm (0, 0.003, 0.03, 0.3, 5, 50, or 600 mg BPA/kg/day, 28 per sex per group). A concurrent positive control group of dietary 17β-estradiol (0.5 ppm; 28 per sex) confirmed the sensitivity of CD-1 mice to an endogenous estrogen. There were no BPA-related effects on adult mating, fertility or gestational indices, ovarian primordial follicle counts, estrous cyclicity, precoital interval, offspring sex ratios or postnatal survival, sperm parameters or reproductive organ weights or histopathology (including the testes and prostate). Adult systemic effects: at 300 ppm, only centrilobular hepatocyte hypertrophy; at 3500 ppm, reduced body weight, increased kidney and liver weights, centrilobular hepatocyte hypertrophy, and renal nephropathy in males. At 3500 ppm, BPA also reduced F1/F2 weanling body weight, reduced weanling spleen and testes weights (with seminiferous tubule hypoplasia), slightly delayed preputial separation (PPS), and apparently increased the incidence of treatment-related, undescended testes only in weanlings, which did not result in adverse effects on adult reproductive structures or functions; this last finding is considered a developmental delay in the normal process of testes descent. It is likely that these transient effects were secondary to (and caused by) systemic toxicity. Gestational length was increased by 0.3 days in F1/F2 generations; the toxicological significance, if any, of this marginal difference is unknown. At lower doses (0.018–30 ppm), there were no treatment-related effects and no evidence of nonmonotonic dose-response curves for any parameter. The systemic no observable effect level (NOEL) was 30 ppm BPA (
5 mg/kg/day); the reproductive/developmental NOEL was 300 ppm (50 mg/kg/day). Therefore, BPA is not considered a selective reproductive or developmental toxicant in mice. Key Words: bisphenol A; CAS No. 80-05-7; 17β-estradiol, CAS No. 50-28-2; CD-1 Swiss mice; two-generation reproductive toxicity study; OECD 416; puberty; andrology.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Bisphenol A (BPA) is a high-production volume chemical used primarily in manufacturing polycarbonate plastics/epoxy resins. Human exposure to low doses of BPA occur mostly through food contact uses, with polycarbonate plastics and epoxy resins used in plastic bottles and coatings of food and beverage containers (the "European Panel on Food Additives, Flavourings, Processing Aids and Materials in Contact with Food" [EFSA, 2006
]). There is scientific debate over whether BPA is able to cause reproductive or developmental effects in animal studies, other than effects at high doses that have been associated with parental systemic toxicity.
Reviews by Gray et al. (2004), Goodman et al. (2006), EFSA (2006), Rhomberg et al. (2007), the Center for Evaluation of Risks to Human Reproduction (CERHR, 2007), and Willhite et al. (2008) examined the BPA mammalian toxicology literature on reproduction, including large studies (meeting or exceeding guidelines) by the oral route, as well as small studies that utilized both oral and nonoral (parenteral) routes of administration. Some of the studies with small numbers of animals/group, and/or a limited number of dose levels, reported effects in rodents such as prostate weight changes, reduced sperm count, diminished reproductive function, acceleration of puberty in females, increased body weight, as well as nonmonotonic dose-response curves in rats and/or mice (e.g., Goodman et al., 2006). The reviewers also examined numerous studies with robust study designs that were unable to confirm the results of these smaller studies. All of the reviewers’ evaluations cited above were consistent with the conclusion that BPA is not a selective reproductive or developmental toxicant.
A key, large study with BPA administered via the oral route to rats was reported by Tyl et al. (2002) in which CD (Sprague Dawley [SD]) rats were exposed to dietary BPA at 0, 0.015–7500 ppm (0, 0.001–500 mg/kg/day BPA) for three generations. The adult systemic toxicity no observable adverse effect level (NOAEL) was 75 ppm (5 mg/kg/day), and reproductive/postnatal toxicity postnal day NOAEL was 750 ppm (50 mg/kg/day). There were no treatment-related effects at low doses (0.001–5 mg/kg/day) and no evidence of nonmonotonic dose-response curves on any parameter across generations for either sex. Therefore, Tyl et al. (2002) concluded that BPA was not a selective reproductive toxicant, and that there were no effects at low oral doses. Ema et al. (2001) also confirmed the lack of reproductive effects of BPA using Crj:CD(SD) rats by oral gavage at doses of 0, 0.2–200 µg/kg/day for two generations.
Published work on BPA has been criticized (e.g., by CERHR, 2007; Long et al., 2000; Spearow and Barkley, 2001; vom Saal, 2007; vom Saal and Hughes, 2005) for lack of low doses, lack of and/or failure of positive controls, use of a less sensitive species and/or strain, inappropriate routes of exposure, and improper statistical analyses. These concerns have been addressed in the present study, which employed very low to very high dietary doses, an effective positive control, appropriate statistical analyses, an OECD Guideline protocol (TG 416, enhanced; OECD, 2001) conducted under Organisation for Economic Cooperation and Development (OECD) Good Laboratory Practice (GLP) Principles (OECD, 1998, 2002), and used the CD-1 (Swiss) mouse.
However, the mouse has not been a species of choice in guideline reproduction studies. Therefore, dietary E2 one- and two-generation studies were first performed to evaluate the potential of a known estrogen, administered in the feed to CD-1 (Swiss) mice, to produce possible alterations in parental fertility, maternal pregnancy and lactation, and systemic and developmental toxicity to offspring, and to develop a baseline (historical positive and negative databases) by which to characterize relevant, estrogen-sensitive parameters in CD-1 mice (Tyl et al., 2008a, b).
The present study employed an enhanced OECD protocol under OECD GLP and dietary BPA exposure of mice to six dose groups (28 per sex per group) at 0.018–3500 ppm, 0.003–600 mg/kg/day, to investigate the sensitivity of this species to BPA. The study was conducted as part of the European Union (EU) Risk Assessment, in cooperation and with the oversight of a EU Bisphenol A Steering Group (a group of expert reproductive/developmental toxicologists from EU member countries). It was designed to determine if effects in mice occurred in the low-dose range for BPA and to examine the possibility of nonmonotonic dose responses. It exceeded guideline requirements as there were more dose groups than required, more animals/group than required, additional histopathology (including on weanling tissues), additional F1 males retained postweaning (1/litter in all groups), and two negative control groups (each with 28 per sex) to address biological variability. In addition, a positive control group (0.5 ppm E2; 0.080 mg/kg/day; 28 per sex) was used to confirm mouse sensitivity to a known estrogen.
|
OBJECTIVES |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
The objectives of this study were to (1) evaluate the potential of BPA, administered in the feed to CD-1 (Swiss) mice, to produce parental and offspring systemic toxicity and/or alterations in fertility or pregnancy, or to affect the growth and development of offspring over two generations from low doses (µg/kg body weight/day) through doses that produced adult systemic toxicity (high milligram/kg/day doses); (2) identify the primary target organs and effects of BPA in mice in a multigeneration reproductive toxicity study; and (3) confirm the sensitivity of the mouse model to a known estrogen by the administration of E2 in the feed at 0.5 ppm to serve as a positive control group.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Test Substance and Dose 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 white crystalline solid. Analytical characterization of the bulk BPA confirmed greater than 99.7% purity throughout the study. E2 (CAS No. 50-28-2, Batch No. 021K1267, Supplier Product No. E8875) was received from Sigma-Aldrich (St Louis, MO); purity
99.0%. The vehicle was Purina Certified Ground Rodent Diet, No. 5002 (PMI Feeds, Inc., St Louis, MO). The diets were formulated using stock solutions of BPA or E2 in acetone, diluted with acetone to the appropriate concentrations, and then mixed with feed to make the premixes. The premixes were air dried under a hood in stainless steel pans using stainless steel "rakes" to ensure dryness and to break up any clumps. The premixes were then mixed with additional feed to produce the specified dietary concentrations in the required quantities. The vehicle control group feed also had an acetone premix. The homogeneity and stability of BPA (0.015 and 7500 ppm) and E2 (0.5 ppm) in the dosed feed were assessed and confirmed at room temperature for at least 9 days and at least 50 days when frozen at –20°C. The dosed feed was formulated, based on the stability data, at least monthly and stored frozen. The feed in the animals’ feeding jars was changed at least every 7 days.
Analyses of the negative control feed, as well as feed containing BPA or E2, were conducted on all formulations for all formulation dates. Standards for acceptable accuracy of mixing were mean concentrations of the analyzed samples within ± 20% of nominal, and the % relative standard deviation for triplicate analyses 
15%. Only formulations satisfying the acceptability standards were administered to the animals. The analytical range for all dietary concentrations used for all formulation dates was 80.8–120% of nominal for BPA and 89.0–118% of nominal for E2, with an estimated limit of detection of 0.002 ppm for BPA and 0.00023 ppm for E2.
Animals and Husbandry
Two hundred and eighty virgin female mice and male mice (Cesarean-originated VAF Crl:CD-1 (ICR)BR outbred albino; known as the Charles River CD-1 mouse) were ordered for the study in two shipments, 140 per sex per shipment, for arrival 8 days apart. Animals from the first shipment were designated as Cohort 1, and animals from the second shipment were designated as Cohort 2. Use of cohorts was necessary due to logistical constraints (animal room sizes, staff, and scheduling of necropsies). Each cohort on study consisted of 14 animals per sex per group with all groups represented. Of the 140 animals per sex per shipment ordered, five per sex were used as quality controls for assessment of pathogen antibody status within 1 day after receipt. The remaining animals were quarantined for 1 week upon arrival. Based on the results of the physical examination, serology, and parasitology, the animals were in good health and were used in this study. All antibody titer assessments for quality control animals were negative. Four per sex per cohort were used as sentinels to monitor the health status of study animals, two per sex per cohort for evaluation of antibody titers at the F0 necropsy and the remaining sentinels, up to two per sex per cohort, for the same analyses at the F1 parental and F2 weanling necropsies; five per sex were used to replace any animals inappropriate for use; and 126 per sex per cohort went on study. All pathogen antibody titer assessments for sentinel animals were negative.
F0 animals, prior to initiation of exposure, and F1 offspring selected to be parents of the F2 generation or selected to be F1 retained males, were identified by uniquely numbered ear tags (National Band and Tag Co., Newport, KY). F1 pups, prior to selection at weaning, were not uniquely identified, except for F1 females selected for parents of the next (F2) generation in each litter, because examination for VP began on postnatal day (pnd) 18 prior to weaning.
All necropsies or humane termination of adult animals were performed after terminal anesthesia with CO2. Culled F1 and F2 offspring were terminated on pnd 4 by decapitation. Any F1 or F2 offspring that became moribund during lactation were terminated by decapitation (pnd 0–4) or by CO2 asphyxiation (pnd 4–21).
The animals were individually housed upon arrival, during the acclimation period, and upon the initiation of the treatment periods; by mating pairs (one male:one female) during the mating period; individually as plug-positive females from gestational day (gd) 0 until birth of their litters; as individual dams with their litters throughout the lactational period until weaning (pnd 21); and individually as selected or extra retained F1 postweanlings, in solid-bottom polypropylene cages (5" x 11.5" x 7") with stainless steel wire bar lids (Laboratory Products, Rochelle Park, NJ). Sani-Chip was used as cage bedding (P. J. Murphy Forest Products, Inc., Montville, NJ).
Temperature and relative humidity (RH) in the ARF animal rooms were continuously monitored, controlled, and recorded using an automatic system (Network 8000 System with Signal Software Version 4.4.1; Siebe Environmental Controls/Barber-Colman Company, Loves Park, IL). The target environmental ranges were 19–25°C for temperature and 30–70% RH, with a 12-h light cycle per day (NRC, 1996). There were no excursions in RH or temperature that affected the design, conduct, results, or conclusions of this study. At all times, the animals were handled, cared for, and used in compliance with the NRC Guide (NRC, 1996) and under an Research Triangle Institute (RTI) Institutional Animal Care and Use Committee protocol.
Purina Certified Ground Rodent Chow (No. 5002, PMI Feeds, Inc.) was available ad libitum in glass mouse feeding jars with stainless steel snap-on or screw-on caps and stainless steel wire-mesh inserts. Contaminant levels were provided by the supplier and were below certified levels. In addition, each of the three lots of feed used was analyzed by the supplier for the concentration of the phytoestrogens genistein (mean ± SEM: 192 ± 18.6 ppm), daidzein (177 ± 4.0 ppm), and glycitein (45 ± 8.9 ppm). The feed was stored at 60–70°F, and the period of use did not exceed 6 months from the milling date.
Water (tap water; source: City of Durham, Department of Water Resources, Durham, NC) was available ad libitum in glass water bottles with Teflon-lined, plastic screw caps and stainless steel sipper tubes. Contaminant levels of the Durham City water were measured at regular intervals (per EPA specifications) by the supplier and by Balazs Analytical Services, Inc. (Freemont, CA). Contaminant levels were below the maximal levels established for potable water and did not affect the design, conduct, or conclusions of this study.
Study Design
A graphic representation of the study design is presented in Figure 1. The study began with 28 males per group and 28 females per group (the F0 generation) to yield at least 20 pregnant females/group at or near term, and nine groups, including two negative control and one positive E2 control groups. Animals were assigned to groups by means of randomization stratified by body weight, such that the mean body weights by sex of all groups were homogeneous at treatment initiation. Table 1 provides the target dietary concentrations, target calculated, and actual BPA or E2 intakes for male and female parental F0 mice. BPA dietary concentrations were selected to provide doses that span the range from purported low-dose effects published in the literature (from 0.018 ppm; 0.003 mg/kg/day) to a dose that was anticipated to result in adult systemic toxicity (3500 ppm; 600 mg/kg/day).
|
|
Experimental Evaluations
Parental Animals (F0 and F1)
Observations for mortality were made twice daily (A.M. and P.M.). Clinical examinations were conducted and recorded at least once daily throughout the course of the study. The body weights of the F0 and F1 parental male mice were recorded initially and weekly through mating. The body weights of the F0 and F1 parental female mice were recorded in the same manner until confirmation of mating. During gestation, females were weighed on gd 0, 7, 14, and 17. Dams producing litters were weighed on pnd 0, 4, 7, 14, and 21. Body weight gains were computed for all intervals for all periods.
Feed consumption measurements were recorded for all F0 and F1 parental animals at least every 7 days, each time the feed was changed, throughout the prebreed exposure period. Feed consumption collection periods corresponded with the collection of the animals’ weekly body weight data and were employed to calculate intake of BPA or E2 on a mg of test article/kg body weight ratio. During pregnancy, feed consumption was recorded for gd 0–7, 7–14, and 14–17. During lactation, maternal feed consumption was measured for pnd 0–4, 4–7, 7–14, and 14–21, although maternal feed consumption after pnd 14 was confounded by the contribution from the pups, because pups were self-feeding by this time. Feed was changed at least weekly throughout the study. Feed consumption was not measured during the period of cohabitation because two adult animals (one male and one female) were in the same cage, but consumption was visually monitored and the feed was added or changed as necessary.
Estrous cyclicity and normality were evaluated by daily vaginal smears, using a sterile eyedropper and saline, from all F0 and F1 parental females during the last 3 weeks of their prebreed exposure periods. The smears were air dried, stained with Toluidine Blue, and the stage of each smear identified. The duration of the cycles was then determined by the mean number of days between the end of one stage and the start of the same stage during the 21-day period.
Animals of the F0 generation were 6 weeks of age at the commencement of treatment. They were administered dosed feed at their respective formulations for 8 weeks prior to mating (i.e., until they were 14 weeks of age). The animals were then mated on the basis of 1 male to 1 female, selected randomly within each dose group for a period of 14 days. Females were examined daily during the cohabitation period for the presence of copulation plug in the vaginal tract. The observation of a copulation plug in the vaginal tract was considered evidence of successful mating; the date a vaginal plug was observed was designated gd 0. Once the vaginal plug was observed, the male and female from that mating pair were individually housed.
Selected animals of the F1 generation were administered dosed feed at their respective formulations ad libitum 7 days/week for at least 8 weeks before mating. Developmental landmarks were evaluated as described below. The F1 animals were 11–13 weeks of age at the initiation of the mating period. The F1 animals were mated as described above for the F0 animals, including, but not limited to, pairing one male:one female and cohabitation for 14 days or until a vaginal copulation plug was observed, whichever came first (with no change in pairing). Brother-sister matings were not employed.
Beginning on gd 17, each plug-positive F0 or F1 female was observed twice daily (A.M. and P.M.) for evidence of delivery. The day of delivery was designated as pnd 0. The dams were allowed to rear their young to pnd 21. If a dam with a litter died or was sacrificed moribund, her litter was euthanized (by methods appropriate for their age) at her termination. Any female that did not show evidence of successful mating after 14 days of cohabitation continued on the original weekly weighing schedule and was monitored for evidence of pregnancy and delivery. Maternal and offspring information during lactation was collected for all females that delivered live litters, regardless of whether they had a confirmed gd 0.
Progeny (F1 and F2)
All live pups were counted, sexed, and examined as soon as possible on the day of birth (designated pnd 0) to determine the number of viable and stillborn members of each litter. Thereafter, litters were evaluated for survival on pnd 4, 7, and 14 and at weaning on pnd 21. All live F1 and F2 pups were individually counted, sexed, weighed, and examined grossly at pnd 0, 4, 7, 14, and 21. The body weights and sexes were recorded on an individual basis, but the pups were not uniquely identified at this stage, except for F1 females selected to be parents because they began evaluation for vaginal patency (VP) on pnd 18 prior to weaning (see below). All pups were examined for physical abnormalities at birth and throughout the prewean and postwean periods. Any pups that were stillborn, died, or were euthanized moribund during lactation were necropsied, when possible, to investigate the cause of death and for any malformations or variations, especially of the reproductive system.
At birth (pnd 0), all live F1 and F2 offspring had anogenital distance (AGD) measured by an eyepiece diopter (accurate to 0.01 mm) attached to a stereomicroscope and a microscope stage grid, and individual body weights were recorded. At pnd 21, all necropsied F1 and F2 offspring (up to three per sex per litter) also had AGD measured by the same method, and individual body weights were recorded.
On pnd 4, the size of each F1 and F2 litter was standardized by eliminating extra pups by random selection within sex from litters with more than 10 pups to yield 10 pups, with as nearly as possible five males and five females per litter. Natural litters with 10 or fewer pups were not standardized. All culled pups were sacrificed by decapitation and examined for visceral alterations, especially those of the reproductive system. Pups with gross lesions were retained in buffered neutral 10% formalin (BNF); normal pups were discarded.
Beginning on pnd 18, each F1 female pup selected for the parental generation (by randomization procedures stratified by body weight) to produce F2 litters (see below) was observed for VP, because exposure to the present positive control, E2, at 0.5 ppm resulted in accelerated VP acquisition beginning prior to weaning in previous one- and two-generation studies (Tyl et al., 2008a, b). Observations for VP continued until every selected female had this response. The date, age, and body weight were recorded for each female on the day of acquisition, and body weight was recorded on pnd 21. The F1 male pups were selected at weaning on pnd 21 (by randomization procedures stratified by body weight) to be parents of the F2 generation. Then, 1 additional F1 male pup/litter was randomly selected to be retained, with exposures continuing for 3 months, and necropsied when the F1 parental males were necropsied (at 14 weeks of age). A total of 28 F1 offspring per sex per group were selected as parents. During the prebreed exposure period, each F1 male was observed for cleavage of the balanopreputial gland (preputial separation [PPS]), beginning on pnd 22. Observations for PPS continued daily for each animal until each male had this response. The date, age, and body weight was recorded for each male on the day of acquisition, and body weight was recorded on pnd 30.
In addition, at weaning (pnd 21), up to three F1 and F2 pups/sex/litter were randomly selected and subjected to a gross necropsy, with organs weighed and retained in fixative. Organs were not weighed for any pups that died.
Gross Necropsy and Organ Weights
Euthanasia was by CO2 asphyxiation. Sacrifice of the parental F0 and F1 males occurred after the completion of the gestation period of their F1 and F2 litters, respectively. Sacrifice of the parental F0 and F1 females occurred on the same day as the weaning of their F1 and F2 litters, respectively. In addition, the retained F1 males (one per litter) were sacrificed at the same time as the F1 parental males (at 14 weeks of age). After euthanasia of the F0 and F1 females, prior to necropsy, a vaginal smear was taken from each female, stained (with Toluidine blue), and examined microscopically to determine the stage of estrus at demise. All F0 and F1 parental animals in all groups, as well as the retained F1 males (1/litter), were subjected to a complete gross necropsy. In addition, the uterus of each parental F0 and F1 female was examined, and the number of nidation scars (implantation sites) was documented in fixed uteri (after staining with potassium ferricyanide). Selected tissues were weighed and retained as outlined below. All gross lesions and the tissues listed in Table 2 were retained in BNF, except for one testis per male which was fixed in Bouin's fixative for 24 h and then retained in 70% ethanol, and one testis per male frozen at –20°C. Organ weights are reported as absolute and as relative to terminal body weight in the supplementary tables.
|
A complete gross necropsy and retention of tissues was conducted for any parental animal that appeared moribund (and was humanely sacrificed) or that died during the study. Organ weights were not recorded, and histopathology was not performed for animals found dead.
Histopathology.
Embedment of fixed tissues for examination was in paraffin. Tissue slides were stained with hematoxylin and eosin, except for male testes slides which were stained with Periodic Acid Schiff and hematoxylin. Full histopathology was performed on the organs listed in Table 2 for all males and females in the two vehicle control groups, for 10 per sex of the F0/F1 parental and F1 retained male animals in all BPA groups and in the E2 positive control group.
In addition, reproductive organs of the animals suspected of reduced fertility, not initially selected for histopathology (e.g., those that failed to mate, conceive, sire, or deliver healthy offspring, or for which estrous cyclicity or sperm number, motility, or morphology were affected) from the two vehicle control groups and all BPA and E2 groups were subjected to histopathological evaluation.
Testis.
Testicular histopathological examination was conducted to identify treatment-related effects such as retained spermatids, missing germ cell layers or types, multinucleated giant cells, or sloughing of spermatogenic cells into the lumen. Examination of the intact epididymis included the caput, corpus, and cauda (Russell et al., 1990).
Ovary.
Ovaries from all F0 and F1 adult females from the two vehicle control groups and all females from the high-dose BPA group and E2 positive control group were examined for enumeration of primordial follicles. Five ovarian sections were taken at least 100 µm apart from the inner third of each ovary. Examination included enumeration of the total number of primordial follicles from these 10 sections per female for comparison with numbers in control ovaries. Examination also confirmed the presence or absence of small, growing, and antral follicles and corpora lutea in comparison with control ovaries (Heindel et al., 1989).
Mammary glands.
The most anterior (axillary) pair of mammary glands in F0 and F1 adult females was retained in BNF (adhered to file cards to keep the preparation flat). The retained mammary glands from 10 randomly selected F1 adult females/group in all dose groups (BPA-exposed groups and E2 positive control group) and mammary glands from all F1 control females were subjected to histopathologic examination. The mammary glands from the F0 adult females were not evaluated.
Andrology.
At the time of F0 and F1 parental male sacrifice, one testis from each male was frozen at –20°C for subsequent enumeration of testicular homogenization-resistant spermatid heads and calculation of daily sperm production (DSP) and efficiency of DSP (LeBlond and Clermont, 1952; Robb et al., 1978). In addition, one cauda epididymis was immediately removed, weighed, and seminal fluid from the cauda was removed and placed into Medium 199 plus bovine serum albumin and assessed for sperm number, motility, and morphology. Sperm motility (motile and progressively motile) was assessed immediately after necropsy in all F0 and F1 adult males. Caudal sperm numbers (from homogenized cauda frozen at the time of necropsy) were recorded, and sperm morphology was evaluated from at least 500 sperm per male, if possible, from sperm slides made at necropsy that were stained with Eosin Y. No treatment-related effects were observed in the high-dose samples, so these parameters were not examined in other groups. Motility (and progressive motility) and number of cauda epididymal sperm and testicular homogenization-resistant spermatid head counts were assessed in F0 and F1 parental males using an HTM-IVOS Automated Sperm Analysis System (Version 12.1 c, Hamilton-Thorne Research, Beverly, MA). Sperm morphology was evaluated manually. Complete andrology was also performed on retained F1 males as specified above.
Nonpregnant females.
The fixed (BNF) uterus from any parental F0/F1 female failing to produce a litter was stained with potassium ferricyanide for confirmation of pregnancy status. This staining procedure did not interfere with subsequent histopathologic evaluation.
Gross Necropsy, Organ Weights, and Pathology (F1 and F2 Pups)
On pnd 21, all F1 pups remaining after selection and all F2 pups were euthanized by CO2 asphyxiation for gross necropsy examination on pnd 21, with the organs presented in Table 2 weighed and retained. The status of the F1 weanling testes was evaluated at necropsy after euthanasia by abdominal incision and localization of the testes low in the abdominal cavity at the inguinal ring (undescended) or in the scrotal sacs (descended). Gross lesions also were retained in fixative (BNF or Bouin's for testes). The organs listed in Table 2 were weighed and retained from two randomly selected pups/sex/litter. For the first randomly selected pup per sex per litter, histopathological examination was performed on the reproductive organs. For the second pup per sex per litter, histopathological examination was performed on all retained tissues (systemic and reproductive) and identified target organs (if any). All retained tissues were placed in fixative (10% neutral buffered formalin or Bouin's for testes), and all retained tissues examined histopathologically were embedded in paraffin.
Statistical Analyses
For the purposes of statistical comparisons, the data for the two control groups were combined. The unit of comparison was the F0 and F1 male, the F0 and F1 female, the pregnant F0 and F1 female, the retained F1 male, or the F1 and F2 litter. Quantitative continuous data (e.g., parental and pup body weights, organ weights, feed consumption, day of acquisition of VP and PPS, etc.) were compared among the six BPA treatment groups or the E2 positive control group versus the vehicle control group using either parametric ANOVA under the standard assumptions or robust regression methods (Huber, 1967; Royall, 1986; Zeger and Liang, 1986), which do not assume homogeneity of variance or normality. The homogeneity of variance assumption was examined via Levene's test (Levene, 1960). If Levene's test indicated lack of homogeneity of variance (p < 0.05), robust regression methods were used to test all treatment effects. These methods use variance estimators that make no assumptions regarding homogeneity of variance or normality of the data, and they were used to test for overall treatment group differences (via Wald chi-square test) and individual t-tests for exposed vs. vehicle control group comparisons when the overall treatment effect was significant (REGRESS procedure of SUDAAN Release 8.0; RTI, 2001).
If Levene's test did not reject the hypothesis of homogeneous variances, standard ANOVA techniques were applied for comparing the treatment groups. The General Linear Models procedure in SAS 8.0 was used to evaluate the overall effect of treatment and, when a significant treatment effect was present, to compare data from each exposed group to the control data via Dunnett's test (Dunnett, 1955, 1964). For the litter-derived percentage data (e.g., periodic pup survival indices), the ANOVA was weighted according to litter size. A one-tailed test (i.e. Dunnett's 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, AGD, VP, and PPS (General Linear Models procedure of SAS Release 8.0; SAS Institute Inc., 2000).
Frequency data, such as reproductive indices (e.g., mating and fertility indices), were not transformed. All indices were analyzed by chi-square test for Independence for differences among treatment groups (Snedecor and Cochran, 1967). When chi-square revealed significant (p < 0.05) differences among groups, then a Fisher's Exact Probability Test, with appropriate adjustments for multiple comparisons, was used for pairwise comparisons between each treatment group and the vehicle control group (SAS Institute, Inc., 2000).
Acquisition of developmental landmarks (e.g., VP and PPS), as well as AGD, was also analyzed by analysis of covariance (in addition to ANOVA analysis or robust regression analysis) using body weight at acquisition or measurement as the covariate. In addition, age at acquisition of puberty (VP, PPS) was also analyzed, with the individual body weights on pnd 21 for females and on pnd 30 for males as the covariate.
Correlated data (e.g., body and organ weights at necropsy of pups on pnd 21, with more than one pup per sex per litter) were analyzed using General Estimating Equations techniques (Zeger and Liang, 1986) in the SUDAAN software package (RTI, 2001). General Estimating Equations techniques were used to evaluate the overall effect of treatment and pairwise comparisons of exposed group values to the control group values. Student's t-test was used to analyze quantitative continuous data collected for endpoints from only two groups (e.g., ovarian primordial follicle counts and selected andrological endpoints) (SAS Institute, Inc, 2000).
A test for statistical outliers (SAS Institute, Inc, 1999) was performed on parental body weights, feed consumption (in g/day), and F0 and F1 adult, F1 and F2 weanling, and F1 retained male organ weights. When examination of pertinent study data did not provide a plausible, biologically sound reason for inclusion of the data flagged as "outlier," the data were excluded from summarization and analysis and were designated as outliers. The criterion for statistical significance was p 0.05. If the overall ANOVA or chi-square p value was significant, then the appropriate pairwise comparisons were made, and those pairwise comparisons that were statistically significant are presented in the summary tables. If the overall ANOVA or chi-square p value was not significant, then pairwise comparisons were not made. All values in the text, specified as "reduced" or "increased" relative to the control group value, were statistically significantly different unless otherwise indicated.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Parental Systemic Parameters
Fate and body weights.
There were no treatment-related deaths or clinical signs of toxicity for F0 or F1 males or females. There were no effects of BPA or E2 treatment on F0 male body weights throughout the 8-week prebreed period, the 2-week mating period, or the postmating holding period. Reduced body weights were observed for F1 parental (Fig. 2) and retained F1 males (data not shown), and reduced weight gain was observed in both F1 parental and retained males (4 and 10% of control value at termination, respectively) at 3500 ppm BPA. Mean adult male body weights and weight gains are provided in Supplementary Tables 1 (F0), 2 (F1), and 3 (F1 retained).
|
F0 and F1 female body weights were unaffected across all groups during the 8-week prebreed period. Mean prebreed body weight (Fig. 3) was reduced at 3500 ppm in F1 females on study day (sd) 0 and was considered to be related to reduced body weights in late lactation. There were no treatment-related effects in the E2 positive control group on mortality, clinical signs, or F0 male or F0/F1 female body weights. Slightly reduced body weights were observed for F1 males from the E2 group on sd 0 and 7 (Fig. 2), although this effect was considered to be transient and of questionable biological significance. Mean adult female body weights and weight gains are provided in Supplementary Tables 4–6 (F0 prebreed, gestation, lactation) and 7–9 (F1 prebreed, gestation, lactation).
|
Feed consumption and BPA/E2 intake.
There were no treatment-related effects on F0 and F1 male or female feed consumption (data not shown). BPA and E2 intake (mg/kg/day) exhibited the appropriate increases across groups and the anticipated decreases within groups over time during the prebreed/postwean exposure periods, because body weights were increasing and feed consumption was similar over time (Fig. 4A for F1 males and Fig. 5 for F1 females). Total (designated "maternal") feed consumption and therefore total BPA (and E2) intake were increased during the latter part of the lactational period for F0 and F1 dams, almost exclusively because the F1 and F2 pups began to self feed on or about pnd 14, and
30–40% of maternal feed consumption (and therefore BPA or E2 intake) during the last week of lactation is considered to be the result of pup consumption. There is no literature for the timing of the onset of mouse pup self-feeding, but it was observed in the present study (Fig. 5), and previous reproduction studies in mice (Tyl et al., 2008a, b) and rats (Tyl et al., 2002), and was measured in rats by Hanley and Watanabe (1985). Mean feed consumption, food efficiency (where appropriate), and BPA/E2 intake are provided in Supplementary Tables 10–12 (F0, F1, and retained F1 males), 13–15 (F0 females prebreed, gestation, lactation), and 16–18 (F1 females prebreed, gestation, lactation).
|
|
Organ weights and histopathology.
The absolute organ weights from F0 and F1 parental males and retained F1 males are presented in Table 3 (absolute and relative, as % of sacrifice weight, organ weights are included in Supplementary Tables 19–21). Treatment-related changes observed in F0/F1 adult males were limited to increased weights of the liver and kidneys at 3500 ppm, increased incidence of liver centrilobular hepatocyte hypertrophy at 300 ppm (minimal severity) and at 3500 ppm (minimal to mild severity), and renal nephropathy at 3500 ppm (minimal severity). There were no treatment-related effects of BPA on the histology or organ weights of the paired testes, paired epididymides, seminal vesicles and coagulating glands, or prostate (ventral lobe, dorsolateral lobe, or total prostate). The F0 male kidney weight increase observed at 300 ppm BPA was not regarded as toxicologically significant, because of the very small increase in absolute kidney weight (8% increased compared with the control group), and the absence of any associated histopathological findings. In addition, the statistically significant increases in F1 kidney weights in parental animals at 1.8, 30, and 300 ppm were not considered to be treatment related due to the lack of dose response (the mean kidney weights for these groups, at doses over 2 orders of magnitude, were 107, 111, and 108% of control, respectively), the lack of any correlating histopathological findings, and no effect in F1 retained males. The slight decrease in F1 absolute paired epididymal weights at 3500 ppm was not considered to be treatment related based on the lack of similar response in the F0 males or F1 retained males and the lack of an effect on relative (to body weight) organ weight. Mean adult male absolute and relative organ weights are provided in Supplementary Tables 19–21 (F0, F1, and F1 retained).
|
Selected absolute systemic and reproductive organ weights, ovarian primordial follicle counts, and stage of estrus at scheduled necropsy are presented in Table 4 for F0 and F1 adult parental females (absolute and relative, as % of sacrifice weight, organ weights are included in Supplementary Tables 22–23). Treatment-related changes observed in adult F0/F1 females were limited to increased absolute and/or relative weight of the liver and kidneys and histopathological changes in the liver (centrilobular hepatocyte hypertrophy of minimal severity) at 3500 ppm. There were no effects of BPA on the weights or histopathology of the paired ovaries, or uterus plus cervix plus vagina or on the histopathology of the F1 mammary glands at any BPA dose.
|
For the E2 positive control group, treatment-related findings were increased weights of the uterus + cervix + vagina in F0 and F1 female adults. For F0/F1 males, statistically significant changes included increases in absolute and/or relative weights of the pituitary (F0 and F1, not retained F1), thyroid (F1 only), liver, and kidneys; these findings were not considered to be treatment related, based on the lack of consistency across generations and/or at the same dose in the previous E2 two-generation study in mice (Tyl et al., 2008b
). Mean adult female ovarian primordial follicle counts and absolute and relative organ weights are provided in Supplementary Tables 22 and 23 (F0, F1).
Parental Reproductive Parameters
F0/F1 mating, fertility, pregnancy, and gestational indices and precoital interval were equivalent across all BPA groups and the control group (Table 6). For males, there were no effects on any F0/F1 andrological endpoint (Table 5) at any dose of BPA, except for a slight but significant decrease in epididymal sperm concentration at 3500 ppm BPA (F0 only). Percent motile, progressively motile, and percent abnormal sperm were all unaffected by BPA, as were testicular homogenization-resistant spermatid head counts, DSP, and efficiency of DSP. The decrease in epididymal sperm concentration observed for the F0 3500 ppm group was not considered to be treatment related, based on no effect in the F1 generation, no effects for any other andrological endpoint, and lack of any effects on male fertility. Mean adult male andrology measurements and calculated values are provided in Supplementary Tables 19–21 (F0, F1, and F1 retained).
|
|
For females, prebreed vaginal cytology indicated no treatment-related effects in any F0/F1 BPA-exposed group for number of animals cycling or percentage with at least one abnormal cycle. Cycle length was also biologically and statistically equivalent across all BPA groups. Vaginal cytology at necropsy (Table 4) indicated no difference across all BPA groups for the percentage of F0/F1 females in any stage of the estrous cycle. There were no differences between 0 and 3500 ppm BPA groups or the E2 group for the number of ovarian primordial follicle counts per female (Table 4).
Effects on reproductive parameters for E2 were reduced fertility index (F1 only), reduction in the percentage of females cycling (F1 only), increased number or percentage with at least one abnormal cycle (F0 only), and increased percentage of F0 in metestrus and decreased percentage in diestrus at demise (Tables 4 and 6). Mean adult female vaginal cytology determinations are provided in Supplementary Tables 22 and 23 (F0, F1).
Gestational and Lactational Parameters
There were no effects of BPA treatment for F1/F2 generations on precoital interval, gestational index, the number of implantation sites/litter, total number of live litters on pnd 0, live birth index, and the number of total, live, dead pups and sex ratio (% males) per litter on pnd 0. Percent postimplantation loss per litter was statistically equivalent across all groups, including the vehicle and positive control groups. The still birth index was statistically equivalent across all BPA groups. The higher still birth index for the F1 300 and 3500 ppm BPA groups (9.5 and 9.0%, respectively) was not considered to be treatment related based on the lack of dose response (no increase in the still birth index over a ten-fold increase in dose) and the lack of any effect (2.2 and 0% at 300 and 3500 ppm, respectively) in the F2 generation. Gestational length was statistically significantly increased for both the F0 and F1 generations at 3500 ppm BPA (19.3 vs. 19.0 days for the vehicle controls, for both generations); however, the toxicological significance of this minor difference, if any, is unknown (Table 6).
At 0.5 ppm E2, treatment-related effects were prolonged gestational length, apparent reduced number of live litters on pnd 0, reduced numbers of total and live F1/F2 pups per litter on pnd 0, and increased number of dead pups per litter (F1 only); the still birth index was higher and the live birth index was lower. Mean litter size, pup AGD, pup body weights and sex ratio by litter are provided in Supplementary Tables 24 and 25 (F1, F2).
There were no effects on any post-birth survival index throughout lactation in any BPA-treated group for F1 or F2 offspring. Pup body weights/litter (sexes separately or combined) were reduced at 3500 ppm BPA from pnd 7 to 21 (weaning) for F1 pups, but no effects were observed on body weights at any BPA or E2 dose for the F2 pups during lactation (Fig. 6 and Supplementary Tables 24 and 25). Male and female F1/F2 AGD on pnd 0 (both absolute and adjusted for individual pnd 0 body weight) was unaffected across all BPA groups (Fig. 6). For F1 (but not F2) males on pnd 21, absolute AGD was statistically significantly reduced at 3500 ppm BPA and at 300 and 3500 ppm BPA when adjusted for terminal body weight (Supplementary Table 26). The differences in F1 male AGD on pnd 21 were not considered to be treatment related because of the absence of effects on pnd 0 for F1/F2 males and on pnd 21 for F2 males. F1/F2 female AGD on pnd 21 (absolute or adjusted for terminal body weight) was equivalent across all BPA groups (Supplementary Table 26). Mean pup AGD, pup organ weights, and pup body weights on pnd 21 are provided in Supplementary Tables 26 and 27 (F1 and F2).
|
For E2, all survival indices during lactation (days 0–4 precull, 4 postcull–7, 7–14, and 14–21) and the lactational index (pnd 4 postcull–pnd 21) were equivalent. There were no effects on pup body weights in either generation. F1/F2 AGD on pnd 0 was equivalent to the control group. F1 (absolute and adjusted for body weight) and F2 (adjusted for body weight only) male AGD on pnd 21 was significantly reduced from controls. F1/F2 pup female AGD on pnd 21 (absolute or adjusted for terminal body weight) was not affected (Supplementary Table 26).
Weanling Necropsies
Selected F1 and F2 absolute systemic and reproductive organ weights and parameters are presented in Table 7 for weanling males and Table 8 for weanling females (absolute and relative [as % of sacrifice weight] organ weights are included in Supplementary Table 26 for F1 males and females and Supplementary Table 27 for F2 males and females). At 3500 ppm BPA, treatment-related findings included reduced spleen weight (males and females) and reduced testes weights with corresponding histopathology (increased incidence of minimal to mild hypoplasia of the seminiferous tubules) in the F1 and F2 weanlings. Other statistically significant differences in males or females occurred sporadically and were not considered to be treatment related due to lack of dose response, lack of correlating absolute or relative organ weight change, and/or lack of consistency across generations. Additionally, at 3500 ppm, apparent increases in the incidence of undescended testes (observed as testes low in the abdominal cavity near the inguinal ring and not in the scrotum; the normal location prior to testes descent) were observed at the F1/F2 weanling necropsies. This finding (a delay in testes descent) was not considered an adverse reproductive finding (see "Discussion" and Table 7). Centrilobular hepatocellular cytoplasmic alteration in the liver was a common finding in weanling animals from this study, with an increased incidence in BPA-exposed F1 weanling males at 300 and 3500 ppm. The term "cytoplasmic alteration, hepatocyte, centrilobular" was used to denote those livers in the weanlings that exhibited more pronounced centrilobular areas of clear hepatocyte cytoplasm than the normal variation, with slightly more cytoplasmic basophilia and the presence of minute vacuoles (data not shown), and were not considered to be toxicologically significant in the weanling liver or to be related to centrilobular hepatocyte hypertrophy observed in adults.
|
|
For the E2 positive control group, treatment-related effects in the F1/F2 male weanlings included decreased testes and epididymal weights associated with increased incidences of seminiferous tubule hypoplasia of the testis, and an apparent increased incidence in undescended testes observed at weanling necropsy. For F1/F2 females, effects included increased weights of the uterus plus cervix plus vagina with associated increased incidences (> 90%) of vaginal epithelial keratinization and bilateral luminal dilatation of the uterine horns.
F1 Postwean Parameters
Acquisition of puberty (PPS) was determined as absolute, adjusted for body weight at acquisition (so all males are at the same physiological state regardless of chronological age), and adjusted for body weight on pnd 30 (so all males are the same chronological age regardless of physiological state). Absolute age at acquisition (Fig. 7) and age adjusted for body weight at the time of acquisition (adjusted data not shown) were significantly delayed at 3500 ppm BPA. Body weight on pnd 30 was significantly reduced at 3500 ppm BPA (adjusted data not shown); however, age at acquisition, adjusted for pnd 30 body weight, was not significantly delayed.
|
The absolute age of acquisition of VP and the age adjusted for body weight at acquisition (so all females were in the same physiological state, regardless of chronological age) and for body weight on pnd 21 (so all F1 females were the same chronological age regardless of physiological state) were determined. Absolute age at VP (Fig. 6) was not significantly accelerated for any BPA group. Body weights at acquisition and on pnd 21 were significantly reduced at 3500 ppm BPA, and VP was accelerated when adjusted for the pnd 21 body weight (adjusted data not shown); however, VP was not affected if adjusted for body weight at acquisition.
At 0.5 ppm E2, PPS was delayed and VP was accelerated for absolute age at acquisition, for age adjusted for body weight at the time of acquisition, and for age adjusted for body weight at chronological day 30 for males and day 21 for females (Fig. 6 for absolute ages).
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Because there were limited multigeneration reproductive toxicity data in mice to provide historical background information, we performed an initial one-generation (Tyl et al., 2008a
) and a subsequent two-generation (Tyl et al., 2008b) reproductive toxicity study in mice, with a known estrogen, E2, in the diet. These studies established the appropriate range of doses, identified appropriate sensitive endpoints, and collected data on these endpoints in both control mice and mice exposed to a wide range of dietary E2 doses. The E2 dietary concentration identified for use as a positive control was 0.5 ppm (0.08 mg/kg/day), and the consistent effects from E2 exposure were: prolonged gestational length ( 10–14 h), decreased F1/F2 litter size (both total and live pups due to increased perinatal pup losses), reduced numbers of live F1/F2 litters/group, increased incidence of undescended testes in weanling males that was considered to be a delay in the second phase of testicular descent that occurs perinatally (rather than a failure of the in utero trans-abdominal descent), accelerated acquisition of puberty in females (VP), delayed acquisition of puberty in males (PPS), swollen/enlarged vaginal area in-life with increased weight of the uterus + cervix + vagina (UCV) in weanlings, and decreased weights of testes and epididymides in weanling males. The most sensitive effect observed from E2 was increased weight of the UCV in weanlings, observed from 0.05 to 0.5 ppm (0.008–0.08 mg/kg/day) (Tyl et al., 2008a, b).
Tyl et al. (2002) previously reported a three-generation, reproductive toxicity study of dietary BPA in CD (SD) rats in which the systemic toxicity NOAEL was 5 mg/kg/day, and the reproductive toxicity NOAEL was 50 mg/kg/day. This study is the most comprehensive study conducted on BPA to date, and provides robust and reliable data for hazard identification and risk assessment purposes. The rat was selected because it is the standard species routinely used in studies of reproductive and developmental toxicity. The present two-generation reproductive toxicity study was conducted under the same regulatory guidelines as the prior rat study, also with enhancements, in mice to allow direct comparison of the responses of the mouse and rat models to dietary BPA.
BPA did not result in toxicologically significant effects, including those identified for E2, on reproductive function at any dose. A slight increase in gestational length was observed at 3500 ppm BPA (19.3 vs. 19.0 days for the vehicle controls, for both generations); however, because there were no other indications of reproductive effects (e.g., no dystocia or perinatal pup toxicity), the toxicological significance of this minor difference (7 hours), if any, is unknown. Increased UCV weight was also not observed at any BPA dose in this study.
At the highest dose, 3500 ppm (600 mg/kg/day) of BPA, the delayed PPS and testicular findings in weanlings (delayed descent, decreased weight, and histopathology) may have been due to systemic toxicity, but these findings were also observed in the E2 positive control group. The apparent treatment-related increased incidence of undescended testes in F1/F2 male pups only at weaning at 3500 ppm BPA (i.e., delayed testes decent) was not considered an adverse reproductive finding because (1) all testes were located low in the abdominal cavity near the inguinal canal and not near the kidneys, which would be expected from disruption of the in utero phase of testis descent; (2) there were no F1 males in any group with undescended testes at adult necropsy (the F2 generation was terminated at weaning); (3) this observation was made in the absence of other markers of endocrine-mediated activity (e.g., no hypospadias, no retained nipples, no decreased AGD) that are typically found in males at doses lower than those that cause permanent cryptorchidism (e.g., following antiandrogen exposure); (4) adult F1 male reproductive functions (andrology and male reproductive indices) and testicular histopathology were equivalent across all groups; and (5) there were some F1 and F2 males in all other groups (including a few in the control group) that were observed with undescended testes at weaning. The delay in testes descent was consistent with the increased incidence of mild/minimal seminiferous tubule hypoplasia and reduced testes weight also observed only in weanlings at 3500 ppm. Additional considerations are that although male mice acquire PPS at a younger age than male rats, data available indicate that testis descent is not acquired by all mouse males (or rat males) by weaning, and that the testes of young male mammals, including mice, during the period of postnatal testes descent may be easily displaced through the initially wide inguinal ring. Increased incidence of undescended testes in weanlings was not observed in the three-generation study in rats (Tyl et al., 2002) at comparable doses, indicating that the observation in this study is not common between these two species.
At 300 ppm (50 mg/kg/day), there were no indications of toxicity in the offspring, even in the presence of mild systemic toxicity in parental animals (liver effects). There were also no indications of systemic, reproductive, or offspring toxicity from 0.018 to 30 ppm BPA (0.003 to 5 mg/kg/day).
No effects on prostate weight were observed at any BPA dose in mice in this study, as previously reported in small oral studies. Nagel et al. (1997) reported an increase in prostate weight among offspring after maternal CF-1 mouse exposure to oral 0.002 or 0.02 mg/kg/day BPA from gd 11 to 17. Because these doses were below the reference dose, this finding received a considerable amount of attention. However, neither Cagen et al. (1999) nor Ashby et al (1999) observed any changes in prostate weight at these doses using a study protocol similar to Nagel et al. (1997). Owens and Chaney (2005) recently conducted a thorough analysis of the statistical power of these three studies. They demonstrated that the negative studies (Ashby et al. 1999; Cagen et al. 1999), because of their larger sample size and more complete data reporting, had ample statistical power to detect an effect, even one less pronounced than that reported by Nagel et al. (1997). Gupta (2000) reported an increase in offspring prostate weight at a dose of 0.05 mg/kg/day in mice after maternal oral dosing on gd 16–18. Timms et al. (2005) dosed maternal CD-1 mice by micropipette on gd 14–18 with 10 µg/kg/day BPA and reported increased overall prostate volume and effects on prostate morphogenesis in offspring. Other studies, by the parenteral route of administration, cannot be compared with oral studies and are of limited relevance to human risk assessment, due to the lack of presystemic clearance of BPA by first-pass metabolism in parenteral studies, as well as route-dependent metabolism (Pottenger et al., 2000). The present two-generation study in mice found no evidence that low (or high) oral doses of BPA affected prostate weight. Reviews by (Gray et al. 2004), Goodman et al. (2006), Rhomberg et al. (2007), EFSA (2006), CERHR (2007), and Willhite et al. (2008) also concluded that there is no consistent effect on the prostate. Our negative results at low doses, and the absence of toxicity to reproductive development and adult function at any dose in this study of oral BPA in mice, were expected based on the known pharmacokinetics and metabolism of BPA in rodents and humans (Domoradzki et al., 2003; Pottenger et al., 2000; Völkel et al., 2002). A number of reviews (e.g., CERHR, 2007; EFSA, 2006; Goodman et al., 2006; Willhite et al., 2008) have used these pharmacokinetic data to consider studies using nonoral routes of administration as of limited utility and not relevant to the assessment of risks of oral BPA exposure at low doses to humans of all ages.
An additional important aim of this BPA study in mice was to allow direct comparison with the results of BPA in rats (Tyl et al., 2002) at a comparable dose range. The Tyl et al. (2002) study evaluated BPA effects in CD rats at dietary concentrations of 0, 0.015, 0.3, 4.5, 75, 750, or 7500 ppm (intake of 0, 0.001, 0.002, 0.3, 5, 50, or 500 mg/kg/day). In compliance with regulatory guidelines, the top doses in both the rat and the mouse studies were selected to result in adult systemic toxicity. Table 9 summarizes the key findings in the two species. For both species, systemic toxicity was evident as body weight decreases and organ weight and/or histopathological alterations. In adult and weanling rats (particularly males), body weight effects were greater than in mice, which consequently led to more pronounced reduction in absolute organ weights and increased relative organ weights in rats. In contrast, body weight and organ weight effects in adult and weanling mice were less pronounced.
|
Effects on reproductive parameters observed in the multigeneration study in rats (reduced number of implants per litter and reduced number of pups/litter) were not observed in mice; the slight increase in gestational length observed in this study in mice was not observed in the rat study.
Developmental effects common to both species were the decreased pup body weights and effects on organ weights, noted above, and delayed PPS. Delay in VP was observed only in rats. However, acquisition of developmental landmarks (including PPS and VP) is dependent on both age and body weight, (i.e., lighter animals acquire the landmark later), and the observed delays were likely secondary effects from reduced body weights. All postwean animals in both studies acquired puberty.
No treatment-related effects were observed for either rat or mouse study on mating, fertility and pregnancy indices, andrological parameters (including sperm count, motility and morphology), prebreed estrous cycle length, precoital interval, gestational index, live birth and still birth indices, sex ratio, % postimplantation loss, postbirth survival indices, or AGD (pnd 0).
Based on Tyl et al. (2002) and this study, both rats and mice showed similar responses to dietary BPA at comparable high doses. The no observable effect level (NOEL) for systemic toxicity for both species was 5 mg/kg/day, and the reproductive/developmental NOEL was 50 mg/kg/day in both species. Neither species exhibited low-dose effects or nonmonotonic dose-response curves for any of the parameters evaluated.
These studies in mice and rats, based on regulatory testing guidelines, are important for risk assessment because they are necessarily robust, examine a broad range of appropriate, sensitive parameters to detect adverse outcomes, if any, over a broad range of doses, use administration by a relevant route of exposure during sensitive life stages over multiple generations in appropriate animal models, with validated endpoints and large numbers of animals/group.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Exposure to dietary BPA in CD-1 (Swiss) mice for two offspring generations resulted in minimal systemic toxicity at 300 ppm (centrilobular hepatocyte hypertrophy). At 3500 ppm, systemic toxicity was expressed as reduced body weights and increased liver and kidney weights, with associated histopathology. Slightly increased gestational length was observed at 3500 ppm, with the toxicological significance, if any, unknown. There were no other effects on reproductive parameters at any dose.
At 3500 ppm, toxicity to the offspring was expressed as decreased spleen weight in male and female weanlings and reduced testes weights associated with testicular seminiferous tubule hypoplasia. Because there were no effects on adult spleen weights or on adult testis structure or function, the likely explanation is that these are transient effects from systemic toxicity. Other observations were an apparent increased incidence of undescended testes in weanlings only (considered a delay in the process of normal testes descent) and a delay in the acquisition of PPS in postweanling offspring, both without permanent consequence on adult reproduction.
All treatment-related effects exhibited monotonic dose-response curves, with no evidence of low-dose effects. There were also no statistically or biologically significant treatment-related changes in prostate weight (whole organ or ventral and dorsolateral lobes separately) in either the F0 or F1 parental or F1 retained adult males (not evaluated in weanlings). In addition, no evidence of increased body weights or weight changes at any BPA dose, and no treatment-related reproductive tract gross or histopathologic lesions were observed. Although the purpose of this study was not to identify malformations, detailed gross necropsies were performed on stillborn/dead pups, pnd 4 culled pups, weanlings, and adults. No treatment- or dose-related malformations (external or visceral) were observed. The systemic toxicity NOEL was 30 ppm BPA (5 mg/kg/day), and the reproductive and offspring toxicity NOEL was 300 ppm BPA (50 mg/kg/day). The findings for the 0.5 ppm E2 positive control group were consistent with those from the same dose in the previously conducted two-generation E2 study in mice (Tyl et al., 2008b).
In addition, the CD (SD) rat and CD-1 (Swiss) mouse models proved to be of similar sensitivity to BPA and resulted in equivalent systemic and reproductive/developmental NOELs for BPA in both species. The results of this study, which are consistent with the previous three-generation study in rats, support the conclusion that BPA is not a selective reproductive or developmental toxicant.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Supplementary data are available online at http://toxsci.oxfordjournals.org/.
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Polycarbonate/BPA Global Group, Arlington, VA.
|
ACKNOWLEDGMENTS |
|---|
We wish to thank Mr M. D. Crews, Mr J. E. Gray, Ms C. R. Robinson, Ms N. M. Kuney, Mr C. G. Leach, Ms L. L. Macdonald, Ms A. J. Parham, Ms D. N. Wenzel, Ms L. B. Pelletier, Mr W. P. Ross, Ms K. D. Vick, Mr T. W. Wiley, Ms V. I. Wilson, Ms N. A. Ostin, and Dr J. E. Scott-Emuakpor, DVM, of the RTI Laboratory Animal Sciences Group; Mr D. L. Hubbard, Mr R. A. Price, and Mr J. E. Larson of RTI's Materials Handling Facility; Ms N. P. Castillo, Ms T. Uenoyama, and Dr J. S. Green of RTI's Analytical Chemistry Group; Ms D. A. Drissel, Manager, Ms C. D. Keller, Ms C. A. Ingalls, Ms M. M. Oh, and Ms S. C. Wade of RTI's Quality Assurance Unit; and Ms C. A. Winkie RTI LST administrative staff.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONOBJECTIVESMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSSUPPLEMENTARY DATAFUNDINGREFERENCES |
|---|
Ashby J, Tinwell H, Haseman J. Lack of effects for low dose levels of bisphenol A and diethylstilbestrol on the prostate gland of CF1 mice exposed in utero. Reg. Toxicol. Pharmacol. (1999) 30:156–166.[CrossRef][Web of Science][Medline]
Cagen SZ, Waechter JM, Dimond SS, Breslin WJ, Butala JH, Jekat FW, Joiner RL, Shiotsuka RN, Veenstra GE, Harris LR. Normal reproductive organ development in CF-1 mice following prenatal exposure to bisphenol A. Toxicol. Sci. (1999) 50:36–44.
CERHR. NTP-CERHR expert panel report on the reproductive and developmental toxicity of bisphenol A. National Toxicology Program (NTP). (2007) U. S. Dept. of Health and Human Services (http://cerhr.niehs.nih.gov), dated November 26, 2007.
Domoradzki JY, Pottenger LH, Thornton CM, Hansen SC, Card TL, Markham DA, Drysga MD, Shiotsuka RN, Waechter JM Jr. Metabolism and pharmacokinetics of bisphenol A (BPA) and the embryo-fetal distribution of BPA and BPA-monoglucuronide in CD Sprague Dawley rats at three gestational stages. Toxicol. Sci. (2003) 76:21–34.
Dunnett CW. A multiple comparison procedure for comparing several treatments with a control. J. Am. Stat. Assoc. (1955) 50:1096–1121.[CrossRef][Web of Science]
Dunnett CW. New tables for multiple comparisons with a control. Biometrics (1964) 20:482–491.[Medline]
EFSA. Opinion of the Scientific Panel on food additives, flavourings, processing aids and materials in contact with food (AFC) related to 2,2-bis(4-hydroxyphenyl)propane. EFSA J. (2006) 428:1–75.
Ema M, Fujii S, Furukawa M, Kiguchi M, Ikka T, Harazono A. Rat two-generation reproductive toxicity study of bisphenol A. Reprod. Toxicol. (2001) 15:505–523.[CrossRef][Web of Science][Medline]
Goodman JE, McConnell EE, Sipes IG, Witorsch R, Slayton TM, Yu CJ, Lewis AS, Rhomberg LR. An updated weight of the evidence evaluation of reproductive and developmental effects of low doses of bisphenol A. Crit. Rev. Toxicol. (2006) 36:387–457.[CrossRef][Web of Science][Medline]
Gupta C. Reproductive malformation of the male offspring following maternal exposure to estrogenic chemicals. Proc. Soc. Exp. Biol. Med. (2000) 224:61–68.
Gray GM, Cohen JT, Cunha G, Hughes C, McConnell EE, Rhomberg L, Sipes IG, Mattison D. Weight of the evidence evaluation of low-dose reproductive and developmental effects of bisphenol A. Hum. Ecol. Risk Assess. (2004) 10:875–921.[CrossRef]
Hanley TR, Watanabe PG. Measurement of solid feed consumption patterns in neonatal rats by 141Ce-radiolabeled microspheres. Toxicol. Appl. Pharmacol. (1985) 77:496–500.[CrossRef][Web of Science][Medline]
Heindel JJ, Thomford PJ, Mattison DR. Histological assessment of ovarian follicle number in mice as a screen for ovarian toxicity. In: Growth Factors and the Ovary—Hirshfield AN, ed. (1989) New York: Plenum Publishing Corp. 421–426.
Huber PJ. The behavior of maximum likelihood estimates under nonstandard conditions. Proc. Fifth Berkeley Symp. Math. Stat. Probab. (1967) 1:221–233.
LeBlond CP, Clermont Y. Definition of the stages of the seminiferous epithelium in the rat. N. Y. Acad. Sci. (1952) 55:548–573.[Web of Science][Medline]
Levene H. Robust tests for the equality of variance. In: Contributions to Probability and Statistics—Olkin I, ed. (1960) Palo Alto, CA: Stanford University Press. 278–292.
Long X, Steinmetz R, Ben-Jonathan N, Caperell-Grant A, Young PC, Nephew KP, Bigsby RM. Strain differences in vaginal responses to the xenoestrogen bisphenol A. Environ. Health Perspect. (2000) 108:243–247.[Web of Science][Medline]
Nagel SC, vom Saal FS, Thayer KA, Dhar MG, Boechler M, Welshons WV. Relative binding affinity-serum modified access (RBA-SMA) assay predicts the relative in vivo bioactivity of the xenoestrogens bisphenol A and octylphenol. Environ. Health Perspect. (1997) 105:70–76.[Web of Science][Medline]
NRC. Guide for the Care and Use of Laboratory Animals (1996) Washington, DC: Institute of Laboratory Animal Resources, Commission on Life Sciences, National Research Council. National Academy Press, National Institutes of Health.
Organisation for Economic Cooperation and Development (OECD). OECD Series on Principles of Good Laboratory Practice and Compliance Monitoring, No. 1, OECD Principles of Good Laboratory Practice (1998) Paris: Environment Directorate, OECD.
Organisation for Economic Cooperation and Development (OECD). Guideline for Testing of Chemicals, No. 416 "Two-generation Reproduction Toxicity Study (2001) OECD, Paris.
Organisation for Economic Cooperation and Development (OECD). OECD Series on Principles of Good Laboratory Practice and Compliance Monitoring. No. 13. The Application of the OECD Principles of GLP to the Organisation and Management of Multi-Site Studies (2002) OECD, Paris.
Owens JW, Chaney JG. Weighing the results of differing ‘low dose’ studies of the mouse prostate by Nagel, Cagen, and Ashby: Quantification of experimental power and statistical results. Regul. Toxicol. Pharmacol. (2005) 43:194–202.[CrossRef][Web of Science][Medline]
Pottenger LH, Domoradzki JY, Markham DA, Hansen SC, Cagen SZ, Waechter JM Jr. The relative bioavailability and metabolism of bisphenol A in rats is dependent upon the route of administration. Toxicol. Sci. (2000) 54:3–18.
Reel JR, Tyl RW, Lawton AD, Lamb JC. (1985) Bisphenol A: Reproduction and fertility assessment in CD-1 mice when administered in the feed. Report to the National Toxicology Program from Research Triangle Institute, Research Triangle Park, NC (NTIS PB86-103207).
Rhomberg LR, Goodman JE, McConnell EE, Sipes IG, Witorsch RJ, Slayton TM, Yu CJ, Lewis AS. An updated weight of evidence evaluation of reproductive and developmental effects of low dose of bisphenol A. Toxicologist (2007) 96:427.
Robb GW, Amann RP, Killian GJ. Daily sperm production and epididymal sperm reserves of pubertal and adult mice. J. Reprod. Fertil. (1978) 54:103–107.
Royall RM. Model robust confidence intervals using maximum likelihood estimators. Int. Stat. Rev. (1986) 54:221–226.
RTI. SUDAAN® User's Manual, Release 8.0 (2001) Research Triangle Park, NC: Research Triangle Institute.
Russell LD, Ettlin RA, Hikim APS, Clegg ED. Histological and Histopathological Evaluation of the Testis (1990) Clearwater, FL: Cache River Press.
SAS Institute, Inc. SAS® Language Reference Dictionary, Version 8 (1999) Cary, NC: SAS Institute, Inc.
SAS Institute, Inc. SAS/STAT® Software: Changes and Enhancements, Release 8.1, Version 8 (2000) Cary, NC: SAS Institute, Inc.
Snedecor GW, Cochran WG. Statistical Methods (1967) 6th edn. Ames, IA: Iowa State University Press.
Spearow JL, Barkley M. Reassessment of models used to test xenobiotics for oestrogenic potency is overdue. Hum. Reprod. (2001) 16:1027–1029.
Timms BG, Howdeshell KL, Barton L, Bradley S, Richter CA, vom Saal FS. Estrogenic chemicals in plastic and oral contraceptives disrupt development of the fetal mouse prostate and urethra. Proc. Nat. Acad. Sci. U. S. A. (2005) 102:7014–7019.
Tyl RW, Myers CB, Marr MC, Castillo NP, Veselica MM, Joiner RL, Dimond SS, Van Miller JP, et al. One-generation reproductive toxicity study of dietary 17β-estradiol (E2; CAS No. 50-28-2) in CD-1® (Swiss) mice. Reprod. Toxicol. (2008a) 25:144–160.[CrossRef][Web of Science][Medline]
Tyl R, Myers C, Marr M, Sloan CS, Castillo N, Veselica MM, Seely JC, Dimond SS, Van Miller JP, et al. Two-generation reproductive toxicity evaluation of dietary 17ß-estradiol (E2; CAS No. 50-28-2) in CD-1® (Swiss) mice. Toxicol. Sci. (2008b) 102:392–412. doi:1093/toxsci/kfn002.
Tyl RW, Myers CB, Marr MC. (2005) Thirteen-week range-finding study for the two-generation reproductive toxicity evaluation of Bisphenol A (BPA; CAS No. 80-05-7) administered in the feed to CD-1 (Swiss) mice. Final report to the American Plastics Council, dated August 10, 2005; unpublished.
Tyl RW, Myers CB, Marr MC, Thomas BF, Keimowitz AR, Brine DR, Veselica MM, Fail PA, Chang TY, Seely JC, et al. Three-generation reproductive toxicity study of dietary bisphenol A in CD Sprague-Dawley rats. Toxicol. Sci. (2002) 68:121–146.
Völkel W, Colnot T, Csanady GA, Filser JG, Dekant W. Metabolism and kinetics of bisphenol A in humans at low doses following oral administration. Chem. Res. Toxicol. (2002) 15:1281–1287.[CrossRef][Web of Science][Medline]
vom Saal FS, Hughes C. An extensive new literature concerning low-dose effects of bisphenol A shows the need for a new risk assessment. Environ. Health Perspect. (2005) 113:926–933.[Web of Science][Medline]
vom Saal FS. Chapel Hill bisphenol A expert panel consensus statement: Integration of mechanisms, effects in animals and potential to impact health at current levels of exposure. Reprod. Toxicol. (2007) 24:131–138.[CrossRef][Web of Science][Medline]
Willhite CC, Ball GL, McLellan CJ. Derivation of a bisphenol A oral reference dose (RfD) and drinking-water equivalent concentration. J. Toxicol. Environ. Health B Crit. Rev (2008) 11:69–146.[Web of Science][Medline]
Zeger S, Liang K. Longitudinal data analysis for discrete and continuous outcomes. Biometrics (1986) 42:121–130.[CrossRef][Web of Science][Medline]
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  D. Li, Z. Zhou, D. Qing, Y. He, T. Wu, M. Miao, J. Wang, X. Weng, J.R. Ferber, L.J. Herrinton, et al. Occupational exposure to bisphenol-A (BPA) and the risk of Self-Reported Male Sexual Dysfunction Hum. Reprod., November 10, 2009; (2009) dep381v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. A. Hunt, M. Susiarjo, C. Rubio, and T. J. Hassold The Bisphenol A Experience: A Primer for the Analysis of Environmental Effects on Mammalian Reproduction Biol Reprod, November 1, 2009; 81(5): 807 - 813. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 Y.-J. Li, T.-B. Song, Y.-Y. Cai, J.-S. Zhou, X. Song, X. Zhao, and X.-L. Wu Bisphenol A Exposure Induces Apoptosis and Upregulation of Fas/FasL and Caspase-3 Expression in the Testes of Mice Toxicol. Sci., April 1, 2009; 108(2): 427 - 436. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. N. Vandenberg, M. V. Maffini, C. Sonnenschein, B. S. Rubin, and A. M. Soto Bisphenol-A and the Great Divide: A Review of Controversies in the Field of Endocrine Disruption Endocr. Rev., February 1, 2009; 30(1): 75 - 95. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||