Comprehensive Assessment of a Chlorinated Drinking Water

Aug 2, 2013 - National Risk Management Research Laboratory, Office of Research and Development, ... National Center for Environmental Assessment, Offi...
7 downloads 0 Views 656KB Size
Download PDF
Article pubs.acs.org/est

Comprehensive Assessment of a Chlorinated Drinking Water Concentrate in a Rat Multigenerational Reproductive Toxicity Study Michael G. Narotsky,*,† Gary R. Klinefelter,† Jerome M. Goldman,† Deborah S. Best,† Anthony McDonald,† Lillian F. Strader,† Juan D. Suarez,† Ashley S. Murr,† Inthirany Thillainadarajah,† E. Sidney Hunter, III,† Susan D. Richardson,‡ Thomas F. Speth,§ Richard J. Miltner,§ Jonathan G. Pressman,§ Linda K. Teuschler,⊥ Glenn E. Rice,⊥ Virginia C. Moser,† Robert W. Luebke,† and Jane Ellen Simmons† †

National Health and Environmental Effects Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711, United States ‡ National Exposure Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Athens, Georgia 30605, United States § National Risk Management Research Laboratory, Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati, Ohio 45268, United States ⊥ National Center for Environmental Assessment, Office of Research and Development, U.S. Environmental Protection Agency, Cincinnati, Ohio 45268, United States S Supporting Information *

ABSTRACT: Some epidemiological studies report associations between drinking water disinfection byproducts (DBPs) and adverse reproductive/developmental effects, e.g., low birth weight, spontaneous abortion, stillbirth, and birth defects. Using a multigenerational rat bioassay, we evaluated an environmentally relevant “whole” mixture of DBPs representative of chlorinated drinking water, including unidentified DBPs as well as realistic proportions of known DBPs at low-toxicity concentrations. Source water from a water utility was concentrated 136-fold, chlorinated, and provided as drinking water to Sprague−Dawley rats. Timed-pregnant females (P0 generation) were exposed during gestation and lactation. Weanlings (F1 generation) continued exposures and were bred to produce an F2 generation. Large sample sizes enhanced statistical power, particularly for pup weight and prenatal loss. No adverse effects were observed for pup weight, prenatal loss, pregnancy rate, gestation length, puberty onset in males, growth, estrous cycles, hormone levels, immunological end points, and most neurobehavioral end points. Significant, albeit slight, effects included delayed puberty for F1 females, reduced caput epidydimal sperm counts in F1 adult males, and increased incidences of thyroid follicular cell hypertrophy in adult females. These results highlight areas for future research, while the largely negative findings, particularly for pup weight and prenatal loss, are notable.

■

INTRODUCTION

drinking water, understanding and quantifying associated risks are important public health issues. Although >600 chemically distinct DBPs have been identified,9 approximately half of the mass of total organic halogen in chlorinated water remains unidentified. Thus, in the absence of a definitive epidemiological data set, estimation of the human health risks posed by exposures to DBPs is best developed through toxicological evaluations of the complex mixture itself. In addition, evaluation

Disinfection of drinking water is arguably the most important public health advance of the 20th century, leading to drastic reductions in morbidity and mortality. Oxidizing disinfectants, e.g., chlorine, react with organic matter in the water, forming disinfection byproducts (DBPs) that may pose health risks for reproductive and developmental effects including low birth weight, spontaneous abortion, stillbirth, and birth defects.1−6 However, the epidemiological data are not definitive because studies have yielded contrasting results.7,8 Animal toxicity studies have identified similar adverse effects of individual DBPs but at much higher doses than those in human drinking water exposures.8 Because of the widespread use of disinfected This article not subject to U.S. Copyright. Published 2013 by the American Chemical Society

Received: Revised: Accepted: Published: 10653

June 13, 2013 July 30, 2013 August 2, 2013 August 2, 2013 dx.doi.org/10.1021/es402646c | Environ. Sci. Technol. 2013, 47, 10653−10659

Environmental Science & Technology

Article

2.5-mil modified polytetrafluoroethylene to fully enclose the contents. The lining was equipped with two fittings for pumping water into and out of the drum, adding potassium bromide and sodium hypochlorite, and removing headspace. To replace bromide lost during concentration and to create a mixture of chlorinated/brominated DBPs representative of waters containing moderate bromide levels,18 potassium bromide was added to achieve a concentration corresponding to 74 μg of bromide/L in unconcentrated water. After 5 min of mixing, the solution was held for 10 min and chlorinated with sodium hypochlorite to provide an initial ratio of chlorine to total organic carbon of 1.3 and zero free-chlorine residual at 48 h. After stirring for 10 min, the concentrate was held for 72 h at room temperature. Sampling and Chemical Analyses of the Concentrate. Water samples were collected after each chlorination event immediately prior to administering to the animals and at various intervals afterward. At several points during the study, unchlorinated water concentrates were also analyzed for comparison. Comprehensive analyses were conducted to chemically characterize the unchlorinated and chlorinated concentrates; detailed results have been published.11 Briefly, a total of 106 DBPs and other chemicals were identified or measured in the chlorinated concentrate; 75 DBPs were quantified, whereas 63 were identified qualitatively by broad screen analysis using gas chromatography/mass spectrometry. Animals and Husbandry. The animals were maintained in a facility certified by the American Association for the Accreditation of Laboratory Animal Care, and procedures were approved by the Institutional Animal Care and Use Committee. Timed-pregnant 10−14-week-old Sprague−Dawley rats weighing 185−275 g were obtained from Charles River Laboratories (Raleigh, NC) on gestation day (GD) 0, housed individually, and uniquely identified with ear tags. The day that evidence of mating (copulatory plug or vaginal sperm) was detected was designated as GD 0. The animals were housed in polycarbonate cages with heat-treated pine shavings provided for bedding. Feed (PMI LabDiet Formulab Diet 5008, Richmond, IN) and drinking water were provided ad libitum. Upon receipt on GD 0, timed-pregnant females were provided reverse-osmosis-purified deionized water via amber glass bottles equipped with Teflon-lined caps and stainless steel sipper tubes. On GD 2, all animals were provided their designated water (see the Experimental Design section) via a custom-made water delivery system comprised of 6-L Teflon fluorinated ethylene propylene bags, PFA tubing, stainless steel fittings and tubing, specialized drinking valves, polystyrene coolers, and ice packs.12 Teflon-like materials, glass, and stainless steel were used to avoid known endocrine disruptors in other plastic materials. The system was headspace-free, minimized waste, and kept the water chilled and protected from light. Experimental Design. On the basis of power calculations conducted a priori,13 the study was conducted in two replicates (blocks), with each replicate consisting of 100 animals. To further improve power, the distribution of animals to the treated and control groups was 60:40 rather than 50:50. At a significance level of 0.05, this design yielded >99% statistical power for pup weight in most simulations. For prenatal loss, this analysis showed 50−60% power to detect a difference of 7.3 percentage points between groups. However, retrospective analysis to detect the same prespecified effect size19 and using arcsine square-root transformation (rather than untransformed

of environmentally realistic mixtures of DBPs bridges the gap between toxicity studies conducted on individual DBPs and epidemiological studies and takes into account possible interactions among all DBPs.10 To evaluate the potential toxicity of whole DBP mixtures at environmentally relevant doses, including unidentified DBPs, we conducted a reproductive study where water was concentrated 136-fold, chlorinated, and provided as drinking water to rats through the production of two generations of progeny. Because this was an entirely new approach to the toxicological study of drinking water, substantial efforts were made to (1) create a new method for concentrating and disinfecting whole drinking water mixtures (to retain DBPs lost by other methods), (2) maintain a water matrix suitable for in vivo and in vitro testing, (3) create a specialized water delivery system, (4) optimize experimental designs to increase our ability to detect subtle effects, and (5) improve analytical methods to measure DBPs in water concentrates.11−14 Because of the low exposure levels, the experimental design was optimized to provide enhanced statistical power relative to that which would have been provided by conventional study designs, particularly for developmental end points of pup weight and prenatal loss,13 which were of particular importance given epidemiological data. In addition to developmental and reproductive end points, we examined immunological and neurobehavioral end points. In conjunction with the current effort, we have also conducted a similar multigenerational study examining the reproductive and developmental effects of defined mixtures of the nine regulated DBPs (four trihalomethanes and five haloacetic acids) of chlorination.15 The proportions of the nine DBPs in the defined mixture were based on data from the water utility that provided the source water for the concentrate used here. An evaluation of defined mixtures provides valuable dose−response data and, in comparison with the current study, will provide insight into the effects of regulated versus unregulated water contaminants. This work is part of the “Four Lab Study”, a collaborative effort of researchers from four national laboratories/centers of the U.S. Environmental Protection Agency’s Office of Research and Development as well as extramural partners. An overview of the Four Lab Study, including its genesis and goals, has been provided.10,16,17

■

MATERIALS AND METHODS Preparation of the Water Concentrate. The concentration and chlorination of the water concentrate is described in detail by Pressman et al.11 Briefly, surface water was coagulated and clarified at a water utility, filtered via membrane ultrafiltration, and concentrated by reverse osmosis over a 19day interval. Technical-grade barium hydroxide was added to adjust the pH and remove sulfate as barium sulfate. The final pH adjustment to 6.8 was conducted with sodium carbonate. A total of 1724 L of the water concentrate was produced, homogenized, and stored at 4 °C in 16 30-gallon plastic drums lined with a 2.0-mil perfluoroalkoxy (PFA) film. The final concentration factor, based on total organic carbon, was 136fold greater than the unconcentrated filtered water. The concentrate was chlorinated under headspace-free conditions in 16 chlorination events during the course of the study. The concentrate was pumped into a 30-gallon stainless steel drum and brought to room temperature (mean ± standard deviation = 19.4 ± 0.9 °C). The drum was lined with 10654

dx.doi.org/10.1021/es402646c | Environ. Sci. Technol. 2013, 47, 10653−10659

Environmental Science & Technology

Article

were used to continue exposure for these animals. Six additional litters per group provided two females for immunological testing; for these animals, exposure was halted at weaning. A total of 10 litters per group provided one male and one female for neurobehavioral toxicity testing; exposure for these animals was also halted at weaning. Pubertal Examinations. F1 females were examined daily for vaginal opening starting on PND 25; animals were scored as closed, partially open, or fully open. Females were weighed on the day of vaginal opening, i.e., when fully open. B females were killed by decapitation on the day of vaginal opening, and sera were collected for measurement of progesterone and estradiol. F1 males were examined daily for preputial separation starting on PND 36. All males were weighed daily from PND 41 to 46 as well as on the day of preputial separation. Separation of the prepuce from the glans was scored as none, minimal, at least 50%, or complete. B and D males were killed on PND 55; body weights were recorded, sera were collected for measurement of testosterone, and testis and epididymis weights were recorded. For B males, the left testis and epididymis were placed in Davidson’s fixative and embedded in plastic for histological examination. For D males, both epididymides were used to determine sperm counts. Estrous Cycles. Over a 3-week period beginning on PND 46 or 47, the A females were monitored for estrous cyclicity. For each female, vaginal lavage was performed daily and smears were examined microscopically for vaginal cytology. Estrous cycles were classified as regular (4 or 5 days, with either 1−2 days of estrus or 2−3 days of diestrus), extended (3−4 days of estrus or 4−5 days of diestrus), or abnormal (>4 consecutive days of estrus or >6 days of diestrus).20 Breeding. Immediately following estrous cycle monitoring, the A females were paired with randomly selected nonsibling A males of the same treatment group and transferred to the male’s cage. Vaginal lavage was conducted daily during cohabitation, and smears were examined for spermatozoa. Cohabitation continued for 10 days or until evidence of mating (copulatory plug or vaginal sperm) was observed; at this time (GD 0), the female was weighed and transferred to her own cage. In Utero Insemination. Untreated females were maintained in a separate animal room with a 14/10-h light/dark cycle (lights off at 0900). Estrous cycles were synchronized with 80 μg sc of luteinizing hormone releasing hormone (LHRH) agonist (L4513-5MG; Sigma, St. Louis, MO) 115 h prior to insemination. Just after room lights turn off on the day of proestrus (4 days postinjection of the LHRH agonist), these females were paired with untreated, sexually experienced, vasectomized males for 30 min. Receptive females were inseminated by sperm obtained from C males: within 15 min of sperm diffusion from the proximal cauda epdidymis, each uterine horn was injected with a volume containing 5 × 106 sperm, a value that results in approximately 75% fertility in control males. A single female was inseminated per male; 10 females (5 per group) were inseminated per day for four consecutive days. All inseminations were performed while the recipient female was in a surgical plane of halothane anesthesia. Uterine horns were exposed through a low midventral incision. Fine, curved forceps were used to elevate and create some tension on the uterine horn, while sperm (0.1−0.2 mL) were injected through an 18-guage iv catheter attached to a 0.5-mL syringe. Injection sites were cauterized immediately upon withdrawal of the needle. Nine days later, inseminated females

data), a more appropriate model for proportional data, showed >99% power for this end point as well. Timed-pregnant animals, obtained on GD 0, were designated as parental animals, i.e., the P0 generation, and subsequent generations were designated as F1 and F2 generations (Figure S1 in the Supporting Information, SI). In block 1, P0 animals were randomly assigned to three treatment groups to receive their designated drinking water (reverse-osmosis-purified deionized water, 40 animals; chlorinated concentrate, 60 animals; unchlorinated concentrate, 40 animals). Beginning at GD 2 of the P0 generation and continuing for the remainder of the study, the designated waters were the sole source of drinking water for the animals. The group receiving unchlorinated concentrate, however, was discontinued on GD 19 because of repeated clogging of the water delivery system; this group was excluded from the study. In block 2, animals were assigned to two treatment groups (purified water, 40 animals; chlorinated concentrate, 60 animals). By design, block 2 was conducted for only the initial part of the study and was halted at postnatal day (PND) 6 of the F1 generation. Procedures. See the SI for methods regarding body weights, water consumption, parturition examinations, necropsies, histology, hormone measurements, catecholamine measurements, immunotoxicity testing, and neurobehavioral toxicity testing. Litter Examinations. F1 litters were examined on PND 0 (day of birth), 6, 13, and 21. F2 litters were examined on PND 0 and 6. On PND 0, pups were sexed, counted, weighed, and examined for evidence of nursing (i.e., abdominal milk bands). In block 1, 12 F1 litters from each group were selected randomly and the anogenital distance (AGD) of each pup was measured by an observer blinded to treatment using a dissecting microscope with an ocular micrometer (15×). The AGD was defined as the distance between the base of the genital papilla and the anal opening. On PND 6, pups were sexed, counted, and weighed; F1 litters were reduced to a maximum of eight pups (four males and four females when possible). On PND 13, each F1 pup was sexed and examined for eye opening and nipple retention. Each eye was scored as closed, partially open, or fully open. For each of the 12 normal nipple sites, areolae were scored as absent, faint, moderate, or prominent. On PND 21, F1 pups were sexed, weighed, and weaned. Selection of Weanlings. Upon weaning of the F1 generation, pups were randomly selected for different roles as the experiment continued (Table S1 in the SI). All weanlings selected to continue as part of the reproductive toxicity evaluation were uniquely identified with ear tags. One male and one female per litter (designated as “A” animals) were maintained to adulthood for breeding to produce F2 litters. An additional male and female per litter (“B” animals) were killed at the time of puberty for examination of serum hormones; males were killed on PND 55, whereas females were killed on the day of vaginal opening. In 20 randomly selected litters per group, an additional two males were selected; one male (“C”) was maintained to adulthood, and epididymal sperm were used for in utero insemination, and one male (“D”) was killed on PND 55 for determination of epididymal sperm counts. Additional F1 weanlings were randomly selected for subsequent immunotoxicity testing and neurobehavioral testing. Six litters per group each provided two males and two females for subsequent immunotoxicity testing; water bottles 10655

dx.doi.org/10.1021/es402646c | Environ. Sci. Technol. 2013, 47, 10653−10659

Environmental Science & Technology

Article

preputial separation or vaginal opening was observed, regardless of the actual day of parturition. Birth-based age was defined as the number of days since birth that these landmarks were observed. Incidences per group (e.g., estrous cycles, histopathology, pregnancy rates, breeding performance, and number of fertile breeding pairs) were analyzed with Fisher’s exact test. Water consumption was analyzed using the water bag (shared by up to five cages) as the statistical unit. For each bag and interval, daily consumption per rat was calculated based on bag weights at the beginning and end of the interval, the number of days in the interval, and the number of rats sharing the bag on each day of the interval. For neurobehavioral testing, continuous data were analyzed using ANOVA (Proc GLM in SAS) and ordinal scores were analyzed with a categorical modeling procedure (Proc Catmod in SAS) that fits linear models to functions of response frequencies, which were then analyzed by weighted regression. Because littermates (one male and one female per litter) were tested, sex was nested within litter.

were anesthetized and killed by cervical dislocation. Corpora lutea (reflecting the number of ovulations) and uterine implantation sites were counted. The fertility of each male was expressed as the percentage of implants per corpora lutea. Necropsies. Full necropsies were conducted on P0 females at 21 days postpartum (upon weaning of their litters), on F1 A males at PND 89−93, and on F1 A females at PND 95−103 (after PND-6 examinations of the F2 litters). Further information regarding necropsies, histology, and hormone measurements is presented in the SI. Sperm Measures. Sperm counts (testicular, caput epididymal, and cauda epididymal) as well as cauda epididymal sperm motility and morphology were evaluated as described previously21 in pubertal (PND 55) and adult (PND 89−93) males. In males assessed for fertility by artificial insemination (PND 95−99), SP22, a sperm membrane protein and biomarker of fertility,22 was quantified using an enzyme-linked immunosorbent assay (ELISA).23 Proximal cauda epididymal sperm membrane proteins were extracted with detergent and quantified following two-dimensional gel electrophoresis. Statistics. All statistical tests were conducted at a significance level of 0.05. No adjustments were made for multiple end points. The litter was used as the experimental unit for all developmental and reproductive data. Where appropriate, data from blocks 1 and 2 were analyzed separately as well as pooled. In the pooled analysis, the block was included in the model. Continuous data (e.g., body weights, organ weights, hormonal measures, and immunological measures), counts per litter (e.g., implantation sites and live pups), and proportions per litter (e.g., prenatal loss, implantations per corpora lutea, and males with nipples) were evaluated by analysis of variance (ANOVA; t test for two groups) using the general linear models (GLM) procedure on SAS, release 9.1 (Cary, NC). Arcsine square-root transformation was used prior to GLM analysis of proportional data. It was assumed a priori that treatment could only reduce the number of surviving progeny; therefore, one-tailed tests were used for pertinent data. Gestation lengths were analyzed using the Mann− Whitney U test; i.e., dams were ranked according to the time and stage that parturition was observed, and ANOVA was applied to the ranked data. The gestation length24 and number of live PND-0 pups were used as covariates in analyses of pup weights. Similarly, the number of implants was used as a covariate in analyses of the numbers of live pups. Litter means and frequencies per litter were used as the experimental units for analyzing pup weights and pup examination data (e.g., eye opening and nipple retention). Prenatal loss for each litter was defined as the number of implants minus the number of viable pups at the PND-0 examination. Postnatal loss was defined as the number of pups that were viable on PND 0 but not on PND 6. Perinatal loss was defined as the number of implants minus the number of live pups on PND 6. These end points were analyzed as percentages of the number of implants (prenatal and perinatal loss) or the number of live pups at PND 0 (postnatal loss). AGD was analyzed by analysis of covariance using Proc Mixed in SAS with pup weight as a covariate and litter as a random effect. To address potential bias inherent in the use of birth-based age for assessment of the onset of puberty,24 pubertal data were analyzed using conception-based age as well as age defined by the day of birth. For conception-based age, the age of pubertal onset was defined as the number of days since GD 22 that

■

RESULTS The P0 dams showed no evidence of maternal toxicity. Body weights were comparable between groups throughout gestation and lactation (Figure S2 in the SI). Dams receiving the chlorinated concentrate had significantly increased water consumption during early lactation (Figure S3 in the SI). For the F1 progeny, water consumption was comparable between groups or significantly increased in animals receiving chlorinated concentrate (Figure S5 in the SI). We detected no adverse effects on prenatal survival or birth weight (Table 1), as Table 1. Data Highlights for Prenatal Loss, Pup Weight, Onset of Puberty, and Sperm Countsa control no. of litters prenatal loss (%) pup weight at birth (g) no. of litters prenatal loss (%) pup weight at birth (g)

F1 Litters 79 5.7 ± 0.9 6.5 ± 0.1 F2 Litters 37 6.6 ± 1.0 6.1 ± 0.1 F1 Onset of Puberty 39

no. of litters males: age at puberty (days) 45.8 ± 0.3 body weight (g) 267.8 ± 3.2 females: age at puberty (days) 33.5 ± 0.3 body weight (g) 128.2 ± 2.2 F1 Adult Sperm Counts males examined 38 caput epididymis count 88.0 ± 1.8 count/g of tissue 300.4 ± 5.8

treated 118 6.7 ± 0.9 6.6 ± 0.1 55 5.9 ± 1.0 6.3 ± 0.1* 57 46.2 ± 0.2 267.2 ± 2.5 34.3 ± 0.3* 132.5 ± 2.1 57 81.8 ± 2.0* 276.5 ± 7.3*

Data represent mean ± SEM of litter percentages for prenatal loss; litter averages for pup weight on PND 0, age of onset of puberty (preputial separation in males and vaginal opening in females), and body weight on the day of puberty onset; and sperm counts determined in one male per litter. Ages are expressed as days after GD 22. Asterisks indicate significant differences from control (p < 0.05). a

10656

dx.doi.org/10.1021/es402646c | Environ. Sci. Technol. 2013, 47, 10653−10659

Environmental Science & Technology

Article

neous abortion in humans, also had enhanced statistical power, yet there was clearly no adverse effect on this end point. We found several significant, albeit slight, effects associated with treatment: delayed puberty in females, reduced sperm counts in adult males, and thyroid follicular cell hypertrophy in adult females. Several DBPs (dibromoacetic acid, bromochloroacetic acid, and bromodichloromethane), when tested individually, have been shown to delay puberty in rats.26−29 Also, the present data are consistent with the dose response for pubertal delays caused by defined mixtures of nine regulated DBPs (four trihalomethanes and five haloacetic acids).15 The proportions of the nine DBPs in the defined mixture were based on data from the water utility that provided the source water for the concentrate. The consistency of the pubertal data in the two studies suggests that the regulated trihalomethanes and haloacetic acids may be contributors to the effect observed with the whole mixture. A decrease in the caput sperm counts without decreases in the testicular and cauda epididymal sperm indicates accelerated transit through the proximal epididymis often associated with a deficit in sperm maturation and reduced fertility.30 Two DBPs, dibromoacetic acid and bromochloroacetic acid, individually and in combination, have been shown to reduce sperm quality in rats23,27 and rabbits.31 Although fertility was unaffected in the current males (Table S5 in the SI), this was unsurprising given that sperm quality in rats, unlike humans, must be substantially reduced to impact fertility.32 These findings highlight the need for further research on DBPs and male fertility. Follicular cell hypertrophy may reflect a hypothyroid state, which can lead to hyperplasia and then to follicular cell carcinogenesis.33 Several drinking water contaminants have been reported to cause such thyroid effects in rodents; these include inorganics (e.g., iodide and bromide),34,35 DBPs [e.g., chlorate, bromate, and 3-chloro-4-(dichloromethyl)-5-hydroxy2(5H)-furanone (MX)],36−38 and other compounds unrelated to chemical disinfection (e.g., perchlorate).39,40 On the basis of chemical analyses of the chlorinated concentrate11 and toxicity reports, it is likely that chlorate,37 perhaps with perchlorate, MX,38 or unidentified contaminants contributed to this thyroid effect. Overall, it is reassuring that multigenerational reproductive and developmental toxicity testing of an environmentally relevant whole mixture of drinking water DBPs yielded predominantly negative results. Nonetheless, slight but significant effects on puberty, sperm production, and thyroid cells warrant further investigation. Because the composition of DBP mixtures is highly variable, the toxicological testing of other concentrated DBP mixtures, including those generated by other disinfectants, also merits additional evaluation.

well as postnatal survival and growth, pregnancy maintenance, gestation length, pup morphology, eye opening, nipple retention, male puberty, estrous cycles, mating, fertility, or immune response (Tables S2−S6 and S9 and Figure S4 in the SI). F2 pup weights at birth, but not PND 6 (Table S5 in the SI), were significantly increased compared to controls. Prenatal loss (Table 1), analogous to spontaneous abortion in humans, was also clearly negative, with values well within the range of historical controls. Findings of uncertain toxicological importance were an increase in F1 male AGD (Table S2 in the SI), slight, sex-specific changes in lung, liver, and pituitary organ weights (Table S7 in the SI), neurobehavioral end points (forelimb grip strength and click response) that were nonsignificant upon step-down analysis (Table S11 in the SI), and dental malocclusion in all weanlings of an F1 litter. Highly significant increases (7−10%) in relative kidney weights for F1 adults of both sexes (Table S7 in the SI) likely reflected the higher solute levels in the concentrate. Vaginal opening, a marker for the onset of puberty, was slightly, but significantly, delayed 0.8 days in F1 females (Tables 1 and S4 in the SI). Serum levels of progesterone and estradiol on the day of vaginal opening, however, were unaffected. Although sperm counts were comparable between groups at puberty, significant reductions (7−8%) were observed in adult, fully mature F1 treated males in the caput epididymis (Tables 1 and S8 in the SI). Cauda epididymal sperm counts were unaffected, and no pathology was observed in the testis or epididymis. Thyroid follicular cell hypertrophy was observed in P0 and F1 females with significantly increased incidences and severity in the treated group compared to controls; the incidence of mild or moderate hypertrophy was also significantly increased in both the P0 and F1 females (Figure 1).

Figure 1. Incidence of thyroid follicular cell hypertrophy significantly increased (p < 0.05) in both P0 and F1 females exposed to the water concentrate. Post hoc analysis indicated the incidence of mild or moderate hypertrophy was also significantly increased (p < 0.05) in both generations. (Because of a technical error, male thyroids were not examined.)

■

ASSOCIATED CONTENT

S Supporting Information *

More background information on the Four Lab Study and supplemental methodological information, tables, and figures. This material is available free of charge via the Internet at http://pubs.acs.org.

■

DISCUSSION The lack of an adverse effect on pup weight, comparable to low birth weight in humans, is especially noteworthy because it is a sensitive indicator of developmental toxicity.25 Unusually large sample sizes provided extraordinary statistical power to detect subtle changes in pup weight;13 nonetheless, there were clearly no reductions. Similarly, prenatal loss, analogous to sponta-

■

AUTHOR INFORMATION

Corresponding Author

*Phone: 919-541-0591. Fax: 919-541-4017. E-mail: narotsky. [email protected]. 10657

dx.doi.org/10.1021/es402646c | Environ. Sci. Technol. 2013, 47, 10653−10659

Environmental Science & Technology

Article

Notes

administering volatile chemicals while minimizing chemical waste in rodent toxicity studies. Lab. Anim. 2010, 44 (1), 66−68. (13) Dingus, C. A.; Teuschler, L. K.; Rice, G. E.; Simmons, J. E.; Narotsky, M. G. Prospective power calculations for the Four Lab Study of a multigenerational reproductive/developmental toxicity rodent bioassay using a complex mixture of disinfection by-products in the low-response region. Int. J. Environ. Res. Public Health 2011, 8, 4082−4101. (14) Narotsky, M. G.; Pressman, J. G.; Miltner, R. J.; Speth, T. F.; Teuschler, L. K.; Rice, G. E.; Richardson, S. D.; Best, D. S.; McDonald, A.; Hunter, E. S., 3rd; Simmons, J. E. Developmental toxicity evaluations of whole mixtures of disinfection by-products using concentrated drinking water in rats: gestational and lactational effects of sulfate and sodium. Birth Defects Res., Part B 2012, 95 (3), 202−212. (15) Narotsky, M. G.; Klinefelter, G. R.; Goldman, J. M.; Strader, L. F.; Suarez, J. D.; Pressman, J. G.; Miltner, R. J.; Speth, T. F.; Teuschler, L. K.; Rice, G. E.; Richardson, S. D.; Best, D. S.; McDonald, A.; Murr, A. S.; Hunter, E. S.; Simmons, J. E. Assessment of reproductive and developmental toxicity of mixtures of regulated drinking water chlorination by-products in a multigenerational rat bioassay. Birth Defects Res., Part A 2010, 88 (5), 360. (16) Simmons, J. E.; Richardson, S. D.; Teuschler, L. K.; Miltner, R. J.; Speth, T. F.; Schenck, K. M.; Hunter, E. S., 3rd; Rice, G. Research issues underlying the four-lab study: integrated disinfection byproducts mixtures research. J. Toxicol. Environ. Health, Part A 2008, 71 (17), 1125−1132. (17) Simmons, J. E.; Teuschler, L. K.; Gennings, C.; Speth, T. F.; Richardson, S. D.; Miltner, R. J.; Narotsky, M. G.; Schenck, K. D.; Hunter, E. S., 3rd; Hertzberg, R. C.; Rice, G. Component-based and whole-mixture techniques for addressing the toxicity of drinking-water disinfection by-product mixtures. J. Toxicol. Environ. Health, Part A 2004, 67 (8−10), 741−754. (18) McGuire, M. J.; McLain, J. L.; Obolensky, A. Information Collection Rule Data Analysis; AWWA Research Foundation: Denver, CO, 2002. (19) Thomas, L. Retrospective power analysis. Conserv. Biol. 1997, 11 (1), 276−280. (20) Goldman, J. M.; Murr, A. S.; Cooper, R. L. The rodent estrous cycle: characterization of vaginal cytology and its utility in toxicological studies. Birth Defects Res., Part B 2007, 80 (2), 84−97. (21) Klinefelter, G. R.; Strader, L. F.; Suarez, J. D.; Roberts, N. L. Bromochloroacetic acid exerts qualitative effects on rat sperm: implications for a novel biomarker. Toxicol. Sci. 2002, 68 (1), 164− 173. (22) Klinefelter, G. R. Saga of a sperm fertility biomarker. Anim. Reprod. Sci. 2008, 105 (1−2), 90−103. (23) Kaydos, E. H.; Suarez, J. D.; Roberts, N. L.; Bobseine, K.; Zucker, R.; Laskey, J.; Klinefelter, G. R. Haloacid induced alterations in fertility and the sperm biomarker SP22 in the rat are additive: validation of an ELISA. Toxicol. Sci. 2004, 81 (2), 430−442. (24) Narotsky, M. G. Comparison of birth- and conception-based definitions of postnatal age in developmental and reproductive rodent toxicity studies: influence of gestation length on measurements of offspring body weight and puberty in controls. Birth Defects Res., Part A 2011, 91 (5), 376. (25) Chernoff, N.; Rogers, E. H.; Gage, M. I.; Francis, B. M. The relationship of maternal and fetal toxicity in developmental toxicology bioassays with notes on the biological significance of the “no observed adverse effect level”. Reprod. Toxicol. 2008, 25 (2), 192−202. (26) Sloan, C. S.; Klinefelter, G. R.; Goldman, J. M.; Vick, K. D.; Fail, P. A.; Tyl, R. W. Delayed preputial separation (PPS) and SP22 measurement in rats administered bromochloroacetic acid (BCA) in drinking water. 44th Annual Meeting of the Society of Toxicology, New Orleans, LA, 2005; Vol. 84, p 112. (27) Klinefelter, G. R.; Strader, L. F.; Suarez, J. D.; Roberts, N. L.; Goldman, J. M.; Murr, A. S. Continuous exposure to dibromoacetic acid delays pubertal development and compromises sperm quality in the rat. Toxicol. Sci. 2004, 81 (2), 419−429.

The authors declare no competing financial interest.

■

ACKNOWLEDGMENTS We thank Maria Hoopes, Talitha Peay, and Adina Wiggins for their tremendous technical contributions to this effort. We are also grateful to Dr. Tony DeAngelo and Michael George for their contributions with necropsies, Pam Phillips for neurobehavioral testing, Carey Copeland and Dr. Jamie DeWitt for immunotoxicity testing, Drs. Glen Marrs and John Seeley for pathological examinations, and Dr. Shahid Parvez for discussion of the onset-of-puberty data. Finally, we are grateful to the drinking water utility for the use of their facilities and acknowledge their request to remain anonymous. The information in this document has been funded wholly by the U.S. Environmental Protection Agency. It has been subjected to review by the National Health and Environmental Effects Research Laboratory and approved for publication. Approval does not signify that the contents reflect the views of the Agency, nor does mention of trade name or commercial products constitute endorsement or recommendation for use.

■

REFERENCES

(1) Aschengrau, A.; Zierler, S.; Cohen, A. Quality of community drinking water and the occurrence of late adverse pregnancy outcomes. Arch. Environ. Health 1993, 48 (2), 105−113. (2) Bove, F.; Shim, Y.; Zeitz, P. Drinking water contaminants and adverse pregnancy outcomes: a review. Environ. Health Perspect. 2002, 110 (Suppl1), 61−74. (3) Dodds, L.; King, W.; Allen, A. C.; Armson, B. A.; Fell, D. B.; Nimrod, C. Trihalomethanes in public water supplies and risk of stillbirth. Epidemiology 2004, 15 (2), 179−186. (4) Hwang, B. F.; Jaakkola, J. J. Water chlorination and birth defects: a systematic review and meta-analysis. Arch. Environ. Health 2003, 58 (2), 83−91. (5) Infante-Rivard, C. Drinking water contaminants, gene polymorphisms, and fetal growth. Environ. Health Perspect. 2004, 112 (11), 1213−1216. (6) Waller, K.; Swan, S. H.; DeLorenze, G.; Hopkins, B. Trihalomethanes in drinking water and spontaneous abortion. Epidemiology 1998, 9 (2), 134−140. (7) Nieuwenhuijsen, M. J.; Grellier, J.; Smith, R.; Iszatt, N.; Bennett, J.; Best, N.; Toledano, M. The epidemiology and possible mechanisms of disinfection by-products in drinking water. Philos. Trans. R. Soc., A 2009, 367 (1904), 4043−4076. (8) Hrudey, S. E. Chlorination disinfection by-products, public health risk tradeoffs and me. Water Res. 2009, 43 (8), 2057−2092. (9) Richardson, S. D.; Plewa, M. J.; Wagner, E. D.; Schoeny, R.; DeMarini, D. M. Occurrence, genotoxicity, and carcinogenicity of regulated and emerging disinfection by-products in drinking water: a review and roadmap for research. Mutat. Res. 2007, 636 (1−3), 178− 242. (10) Simmons, J. E.; Richardson, S. D.; Speth, T. F.; Miltner, R. J.; Rice, G.; Schenck, K. M.; Hunter, E. S., 3rd; Teuschler, L. K. Development of a research strategy for integrated technology-based toxicological and chemical evaluation of complex mixtures of drinking water disinfection byproducts. Environ. Health Perspect. 2002, 110 (Suppl 6), 1013−1024. (11) Pressman, J. G.; Richardson, S. D.; Speth, T. F.; Miltner, R. J.; Narotsky, M. G.; Hunter, E. S., 3rd; Rice, G. E.; Teuschler, L. K.; McDonald, A.; Parvez, S.; Krasner, S. W.; Weinberg, H. S.; McKague, A. B.; Parrett, C. J.; Bodin, N.; Chinn, R.; Lee, C. F.; Simmons, J. E. Concentration, chlorination, and chemical analysis of drinking water for disinfection byproduct mixtures health effects research: U.S. EPA’s Four Lab Study. Environ. Sci. Technol. 2010, 44 (19), 7184−7192. (12) McDonald, A.; Killough, P.; Puckett, E.; Best, D. S.; Simmons, J. E.; Pressman, J. G.; Narotsky, M. G. A novel water delivery system for 10658

dx.doi.org/10.1021/es402646c | Environ. Sci. Technol. 2013, 47, 10653−10659

Environmental Science & Technology

Article

(28) Christian, M. S.; York, R. G.; Hoberman, A. M.; Fisher, L. C.; Brown, W. R. Oral (drinking water) two-generation reproductive toxicity study of bromodichloromethane (BDCM) in rats. Int. J. Toxicol. 2002, 21 (2), 115−146. (29) Christian, M. S.; York, R. G.; Hoberman, A. M.; Frazee, J.; Fisher, L. C.; Brown, W. R.; Creasy, D. M. Oral (drinking water) twogeneration reproductive toxicity study of dibromoacetic acid (DBA) in rats. Int. J. Toxicol. 2002, 21 (4), 237−276. (30) Klinefelter, G. R.; Suarez, J. D. Toxicant-induced acceleration of epididymal sperm transit: androgen-dependent proteins may be involved. Reprod. Toxicol. 1997, 11 (4), 511−519. (31) Veeramachaneni, D. N.; Palmer, J. S.; Klinefelter, G. R. Chronic exposure to low levels of dibromoacetic acid, a water disinfection byproduct, adversely affects reproductive function in male rabbits. J. Androl. 2007, 28 (4), 565−577. (32) Klinefelter, G. R.; Laskey, J. W.; Perreault, S. D.; Ferrell, J.; Jeffay, S.; Suarez, J.; Roberts, N. The ethane dimethanesulfonateinduced decrease in the fertilizing ability of cauda epididymal sperm is independent of the testis. J. Androl. 1994, 15 (4), 318−327. (33) Hill, R. N.; Erdreich, L. S.; Paynter, O. E.; Roberts, P. A.; Rosenthal, S. L.; Wilkinson, C. F. Thyroid follicular cell carcinogenesis. Fundam. Appl. Toxicol. 1989, 12 (4), 629−697. (34) Pavelka, S.; Babicky, A.; Vobecky, M.; Lener, J. High bromide intake affects the accumulation of iodide in the rat thyroid and skin. Biol. Trace Elem. Res. 2001, 82 (1−3), 133−142. (35) Collins, W. T.; Capen, C. C. Ultrastructural and functional alterations of the rat thyroid gland produced by polychlorinated biphenyls compared with iodide excess and deficiency, and thyrotropin and thyroxine administration. Virchows Arch. B 1980, 33 (3), 213−231. (36) DeAngelo, A. B.; George, M. H.; Kilburn, S. R.; Moore, T. M.; Wolf, D. C. Carcinogenicity of potassium bromate administered in the drinking water to male B6C3F1 mice and F344/N rats. Toxicol. Pathol. 1998, 26 (5), 587−594. (37) Hooth, M. J.; DeAngelo, A. B.; George, M. H.; Gaillard, E. T.; Travlos, G. S.; Boorman, G. A.; Wolf, D. C. Subchronic sodium chlorate exposure in drinking water results in a concentrationdependent increase in rat thyroid follicular cell hyperplasia. Toxicol. Pathol. 2001, 29 (2), 250−259. (38) Komulainen, H.; Kosma, V. M.; Vaittinen, S. L.; Vartiainen, T.; Kaliste-Korhonen, E.; Lotjonen, S.; Tuominen, R. K.; Tuomisto, J. Carcinogenicity of the drinking water mutagen 3-chloro-4-(dichloromethyl)-5-hydroxy-2(5H)-furanone in the rat. J. Natl. Cancer Inst. 1997, 89 (12), 848−856. (39) Srinivasan, A.; Viraraghavan, T. Perchlorate: health effects and technologies for its removal from water resources. Int. J. Environ. Res. Public Health 2009, 6 (4), 1418−1442. (40) Lewandowski, T. A.; Seeley, M. R.; Beck, B. D. Interspecies differences in susceptibility to perturbation of thyroid homeostasis: a case study with perchlorate. Regul. Toxicol. Pharmacol. 2004, 39 (3), 348−362.

10659

dx.doi.org/10.1021/es402646c | Environ. Sci. Technol. 2013, 47, 10653−10659