Developmental exposure to a commercial PBDE mixture, DE-71: neurobehavioral, hormonal, and reproductive effects

TOXICOLOGICAL SCIENCES 116(1), 297–312 (2010) doi:10.1093/toxsci/kfq105 Advance Access publication April 7, 2010

Developmental Exposure to a Commercial PBDE Mixture, DE-71: Neurobehavioral, Hormonal, and Reproductive Effects

*Neurotoxicology Branch, †Developmental Toxicology Branch and ‡Endocrine Toxicology Branch, Toxicity Assessment Division, National Health and Environmental Effects Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711; §Wadsworth Center, New York State Department of Health and Department of Environmental Health Sciences, State University of New York, Albany, New York 12201; and {National Cancer Institute and National Institute of Environmental Health Sciences, Research Triangle Park, North Carolina 27709 1 To whom correspondence should be addressed at Neurotoxicology Branch, Toxicity Assessment Division, B105-06, NHEERL/ORD, U.S. Environmental Protection Agency, Research Triangle Park, NC 27711. Fax: (919) 541-0484. E-mail: [email protected]. 2 Present address: National Institute of Environmental Health Sciences, National Toxicology Program, MD E1-08, 111 T.W. Alexander Drive, Durham, NC 27709. 3 Present address: Oak Ridge National Laboratories, 1060 Commerce Park Drive, Oak Ridge, TN 37830.

Received December 28, 2009; accepted March 31, 2010

Developmental effects of polybrominated diphenyl ethers (PBDEs) have been suspected due to their structural similarities to polychlorinated biphenyls (PCBs). This study evaluated neurobehavioral, hormonal, and reproductive effects in rat offspring perinatally exposed to a widely used pentabrominated commercial mixture, DE-71. Pregnant Long-Evans rats were exposed to 0, 1.7, 10.2, or 30.6 mg/kg/day DE-71 in corn oil by oral gavage from gestational day 6 to weaning. DE-71 did not alter maternal or male offspring body weights. However, female offspring were smaller compared with controls from postnatal days (PNDs) 35–60. Although several neurobehavioral endpoints were assessed, the only statistically significant behavioral finding was a dose-by-age interaction in the number of rears in an openfield test. Developmental exposure to DE-71 caused severe hypothyroxinemia in the dams and early postnatal offspring. DE-71 also affected anogenital distance and preputial separation in male pups. Body weight gain over time, reproductive tissue weights, and serum testosterone concentrations at PND 60 were not altered. Mammary gland development of female offspring was significantly affected at PND 21. Congener-specific analysis of PBDEs indicated accumulation in all tissues examined. Highest PBDE concentrations were found in fat including milk, whereas blood had the lowest concentrations on a wet weight basis. PBDE concentrations were comparable among various brain regions. Thus, perinatal exposure to DE-71 leads to accumulation of PBDE congeners in various tissues crossing blood-placenta and Disclaimer: The research described in this article has been reviewed by the National Institute of Health and the National Health and Environmental Effects Research Laboratory of the U.S. Environmental Protection Agency and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Agency nor does mention of trade names or commercial products constitute endorsement or recommendation for use.

blood-brain barriers, causing subtle changes in some parameters of neurobehavior and dramatic changes in circulating thyroid hormone levels, as well as changes in both male and female reproductive endpoints. Some of these effects are similar to those seen with PCBs, and the persistence of these changes requires further investigation. Key Words: polybrominated diphenyl ethers; neurotoxic effects; reproductive effects; thyroid hormones; mammary gland; PBDE levels; DE-71; motor activity; neurobehavior.

Polybrominated diphenyl ethers (PBDEs) are high– production volume chemicals and are mainly used as flame retardants in construction materials, coatings, and textiles and in polymers found in electronic equipment such as computers and televisions (Birnbaum and Staskal, 2004; Sjodin et al., 2003; World Health Organization, 1994). PBDEs have been marketed as penta- (e.g., DE-71), octa- (e.g., DE-79), and decabrominated (e.g., DE-83) mixtures. Because PBDEs are not chemically bound to the polymer products, they can be released from products into the environment (Hutzinger et al., 1976). Due to their lipophilicity and environmental stability, PBDEs accumulate in sediment and biota (Sellstrom, 1996). High concentrations have been found in sentinel animal species such as marine mammals (de Boer et al., 1998; Shaw and Kannan, 2009), cats (Dye et al., 2007), and freshwater fish (Johnson and Olson, 2001), as well as in human tissue and breast milk (Frederiksen et al., 2009; Johnson-Restrepo et al., 2005; Rahman et al., 2001; Sjodin et al., 2001). PBDEs have also been found in several U.S. commercial food products such


The Author 2010. Published by Oxford University Press on behalf of the Society of Toxicology. All rights reserved. For permissions, please email: [email protected]

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

Prasada Rao S. Kodavanti,*,1 Cary G. Coburn,* Virginia C. Moser,* Robert C. MacPhail,* Suzanne E. Fenton,†,2 Tammy E. Stoker,‡ Jennifer L. Rayner,†,3 Kurunthachalam Kannan,§ and Linda S. Birnbaum{

298

KODAVANTI ET AL.

respectively (Stoker et al., 2004, 2005). In addition to the delay in pubertal onset, the weights of certain androgen-dependent tissues including the ventral prostate and seminal vesicles were decreased. Studies designed to evaluate changes in steroid hormones/androgen receptor (AR) function following DE-71 exposure showed that this mixture as well as the predominant congeners present in this mixture were AR antagonists both in vivo and in vitro suggesting anti-androgenic activity (Stoker et al., 2005). A delay in the onset of vaginal opening (VO; a marker of pubertal onset) in female rats was also observed following peripubertal exposure to 60 mg/kg of DE-71 and higher doses; however, it is not known whether an antiestrogenic mechanism is responsible for this effect. In addition, perinatal exposure to low concentrations of PBDE 47 (0.4–0.7 mg/kg) decreased ovarian weight and produced alterations in folliculogenesis in female offspring (Talsness et al., 2008). With increasing concentrations being detected in the environment and in tissue/milk samples of humans and other animals (Frederiksen et al., 2009; Rayne et al., 2003; Schecter et al., 2005), the anti-androgenic and neurotoxic effects of the PBDEs raise new concerns about the possible effects on human reproductive and nervous system development. Because antiandrogenic environmental chemicals have the potential to adversely impact human reproductive health and are observed in human and wildlife populations (Guillette et al., 1994; Shaw and Kannan, 2009), such an effect would be important, especially because these chemicals appear to be fairly persistent in the environment and are ubiquitous, being found in such places as dust in window sills and dryer lint (Stapleton et al., 2005). Due to limited reports regarding the developmental effects of these PBDEs, we have evaluated the effects of the pentabrominated diphenyl ether mixture (DE-71) on general growth, neurobehavior, and hormones and on reproductive tissue development in rats following perinatal exposure. We selected DE-71 commercial mixture as this contains tetra- and pentabrominated congeners that are prevalent in the environmental and biological samples. The results indicate subtle effects on some parameters of neurobehavior and substantial effects on the development of the female reproductive system and on TH status. MATERIALS AND METHODS Animals. Timed-pregnant Long-Evans rats were obtained from Charles River Laboratory (Portage, MI) on GD 3 (day of insemination as indicated by copulatory plug was GD 0). The animals were housed in American Association for Accreditation of Laboratory Animal Care–approved animal facilities. The animals were housed individually in standard polycarbonate plastic cages with heat-treated pine shavings as bedding. Food (Purina Lab Diet 5008) and water were provided ad libitum. Temperature was maintained at 21C ± 2C and relative humidity at 50 ± 10% under a 12/12-h light/dark cycle (lights on from 0600 to 1800 h). All experiments were approved by the animal care and use committee of the National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency. Experimental treatment. The commercial PBDE mixture, DE-71 (Lot #1550OI18A), was a gift from Great Lakes Chemical Corporation (El Dorado,

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

as canned fish, meat, and dairy products (Frederiksen et al., 2009; Schecter et al., 2006). PBDEs can exist, based on the number and position of the bromine atoms, as 209 possible congeners that are noncoplanar with low vapor pressure and high lipophilicity (log Kow ranges from 4.28 to 9.9) (International Programme on Chemical Safety, 1994). These characteristics are similar to those of noncoplanar polychlorinated biphenyls (PCBs). Despite these structural similarities, few reports have demonstrated toxicological similarities between PBDEs and PCBs in vitro (Kodavanti and Derr-Yellin, 2002; Kodavanti et al., 2005) or in vivo (Eriksson et al., 2001; Viberg et al., 2003a). Some studies indicated that developmental exposure to PBDEs can lead to a wide array of effects including permanent aberrations in motor activity and learning and memory (Driscoll et al., 2009; Eriksson et al., 2001; Viberg et al., 2003a,b, 2004). Mice exposed to a single dose of PBDE 47 on postnatal day (PND) 10 demonstrated delayed ontogeny of neuromotor function and hyperactivity when they attained adult age without any alterations in circulating thyroid hormone (TH) levels (Gee and Moser, 2008; Gee et al., 2008). These studies are in agreement with previous reports on PBDE 99 where hyperactivity was observed in CD-1 Swiss mice following developmental exposure in the absence of significant changes in circulating TH levels when assessed at PND 22 (Branchi et al., 2005). Other studies showed developmental delays in the acquisition of the palpebral reflex following neonatal exposure to PBDE 209 along with changes in circulating thyroxine (T4) levels (Rice et al., 2007). Because of these differential reports, the role of hypothyroxinemia in the behavioral effects of PBDEs is unclear. Kuriyama et al. (2005) indicated that gestational exposure of pregnant rats to low doses of PBDE 99 caused hyperactivity in the offspring, an effect that extended into adulthood. Similarly, Branchi et al. (2002, 2003, 2005) reported that prolonged developmental exposure to PBDE 99 affected sensory motor development as indicated by a delay in screen climbing response and spontaneous behavior as indicated by hyperactivity and impaired habituation in mice. During developmental exposure in rats, increased locomotor activity was produced by PBDE 47 (Suvorov et al., 2009), whereas Cheng et al. (2009) reported a delayed appearance of cliff drop and negative geotaxis reflexes by PBDE 99. Until recently, very little was known about the effects of PBDEs on male or female reproductive development or function. Some work has been conducted on individual PBDE congeners following gestational exposures, evaluating effects on reproductive outcomes in the fetus. For example, Kuriyama et al. (2005) showed a decrease in spermatogenesis in the male rat offspring of dams exposed to a single dose of 60 or 300 lg/kg of PBDE 99 on gestational day (GD) 6. Male rats treated peripubertally with oral exposure to DE-71, a commercial pentabrominated diphenyl ether mixture, not only developed hypothyroxinemia but also displayed significant 2-, 3.5-, and 5-day delays in puberty at 30, 60, and 120 mg/kg body weight,

DEVELOPMENTAL EXPOSURE TO A COMMERCIAL PBDE MIXTURE, DE-71

AR). The presence of impurities (such as brominated biphenyls, dioxins, and furans) in this mixture has been reported elsewhere (Hanari et al., 2006). The dosages 0, 1.7, 10.2, and 30.6 mg/kg/day in corn oil were selected to match doses of Aroclor 1254 (1, 6, and 18 mg/kg) on a molar basis where we have extensive information from both in vitro and in vivo studies (for a review, see Kodavanti, 2005). This will allow us to compare the effects between these two structurally related chemicals (PCBs and PBDEs). Dams (n > 15 per dose group) were weighed and given DE-71 or corn oil (2 ml/kg body weight) by oral gavage daily (between 8:00 and 10:00 A.M.) from GD 6 through PND 21, except on the day of birth (PND 0) when the dams were left undisturbed (Fig. 1). All dams (100% pregnancy rate) delivered within a few hours of each other. Litter sizes between 8 and 15 pups were included in this study. On PND 4, litter size was reduced to eight pups per litter, six males and two females. Six males were used in neurobehavioral testing as well as their reproductive endpoints, whereas the two females were tested for motor activity and mammary gland development. Two litters that had less than eight pups were discarded. The pups were weaned at PND 21. Animals within a treatment group were housed in same-sex pairs. Pup growth and development. Total litter weight was documented twice a week until weaning (Fig. 1). In addition, one male and one female pup from each litter were marked with a permanent pen (Flipchart nontoxic marker) every 2–3 days and their weights were taken twice a week. Neurobehavioral testing. In one series of experiments, neurobehavioral testing was conducted only in male offspring on PNDs 24 and 60 (n ¼ 11 control, 8 low dose, 10 mid-dose, and 13 high dose) using a functional observational battery (FOB) and motor activity. A subset of the same rats was again tested at PND 273 (n ¼ 7 control and low dose and 8 high dose) using open-field rearing; this was done to follow up a suggestive finding obtained on PND 60. Also, previous reports indicate that decabrominated diphenyl ether caused behavioral changes in aged mice but not in young mice after neonatal exposure (Rice et al., 2009). Procedural details and scoring criteria for the FOB protocol are provided elsewhere (McDaniel and Moser, 1993). Briefly, the rats were evaluated for changes in general appearance, lacrimation, salivation, as well as a ranking of the rat’s reactivity to being handled. Open-field measurements included ranking the rat’s arousal and activity level, number of rears, as well as any tremorigenic activity. The rat’s reactions to a click stimulus, tail pinch, toe pinch, or penlight were ranked. Also tested were the

palpebral, pinna, and proprioceptive placing (both forelimb and hind limb) responses. Forelimb and hind limb grip strength, landing foot splay, and rectal temperature were quantified. Motor activity data were collected shortly after FOB testing, using a photocell chamber shaped like a figure eight (Reiter, 1983). Photobeams were placed within the chamber to compile both horizontally and vertically directed activity. Activity counts were recorded in twelve 5-min intervals to evaluate habituation of activity during the session. For all testing, the observer was blind with respect to the treatment group. In a second experiment, motor activity of both male and female offspring (littermates of those tested in the FOB) was measured in photocell devices of a different configuration (Motron Electronic Motility Meter, Stockholm, Sweden). Each device had a platform that contained a 5 3 8 matrix of photodetectors that were illuminated by an overhead incandescent (30 W) bulb. Any movement that occluded a photodetector was recorded as a count of horizontal activity. Each device also had an array of six photoemitters and detectors oriented in a horizontal plane 16.5 cm above the platform floor. An interruption of this plane of light was recorded as a count of vertical activity (or rearing). A clear Plexiglas chamber (33 3 21 3 26 cm), with a removable lid, was placed over each platform to contain the rat during testing. Unlike the figure-eight maze, rearing behavior could be recorded throughout the chamber. Each device was housed in a ventilated enclosure that eliminated extraneous sound and light. Motor activity was recorded in five successive 6-min intervals (total test time ¼ 30 min). All testing took place in the morning. Test sessions were conducted when the rats were 100 days of age and again at 114 days of age. Order of testing at each age was counterbalanced across dose groups and testing device; however, males and females were always tested on alternate days. This experiment was designed (1) as a check on the generality of any activity changes found in the open-field and figure-eight devices and (2) an evaluation of the potential effects of DE-71 on between-session (long-term) habituation. Anogenital distance. On PND 7, anogenital distance (AGD; the distance between the genital papilla and the anus) was measured using calipers in the male offspring. Normally, male rodents have AGDs that are approximately twice the length of those of females (Gray et al., 1999; Vandenbergh and Huggett, 1995). The AGD has been shown to be a predictor of a number of adult physiological and behavioral characteristics and is also sexually dimorphic in humans (Salazar-Martinez et al., 2004). One male per litter was used in the analysis (n ¼ 7–10). Preputial separation. The separation of the foreskin from the glans penis, preputial separation (PPS), is an early reliable marker of the onset of puberty in male rats (Korenbrot et al., 1977). In the present study, PPS was monitored beginning on PND 30 and continued until all males showed separation. All males were examined once daily at approximately the same time each day. A partial separation with a thread of cartilage remaining was recorded as ‘‘partial,’’ but only the day of complete separation was used in the data analyses. VO in the females was not assessed in the current study because the highest dose in this study was the no observed adverse effect level in the previous pubertal study (Stoker et al., 2004). Two pups per litter were examined for PPS, and litter means were used in the statistical analysis (n ¼ 7–13). Tissue collection and weights. Following decapitation of animals dedicated for male reproductive endpoints at PND 60, the seminal vesicles (fluid-filled), ventral prostate lobes, right epididymis, and right testis were removed and weighed. Sera and gut (stomach) contents were collected from pups on PND 4. Dam sera and mammary gland tissues were collected on PND 22 (the day after weaning) for TH and PBDE analyses (see below). Milk collection. Lactating dams with equal number of pups in each dose group were milked on PND 10 to determine the amount and types of PBDEs in milk. Pups were removed from dams for 2 h prior to injecting dams with oxytocin (5 IU from Sigma; im). Twenty minutes later, dams were injected with ketamine:xylazine (35:5 mg/kg, ip; Sigma). A milking device was created using a vacuum system, a side-arm flask with a one-holed rubber stopper, and tubing from a commercial breast pump. Milk was collected until a stream of milk was no longer available. All milk samples were brought to 1 ml with Krebs buffer and used for analysis of PBDEs.

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

FIG. 1. Dosing and testing paradigm for the developmental exposure to a commercial PBDE mixture, DE-71. The exposure started at GD 6 and continued through PND 21. Body weights of offspring were taken from birth until PND 60. Neurobehavioral testing started on PND 24 and continued through PND 273. Mammary gland development was evaluated on PNDs 4 and 21. Male reproductive endpoints were examined on PND 7 through PND 60. THs were measured from PND 4 through PND 60.

299

300

KODAVANTI ET AL.

Radioimmunoassay of circulating THs and testosterone. Trunk blood samples were collected on PNDs 4, 7, 14, 21, and 60 from at least one male and female pup per litter per treatment after decapitation. After collection, blood samples were placed on ice for 1 h to allow for blood clotting. The samples were centrifuged at 1000 3 g for 15 min to separate serum. These serum samples were stored at
80C until radioimmunoassay (RIA). Total T4, triiodothyronine (T3), and testosterone were determined using Coat-a-Count kits from Siemens Medical Solution Diagnostics (Los Angeles, CA). All samples were measured in duplicate. The intra- and interassay variations were below 10% for these hormones. RIA of thyroid-stimulating hormone. The thyroid-stimulating hormone (TSH) RIA was performed using the following materials supplied by the National Hormone and Pituitary Agency: iodination preparation (I-9-TSH), reference preparation (RP-3), and antisera (S-6-TSH). Iodination material was radiolabeled with 125I (Perkin Elmer, Shelton, CT) by a modification of the chloramine-T method (Greenwood et al., 1963). Labeled antigen was separated from unreacted iodide by gel filtration chromatography as described previously (Goldman et al., 1986). Sample serum was diluted to a final assay volume of 500 ml with 100mM phosphate buffer containing 1% bovine serum albumin (BSA). Standard reference preparation was serially diluted for the standard curve. Primary antisera (200 ll) in 100mM potassium phosphate, 76.8mM EDTA, 1% BSA, and 3% normal rabbit serum (pH 7.4) were pipetted into each assay tube, vortexed, and incubated at 5C for 24 h. A 100-ll aliquot of the iodinated hormone was then added to each tube, and the tube was vortexed and incubated for 24 h. A second antibody (goat anti-rabbit gamma globulin, Calbiochem, at a dilution of 1 unit/100 ll) was then added, vortexed, and incubated for 24 h. The samples were centrifuged at 1260 3 g for 30 min and the supernatant aspirated, and the sample tube, with pellet, was counted on a gamma counter. PBDE analysis. Concentrations of PBDEs in the gut contents at PND 4 (culled pups used in mammary gland analyses), expressed milk samples at PND 10, and dam mammary gland tissue at PND 22 (all dams were euthanized at this time) were determined by a method previously described (Johnson-Restrepo et al., 2005). Different brain regions (frontal cortex, cerebellum, hippocampus, striatum, and hypothalamus), liver, blood, and fat tissues from PND 22 male offspring were also analyzed for congener-specific PBDEs to understand distribution of PBDEs in circulation and some target organs. Samples (0.1–1.1 g wet) were homogenized with anhydrous sodium sulfate and were Soxhlet extracted using dichloromethane and hexane (3:1 vol/vol; 400 ml). Fat content was determined gravimetrically from an aliquot of the extract. PCB congeners CB-30 and CB-204 were spiked as surrogate standards prior to extraction. The extracts were spiked with 13C-labeled PBDE congeners (EO-51000-10X; Cambridge Isotope Laboratories, Andover, MA) as internal standards and were then purified by passage through 6 g of a multilayer silica

gel (100–200 mesh) packed in a glass column (10 mm inner diameter). The silica gel column was prepared by packing 2 g of silica gel, followed by 2 g of 40% acidic silica, and then by 2 g of silica gel. The silica gel bed was washed with 100 ml of hexane prior to loading of the sample extract. Extracts were then eluted with 150 ml of 25% dichloromethane in hexane. Samples were further subjected to lipid removal by treatment with concentrated sulfuric acid, if and when needed. Sample extracts (2 ll) were injected into a Hewlett-Packard 6890 gas chromatograph interfaced with a Hewlett-Packard 5973 mass spectrometer. Injections were made in the splitless mode, and samples were separated on a 30-m DB-5 (5% diphenyl/dimethylpolysiloxane) analytical capillary column with a 250-lm i.d. and a 0.25-lm film thickness. The oven temperature program was set to 100C for 1 min, 10C/min to 160C, hold 3 min, and 2.5C/min to 260C, hold 10 min. The inlet and interface temperatures were set to 270C and 300C, respectively. The mass spectrometer was operated in an electron impact mode (70 eV) and selected ion-monitoring (SIM) mode. PBDE congeners were identified and quantified by SIM at m/z 406, 408; 486, 484; 564, 566; and 642, 644 for tri-, tetra-, penta-, and hexa-BDEs, respectively. Quantification was based on an external calibration standard. Twenty-three congeners representing tri- through hexa-BDEs (congener #17, 28, 51, 49, 48, 47, 66, 102, 100, 119, 91, 99, 85, 155, 154, 153, 139, 140, and 138) including one unidentified tetra-BDE, one penta-BDE, and two hexa-BDE were quantified. The detection limit of individual PBDE congeners was 0.01 ng/g wet weight. The mean recoveries of PCB-30 and PCB-204, spiked into samples prior to extraction, were 86% and 103%, respectively. Recovery of 13C-PBDEs, spiked prior to lipid removal, ranged from 86 to 99%. The measured concentrations were not corrected for the recoveries of internal or surrogate standards. Statistics. Statistical analyses on the behavioral data in the first series of studies were conducted using analysis of variance (ANOVA) for continuous measures and the Kruskal-Wallis test for ranked (nonparametric) data. The two early test times (PNDs 24 and 60) were included as repeated measures, but the data from the PND 273 (open-field) test were analyzed separately because many fewer rats were included. For the second neurobehavioral experiment (using Motron photocell device), session-total motor activity counts were analyzed using a three-way repeated measures ANOVA, with dose and sex as main factors and test-session number (initial test and retest) as the repeated factor. Separate analyses were carried out for horizontal and vertical activity. Behavioral data were analyzed using SAS software (SAS, Inc., Cary, NC). Body weight and TH data were analyzed by two-way ANOVA with age and treatment as two factors using SigmaStat software, version 3.0 (SPSS Inc., Chicago, IL) followed by Fisher’s least significant difference post hoc test. Male reproductive data were analyzed for age and treatment effects by ANOVA using Prism (GraphPad Software, San Diego, CA) and then tested for homogeneity of variance using Bartlett’s test. Mammary gland data analyses were performed by ANOVA in SAS (SAS, Inc.). Statistical significance was taken as p < 0.05. When significant treatment effects were indicated by ANOVA (i.e., significant F-statistic), Dunnett’s multiple comparison test was used to compare each treatment group with the control group. Data are presented as mean ± SEM for each treatment group and endpoint. All the data in this study were analyzed using the dam or litter as the experimental unit.

RESULTS

Maternal Body Weight Gain and Offspring Development The body weights of all the dams steadily increased throughout gestation followed by an expected drop after delivery and remained steady thereafter. DE-71 exposure did not affect the normal pattern of maternal body weights throughout gestation and lactation regardless of dose (Fig. 2A). In support of this, ANOVA indicated no interaction between age and treatment; as expected, there was a significant effect only with

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

Mammary gland development. Mammary tissue was collected from female pups for evaluation of puberty-related mammary gland development. On PNDs 4 (five litters; four to eight pups per litter) and 21 (eight litters; one pup per litter), the entire fourth and fifth mammary glands from both sides were removed and mounted skin side down on glass slides. Whole mounts were fixed in Carnoy’s solution, stained in alum carmine stain, and dehydrated and cleared in xylene, as previously described (Fenton et al., 2002). Flattened whole mounts from female offspring were scored on a 1–4 age-adjusted developmental scale (as described by Rayner et al., 2004; 1 ¼ poor development/structure; 4 ¼ normal development/structure, for a given age). The developing tissue was assessed for number of primary ducts, number of large secondary ducts, lateral side branching, terminal end buds as a percentage of all duct ends, appearance of budding from the ductal tree, and longitudinal outgrowth of the epithelia. Slides were separated by score as they were evaluated, compared within a score for consistency, and then recorded. Two independent scorers, blind to treatment, scored glands within the age groups. Mean scores for the two ages, within treatment groups, were calculated and analyzed statistically for treatment and body weight or time-related differences using the dam or litter as the unit. Mammary glands representative of the mean score of the group were photographed on a Leica WILD M420 macroscope.

DEVELOPMENTAL EXPOSURE TO A COMMERCIAL PBDE MIXTURE, DE-71

age (F82,1872 ¼ 5.87, p < 0.001). The pregnancy rate was 100% in all dose groups. The litter size ranged from 7 to 15 pups with an overall mean of 9.6 pups per litter and did not differ significantly across dose groups. Two litters had only female pups (one litter in control and one litter in 30.6 mg/kg dose groups) and were not included in the study. Pup mortality ( 0.05) and there was no treatment-related effect on PND 273. Motor activity measured in the figure-eight chambers on PNDs 24 and 60 (data not presented) showed a significant age difference, but there was no age-by-treatment interaction (p > 0.05) on either horizontal or vertical activity. Perinatal treatment with DE-71 was also ineffective when the activity of adult rats was tested at an intermediate age (ca. 110 days) in a different photocell (Motron) device (Fig. 5). Overall, horizontal activity was significantly higher in male versus female rats (F1,47 ¼ 24.27, p < 0.0001). Activity levels on retest (session 2) were slightly, yet uniformly, lower in males but not in females. These results are supported by a significant effect of test session (F1,47 ¼ 4.20, p < 0.04) and an interaction with gender (F1,47 ¼ 4.41, p < 0.04). There was no significant effect of dose and no interaction of either gender or test session with dose. For vertical activity, there was no significant effect of sex and no significant dose-by-sex interaction. Retesting (session 2) resulted in significantly lower vertical activity levels overall (F1,47 ¼ 8.68, p < 0.005), although there was no significant interaction with either dose group or sex (and no significant three-way interaction). Circulating THs and TSH

Male Offspring Reproductive EndPoints

Developmental exposure to DE-71 induced significant hypothyroxinemia in the dams (Fig. 6). The ANOVA on total T4 levels in the sera of dams at PND 22 indicated a significant

There were no significant differences in the AGD of the male offspring on PND 7 in any of the DE-71 dose groups as compared with the controls (Fig. 9A). There was a 0.4-mm or

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

FIG. 4. Open-field rearing of males at PNDs 24, 60, and 273 following in utero and lactational exposure to DE-71. Each value is a mean ± SE. The increase between PNDs 24 and 60 in treated males did not reach statistical significance (p < 0.10).

effect of DE-71 dose (F3,40 ¼ 39.28, p < 0.001) as T4 levels were significantly lower in 10.2 and 30.6 mg/kg dose groups compared with controls. However, total T3 levels in maternal sera were not altered by DE-71 exposure (Fig. 6). Maternal DE-71 exposure produced marked hypothyroxinemia in offspring of exposed dams (Fig. 7). In both male and female offspring, perinatal exposure produced an age-dependent depression in circulating total T4 concentrations that suppressed the normal increase in T4 levels seen between birth and PND 21. At 30.6 mg/kg, the effect on circulating T4 levels in male offspring started with a 52% decrease at PND 4 and 60% decrease at PND 7 (Fig. 7). This effect increased to 78% drop on PND 14 and 74% drop on PND 21. Similarly in female offspring of 30.6 mg/kg dose group, the decreases in total T4 levels were 56%, 50%, 80%, and 76% at PNDs 4, 7, 14, and 21, respectively (Fig. 7). The dose-related effects on circulating T4 levels were confirmed by a significant dose-by-age interaction for both sexes (F12,92 ¼ 5.812, p < 0.001 for males; F12,104 ¼ 6.382, p < 0.001 for females). The post hoc test indicated a significant effect of DE-71 starting at 10.2 mg/kg dose group, between ages PNDs 7 and 14. At PND 60, after cessation of exposure at weaning, total T4 levels in male offspring of DE-71–exposed groups returned to control levels (Fig. 7). However, total T4 levels in female offspring were higher at PND 60 following DE-71 exposure compared with controls. Total T3 levels were not significantly altered in either male or female offspring at any age or in any dose group following perinatal exposure to DE-71 (Fig. 8). Developmental exposure to DE-71 significantly increased serum TSH levels in dams at PND 22 and in female offspring at PND 60 (Table 2). In the dams, ANOVA indicated a significant effect of DE-71 dose (F3,42 ¼ 2.894, p < 0.046) with a significant increase at 30.6 mg/kg dose (Table 2). The offspring TSH data were highly variable. In the offspring, serum TSH levels increased during postnatal development. ANOVA indicated a significant effect of PND in male (F4,113 ¼ 24.69, p < 0.001) and female offspring (F4,92 ¼ 10.06, p < 0.001). Although DE-71 exposure did not significantly alter serum TSH levels in male offspring, serum TSH levels were lower in female offspring at PND 60 (Table 2). In support of this, ANOVA indicated a significant interaction of PND by DE-71 treatment (F12,92 ¼ 2.176, p < 0.019). However, the decreases in serum TSH levels in females should not be given much weight as we noticed that the control TSH values at PND 60 were unusually higher than the historic values that typically range from 1.5 to 2.5 ng/ml. The replicates for these control TSH levels were comparable, and the quality control serum was within the expected range.

DEVELOPMENTAL EXPOSURE TO A COMMERCIAL PBDE MIXTURE, DE-71

303

5.5% difference between the mean AGD in the high-dose DE-71 group and the control mean. Although this difference was not significant, there does appear to be biological relevance as this measure is not confounded by body weight (Fig. 2) and there was also a delay in PPS and a 20% decrease in mean testosterone concentration in the 30.6 mg/kg dose group on PND 60. Therefore, a higher gestational dose may be required to cause significant changes in this endpoint and subsequently more permanent alterations in testosterone secretion in the adult. DE-71 exposure to the dam from pregnancy through PND 21 dose dependently delayed the age of PPS by 1.0 and 1.8 days in the 10.2 and 30.6 mg/kg dose groups, respectively (Fig. 9B). However, only the 1.8-day delay in the 30.6 mg/kg group was significantly different compared with the control males. As described in Figure 3, there was no significant difference in body weight at the time of puberty (PNDs 40–45), so the delay in PPS is not a result of altered body weight. There were no differences in the PND 60 weights of the seminal vesicles,

epididymis, testis, or ventral prostate lobes as compared with the control weights (Fig. 9C). Although there appeared to be a decrease in the mean serum testosterone concentration at the 30.6 mg/kg dose (~20%), there were no significant differences between the control mean and any of the DE-71 dose groups (Fig. 9D). Mammary Gland Development in Female Offspring To determine whether or not the commercial PBDE mixture acts on the developing mammary tissue of the female neonate, an important endpoint in female pubertal progression, mammary glands were removed from female pups on PNDs 4 and 21, fixed and stained with carmine, and evaluated for agedependent development on a scale of 1–4. As shown in Table 1 and Figure 10, there was no significant treatment-related effect on PND 4 mammary gland development (F3,67 ¼ 2.03, p < 0.1). This may be due to the unusually low scores of the

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

FIG. 5. Horizontal and vertical motor activity in males and females at PNDs 100 and 114 following in utero and lactational exposure to DE-71. Each value is a mean ± SE. There were no significant effects of treatment with DE-71. Numbers in parenthesis are sample sizes.

304

KODAVANTI ET AL.

controls, which are typically closer to 4 than 3 in animals of this strain given other vehicle types (water or 1% methylcellulose in water; Fenton et al., 2002; Rayner et al., 2004). The glands of the 10.2 and 30.6 mg/kg groups displayed abnormally less longitudinal epithelial growth than is typical. However, there was a significant effect of dose in mammary developmental scores in DE-71–exposed female offspring by PND 21 (F3,28 ¼ 12.34, p < 0.001). The epithelial portion of glands of female pups exposed to 10.2 and 30.6 mg/kg of DE-71 throughout the lactation period exhibited a significant lack of outgrowth, fewer lateral branches, and limited terminal end bud development compared with control glands (p < 0.01). The DE-71 effect was most pronounced at 10.2 mg/kg. Concentrations of PBDEs in Pup Gut, Milk, and Lactating Mammary Gland PBDE concentrations were measured in neonate gut contents at PND 4, in mother’s milk at PND 10, and in dam mammary gland at PND 22, 1 day after weaning. These measures were

FIG. 7. Total serum thyroxine (T4) levels in pups at PND 4 through PND 60 following developmental exposure to DE-71. Each value is a mean ± SE. *Significantly different from control at p < 0.05.

taken to elucidate early exposure in neonates, during the period of rapid brain development, and to determine residual concentrations post lactation, respectively. This was done only for the samples collected in the control and 30.6 mg/kg dose groups. On a wet weight basis, total PBDE concentrations increased with dosing time to nearly 500 lg/g in the involuting mammary gland. The milk data that were represented on a wet weight basis are confounded by the buffer dilution and do not represent the actual value. Therefore, we normalized all the values to lipid content of that tissue for uniformity. When normalized to lipid content, PBDE levels were similar in all three measures (Fig. 11A), indicating similar PBDE exposure to the offspring with time. Because all three measures belong to the dam, similar levels with time suggest that PBDE levels are in equilibrium in the dam’s body. When the individual congener profiles in the three selected sample types were evaluated, the profiles of various congeners in mammary gland resembled those of the PBDE mixture,

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

FIG. 6. Total serum thyroxine (T4) and triiodothyronine (T3) levels in dams at PND 21 following exposure to DE-71 from GD 6 through PND 21. Each value is a mean ± SE. *Significantly different from control at p < 0.05.

DEVELOPMENTAL EXPOSURE TO A COMMERCIAL PBDE MIXTURE, DE-71

305

DISCUSSION

FIG. 8. Total serum triiodothyronine (T3) levels in pups at PND 4 through PND 60 following developmental exposure to DE-71. Each value is a mean ± SE. No significant differences were detected among treatment groups.

DE-71, with slight increases in the proportions of PBDEs 100 and 153. On the other hand, the proportions of PBDE 47 were greater, whereas those of PBDE 99 were lower in the neonate gut and milk when compared with those in the DE-71 mixture (Fig. 11B). Concentrations of PBDEs in Brain Regions, Blood, Liver, and Fat PBDE congeners were found in all the rat tissues examined on PND 22 including different brain regions (Fig. 12). Concentrations were greatest in fat tissue followed by liver. The PBDE concentrations in fat tissues ranged from 627 to 923 lg/g wet weight. On a wet weight basis, concentrations in blood were more than fivefold lower than the concentrations found in brain and about three orders of magnitude different than the amount in fat. PBDE concentrations in frontal cortex, cerebellum, hippocampus, striatum, and hypothalamus were comparable (Fig. 12A). When adjusted for lipid content,

Few reports exist in the literature regarding developmental effects of PBDEs on human reproductive and nervous systems. Our previous studies indicated that PBDE mixtures as well as individual congeners are anti-androgenic (Stoker et al., 2005) and that they could perturb intracellular signaling pathways, including protein kinase C and calcium homeostasis in neuronal cells, both of which are processes that are involved in nervous system development (Coburn et al., 2008; Kodavanti et al., 2005). Based on these previous results as well as other animal studies (Birnbaum and Staskal, 2004; Costa and Giordano, 2007), we have now conducted a study to understand the developmental effects of a pentabrominated diphenyl ether mixture, DE-71, on reproductive and nervous system development in rats. The results indicate subtle changes in some parameters of neurobehavior but substantial effects on TH homeostasis and female reproductive system development. Perinatal exposure to DE-71 commercial mixture did not cause any gross changes in maternal body weight or male offspring weight. However, female offspring following DE-71 exposure demonstrated reduced body weights compared with controls starting at PND 29 that remained lower until PND 60. This observation is consistent with recent reports in the Center for the Health Assessment of Mothers and Children of Salinas population where Harley et al. (2009) reported that human maternal serum PBDE concentrations at the beginning of the third trimester of pregnancy are inversely associated with birth weight. Harley et al. (2009) indicated that every 10-fold increase in total PBDEs was associated with a 148-g reduction in birth weight. Neurobehavioral testing included an FOB, open-field activity, and measurements of motor activity in two automated

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

corrected PBDE concentrations in the liver and fat were more similar, and the brain levels were generally lower than those in blood except striatum. Among the 23 congeners measured, PBDE 47 was predominant in all the tissues, accounting for 44–50% (mean 47%) of the total PBDE concentrations (Fig. 12B). This was followed by BDE 99 and 100, respectively. These three congeners collectively accounted for 84% of the total PBDE concentrations in tissues. The distribution of PBDE congeners in all the tissues including different brain regions was similar. PBDEs 47, 153, and 138 were enriched in rat tissues relative to those found in DE-71 mixture. On the contrary, PBDEs 99 and 154 were depleted in rat tissues relative to those in DE-71 mixture. These results imply differences in the accumulation potential of PBDE congeners. However, it is not clear if the depletion of PBDEs 99 and 154 could contribute to the enrichment of PBDE 47. This study documents the occurrence of PBDE congeners in rat brain and that the profile of brain PBDE congeners was similar to those in other tissues.

306

KODAVANTI ET AL.

devices. Open-field rearing in the DE-71–exposed groups increased from PND 24 to 60 but decreased again on PND 273. In contrast, control levels remained the same. Although these differences are suggestive, they did not attain significance with the sample size available. Increased rearing in adult male mice has been consistently reported by Eriksson, Viberg, and colleagues for acute exposure to several PBDE congeners during rapid brain development (Eriksson et al., 2001, 2002; Viberg et al., 2002, 2003a,b, 2004, 2006, 2007). More recently, increases in rearing have been reported for adult mice treated with PBDE 47 and tested in a different activity device (Motron chambers) (Gee and Moser, 2008). Furthermore, there is a growing literature suggesting altered activity in developing and/or adult offspring following various perinatal treatments with PBDEs (Branchi et al., 2002, 2003, 2005; Kuriyama et al., 2004, 2005). The finding of increased rearing in the present study was expected in light of these other studies, but its reliability is suspect due to both the lack of statistical

significance and concordance with two automated measures of rearing. The lack of effects on any other measures of the FOB was striking. It should be noted that the rats were tested only as young adult/adults. In contrast, others have reported altered ontogeny of specific neuromotor functions in developing offspring (Gee and Moser, 2008; Rice et al., 2007). Whereas many neurobehavioral studies of PBDEs have used mice, other studies have also tested PBDE-exposed rats and reported neurological alterations such as changes in activity, visual discrimination, and reproductive behaviors (Dufault et al., 2005; Kuriyama et al., 2004, 2005; Viberg et al., 2007). In the present study, some of the neurobehavioral effects of DE-71 exposure in rats were subtle. Recent epidemiological study shows neurodevelopmental effects in relation to cord blood PBDE concentrations where children with higher PBDE concentrations scored lower on tests of mental and physical development (Herbstman et al., 2010). Roze et al. (2009)

TABLE 1 Mammary Gland Development Scores in Offspring Exposed to DE-71 during In Utero and Lactation Periods Age at scoring PND 4 PND 21

Sample n

Corn oil controla

DE-71 (1.7 mg/kg)

DE-71 (10.2 mg/kg)

DE-71 (30.6 mg/kg)

5 litters 4–8 pups/litter 8 litters 1 pup each

3.0 ± 0.04 3.3 ± 0.1

2.8 ± 0.04 2.9 ± 0.06

3.0 ± 0.05 1.5 ± 0.08b

2.6 ± 0.04 2.0 ± 0.05b

Note. Data are shown as mean ± SE. a Developmental scores in this group are low compared with historical vehicle controls (water or 1% methylcellulose in water). b Significantly different from control by Dunnett’s test, p < 0.01.

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

FIG. 9. Measures of male reproductive development following gestational and lactational exposure to DE-71. (A) AGD on PND 7; (B) age of PPS; (C) reproductive tissue weights (VP, ventral prostate lobe; Sem Ves, seminal vesicle; R. Testis, right testis; R. Epi, right side epididymis) on PND 60; and (D) serum testosterone concentrations of the male offspring on PND 60. Each value is a mean ± SE. *Significantly different from control at p < 0.05. AGD was done on one male pup per litter, and PPS was done on two pups per litter, but litter mean was used in statistics.

DEVELOPMENTAL EXPOSURE TO A COMMERCIAL PBDE MIXTURE, DE-71

307

reported that PBDEs as well as hydroxyl-PCBs correlated with motor performance, attention, and visual perception, and behavior in school-age children. THs play a critical role in brain function (Ford and Cramer, 1977; Porterfield and Hendrich, 1993). Altered TH status during development is known to produce motor incoordination, hearing loss, and memory deficits (Porterfield, 1994). Several reports indicate decreases in circulating THs following developmental exposure to commercial PBDE mixtures such as DE-71 (Zhang et al., 2009) or individual congeners such as PBDE 99 (Kuriyama et al., 2007). Our results support the finding that in utero and lactational exposure to DE-71 causes a hypothyroid state in the dam and offspring, leading to a maximal 80% drop in circulating total T4 levels. Importantly, these findings confirm that PBDEs can affect thyroid function in adults as well as their developing offspring. We have recently investigated the possible mechanisms by which

PBDEs decrease circulating THs in the same cohort of animals (Szabo et al., 2009). The results indicate that deiodination, active transport, and sulfation, in addition to glucuronidation, may be involved in the disruption of TH homeostasis due to perinatal exposure to DE-71 in rat offspring (Szabo et al., 2009). In addition, literature reports indicate the presence of THs in human and animal milk (Karimova et al., 1983), and in suckling rats, maternal milk is a source of iodide and nonhormonal protein–bound iodine (Vigouroux et al., 1980). The significant decreases in TH in offspring could also be due to maternal hypothyroxinemia caused by DE-71 exposure. The significant effect seen in dams on TSH increases (Table 2) is in agreement with reports where increased serum PBDE concentrations were associated with increased TSH levels in Chinese workers from an E-waste dismantling site (Yuan et al., 2008). Recent studies in humans showed that serum PBDE concentrations were positively associated with free T4 levels

FIG. 11. Concentrations of PBDEs in neonate’s gut (stomach) contents at PND 4, in milk samples at PND 10, and in dam mammary gland at PND 22 following developmental exposure to DE-71 (30.6 mg/kg/day). (A) Total PBDE concentrations on wet weight and lipid weight basis and (B) congener composition across sample type are presented. Each value is a mean ± SD of three samples from three different animals. Control values were 3.1, 35.4, and 41 ng/g wet weight and 11.7, 1230, 149 ng/g lipid weight for gut, milk, and mammary gland, respectively.

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

FIG. 10. Prepubertal mammary gland development in pups at PNDs 4 and 21 following developmental exposure to DE-71. Significant developmental delays were observed at PND 21 at 10.2 and 30.6 mg/kg DE-71. Magnifications differ between PNDs 4 and 21 and are noted.

308

KODAVANTI ET AL.

(Meeker et al., 2009). The preferential effect of PBDEs on serum T4 levels could have neurological implications because 80% of the THs in the brain are derived from serum T4 (Silva and Matthews, 1984). Similar decreases in circulating T4 levels were seen in pups following developmental exposure to PCBs (Crofton et al., 2000), suggesting that PCBs and PBDEs not only are structurally similar but also exert some similar biochemical effects. Not only do PBDEs alter thyroid function, they have also been shown to be anti-androgenic. With increasing concentrations being detected in the environment and in tissue/milk samples of humans and animals (Frederiksen et al., 2009; Schecter et al., 2005), these dual-role endocrine disruptors have raised new concerns about possible effects on human reproductive development. Anti-androgenic environmental chemicals are observed in wildlife populations (Guillette et al., 1994; Kloas et al., 2009) and have the potential to adversely impact human reproductive health. Such an effect would be important because these chemicals are persistent in the environment and are relatively ubiquitous, being found in such places as indoor dust (Johnson-Restrepo and Kannan, 2009; Sjodin et al., 2008; Stapleton et al., 2005). In a recent human study, Meeker et al. (2009) reported a significant inverse relationship between dust PBDE concentrations and free serum androgen index. Dust PBDE concentrations were strongly and inversely associated with luteinizing hormone and follicle-stimulating hormone and positively associated with inhibin B and sex hormone–binding globulin. In addition, Main et al. (2007) reported a positive correlation between the sum of PBDEs and serum luteinizing hormone, and concentrations of PBDEs in mothers’ breast milk were significantly higher in boys with cryptorchidism.

The observation of a delay in PPS in the DE-71–exposed male offspring in this study indicates that there is a decreased response to the circulating androgens or a suppression of androgen function. As mentioned, DE-71 was shown to be anti-androgenic in several in vivo and in vitro assays

TABLE 2 TSH Concentrations in Dams at PND 22 and in Offspring from PND 4 through 60 Following Exposure to DE-71 from GD 6 through PND 21 TSH (ng/ml) Dams/ offspring Dams at PND 22 Male offspring PND 4 PND 7 PND 14 PND 21 PND 60 Female offspring PND 4 PND 7 PND 14 PND 21 PND 60

Control

1.7 mg/kg

2.543 ± 0.248 5.080 ± 1.019

10.2 mg/kg

30.6 mg/kg

3.810 ± 0.468

5.768 ± 1.272a

1.209 1.204 1.602 2.495 1.839

± ± ± ± ±

0.191 0.151 0.106 1.455 0.650

1.299 1.264 1.668 1.726 4.202

± ± ± ± ±

0.074 0.178 0.247 0.128 0.952

1.175 1.581 1.918 1.052 7.632

± ± ± ± ±

0.168 0.128 0.216 0.059 2.093

1.275 1.346 1.845 2.103 6.953

± ± ± ± ±

0.161 0.173 0.110 0.444 1.256

0.759 1.417 1.895 1.713 3.289

± ± ± ± ±

0.164 0.112 0.276 0.455 0.370

1.406 2.086 2.832 1.355 2.012

± ± ± ± ±

0.950 0.172 0.400 0.149 0.401a

0.856 1.585 1.568 1.293 2.243

± ± ± ± ±

0.222 0.240 0.091 0.141 0.551a

0.924 1.133 1.754 1.752 1.868

± ± ± ± ±

0.141 0.233 0.313 0.239 0.253a

Note. Values are expressed as mean ± SE. n ¼ 5–12 rats. a Significantly different from control at p < 0.01.

Downloaded from http://toxsci.oxfordjournals.org at Environmental Protection Agency Library on June 23, 2010

FIG. 12. Concentrations of PBDEs in selected brain regions (frontal cortex, cerebellum, hippocampus, striatum, and hypothalamus), blood, liver, and fat tissues from male offspring at PND 22 following developmental exposure to DE-71 (30.6 mg/kg/day). (A) Total PBDE concentrations on wet weight and lipid weight basis and (B) congener composition across sample type are presented. Each value is a mean ± SD of three samples from three different animals. Control values were