Developmental Atrazine Exposure Suppresses Immune Function in Male, but not Female Sprague-Dawley Rats

Andrew A. Rooney*,1, Raymond A. Matulka{dagger} and Robert W. Luebke{ddagger}

* College of Veterinary Medicine, Anatomy, Physiological Sciences and Radiology, North Carolina State University, Raleigh, North Carolina 27695; {dagger} Curriculum in Toxicology, University of North Carolina, Research Triangle Park, North Carolina 27710; and {ddagger} Immunotoxicology Branch, Experimental Toxicology Division, National Health and Environmental Effects Research Laboratory, U. S. Environmental Protection Agency, Research Triangle Park, North Carolina 27711

Received August 8, 2003; accepted September 12, 2003


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Each year, 75 million pounds of the broadleaf herbicide atrazine (ATR) are applied to crops in the United States. Despite limited solubility, ATR is common in ground and surface water, making it of regulatory concern. ATR suppresses the immunomodulatory hormones prolactin (PRL) and the thyroid hormones (THs), with developmental exposure to ATR permanently disrupting PRL regulation. We hypothesized that ATR may cause developmental immunotoxicity through its disruption of PRL or THs. To test this hypothesis, pregnant Sprague-Dawley (SD) rats were exposed to 35-mg ATR/kg/d from gestational day (GD) 10 through postnatal day (PND) 23. Separate groups were exposed to bromocryptine (BCR) at 0.2 mg/kg/2x/day to induce hypoprolactinemia or to propylthiouracil (PTU) at 2 mg/kg/day to induce hypothyroidism. After the offspring reached immunologic maturity (at least 7 weeks old), the following immune functions were evaluated: natural killer (NK) cell function; delayed-type hypersensitivity (DTH) responses; phagocytic activity of peritoneal macrophages; and antibody response to sheep erythrocytes (SRBC). ATR decreased the primary antibody and DTH responses in male offspring only. Neither PTU nor BCR caused immunosuppression in any measured variable, although PTU increased phagocytosis by peritoneal macrophages. These results demonstrate that developmental exposure to ATR produced gender-specific changes in immune function in adult rats and suggest that immune changes associated with ATR are not mediated through the suppression of PRL or THs.

Key Words: atrazine; developmental immunotoxicity; prolactin; thyroid hormones; rats; pesticides.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The ability of atrazine (ATR) to control broadleaf weeds in the production of corn, sorghum, and sugar cane make it among the most widely used pesticides in the world, with approximately 75 million pounds spread annually in the United States alone (Short and Colborn, 1999Go; U.S. EPA, 2001Go). ATR is moderately persistent under normal soil conditions and has low to moderate solubility in water; however, breakdown of ATR is negligible in neutral or low-alkaline water resulting in a half-life of 2 or more years (Cohen et al., 1984Go). ATR is the most commonly detected pesticide in both ground and surface water in the United States, with spring-associated spikes that sometimes exceed the Environmental Protection Agency’s (EPA’s) maximum contaminant level (MCL) (U.S. EPA, 1991Go). The prevalence of ATR in source water explains why drinking water is the most common route of human exposure to ATR (U.S. EPA, 2001Go) and why ATR is included on the EPA’s candidate contaminant list for health effects testing. Although management practices improved during the 1990s and reduced the likelihood of ATR entering ground and surface water, there has been no downward trend in ATR detected in bodies of water draining areas of high ATR use such as the Mississippi drainage into the Gulf of Mexico (Clark and Goolsby, 2000Go).

Animal data indicate that maternal exposure to ATR reduces the suckling-induced rise in prolactin (PRL) and thereby decreases PRL concentrations in the milk during a critical ontogenetic set-period for hypothalamic regulation of pituitary PRL secretion (Stoker et al., 1999aGo). Decreased PRL during this critical stage (postnatal days (PNDs 2–5)) hinders adult regulation of PRL production and may result in permanent hyperprolactinemia in adults (Stoker et al., 1999aGo). In contrast, experimentally induced decreases in PRL are temporary in rats of any other age and do not effect the regulation of PRL after the experimental agent is removed (Shyr et al., 1986Go). PRL is a known immunomodulator that up-regulates immune function at physiologic concentrations and down-regulates function at supraphysiologic doses (Gala, 1991Go). Increased release of PRL has been reported in adult mice exposed to diethylstilbestrol (DES) for the first 5 days of life, an exposure regimen that causes long-term suppression of natural killer (NK) cell activity (Kalland, 1984Go), antibody responses (Kalland, 1980Go), and the delayed-type hypersensitivity (DTH) response (Kalland and Forsberg, 1978Go).

In a mixture study, ATR (10 parts per billion (ppb) or approximately 2.0 ng/kg/day) in combination with aldicarb (10 ppb) and nitrate (28 parts per million (ppm)) in drinking water suppressed thyroid hormone (TH) concentrations (Porter et al., 1999Go). High doses of ATR alone (618 mg/kg/day for 6 days) caused a dose-dependent decrease in serum triidothyronine in adult female Wistar rats (Kornilovskaya et al., 1996Go). Although thyroid–immune interactions are not as well studied as PRL immunomodulation, there are several lines of evidence suggesting that THs are also immunomodulatory (reviewed in Fabris et al., 1995Go).

Despite evidence of ATR’s endocrine toxicity in adults and neonates, there is little evidence of immunotoxicity associated with adult exposure, and, to our knowledge, ATR has not been examined for developmental immunotoxicity. Fournier (1992)Go concluded that female C57/BL6 mice exposed to a single high dose (27.3 to 875 mg/kg) of ATR displayed no cell-mediated or humoral immune changes because their plaque-forming cell (PFC) assay response and mixed leukocyte response (MLR) displayed no dose–response relationships. Exposure of female B6C3F1 mice to medium and high doses of ATR (25 or 250 mg/kg/day) for 14 days did not affect NK cell function, mitogen-induced lymphocyte proliferation, MLR, antibody response to sheep erythrocytes (SRBC), or lymphocyte cell surface markers (Munson et al., 1994Go). In a 3-week oral exposure to ATR (100 to 900 mg/kg) in adult male Wistar rats, the highest dose (900 mg/kg) was associated with decreased body weight, and ATR did not affect serum IgM or IgG concentrations (Vos et al., 1983Go), although antigen-specific antibody production was not measured. High doses (500 mg/kg) of ATR decreased the double positive CD4CD8 thymocytes and increased the MLR of female B6C3F1 mice; however, body weight was also decreased at 500 mg/kg (Munson et al., 1994Go), suggesting the general toxicity of ATR at this dose.

Permanent hypothalamic modification of PRL regulation following developmental exposure to ATR, and the low order of ATR toxicity for adult immune function, suggests that ATR may be a developmental immunotoxicant if exposure occurs during a critical developmental window. Alternatively, because ATR has been shown to decrease the production of THs, ATR might also affect normal immune system development via a thyroid pathway. The current studies were designed to test the hypothesis that developmental ATR exposure causes immunotoxicity, and that immunotoxicity is linked to the suppression of either hormone (i.e., PRL or THs) during a critical, early stage of immune development. Exposure was timed to include the second half of gestation and all of lactation, the period of organogenesis and early development of immune function in rodents. Antibody responses to SRBC and DTH responses to BSA were monitored to determine the effect of ATR on adaptive immune responses. Assays of NK cell activity and macrophage phagocytosis were also evaluated as measures of innate immune function.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Chemicals
Atrazine (2-chloro-4-ethylamino-6-isopropylamino-S-triazine; lot 254–140B for the first experiment and lot 273–52A for the repeat experiment) was obtained from ChemServices (West Chester, PA). Bromocryptine (2-bromo-a-ergocryptine; Sigma, St. Louis, MO), a dopamine receptor agonist known to reduce PRL secretion (Nagy et al., 1983Go, 1991Go), was used to induce hypoprolactinemia. The goitrogen propylthiouracil (6-propyl-2-thiouracil; Sigma) was administered to induce hypothyroidism.

Animals
The experiment was performed twice with two separate groups of timed pregnant Sprague-Dawley (SD) rats purchased from Charles River Laboratories (Raleigh, NC). The rats were housed individually in polycarbonate cages, provided a 12-h light (0600 h) to dark (1800 h) cycle, maintained at 20 ± 1°C, and given access to both food and water ad libitum. The dams arrived at the EPA on gestational day (GD) 2 using the definition of GD 1 equal to the day following overnight mating. All procedures employed in this study were approved in advance by the Institutional Animal Care and Use Committee of the EPA.

Dose Selection
The dose of ATR was chosen as the lowest quantity likely to produce significant endocrine modification, as evident by alterations in reproductive development in both sexes. The no observable effects level (NOEL) is 6.25-mg/kg ATR given twice daily for suppression of the suckling-induced rise in serum PRL in rat dams (Stoker et al., 1999aGo), and the NOEL for ATR-associated delay in prepubertal separation is 6.25 mg/kg/day for male rat offspring (Stoker et al., 2000Go). However, the NOEL for the ATR-associated delay in the onset of puberty in female rats is 25 mg/kg/day (Laws et al., 2000Go). Therefore, 35 mg/kg/day was selected as a dose just above the NOEL for the delay of puberty in females and therefore likely to induce endocrine effects in both sexes. Above 25 mg/kg/day, ATR given to nursing dams during PNDs 1–4 results in developmental induction of prostatitis in the pups after sexual maturity, which is further evidence that ATR at 35 mg/kg/day would have the desired endocrine modifying effects, including reducing PRL (Stoker et al., 1999bGo). Both ATR and BCR increase the incidence of prostatitis, which is induced in Wistar rats by PRL disruption in the critical developmental window of PNDs 1–9 (Stoker et al., 1999bGo). The PRL blocking effects of BCR given to the dam are well characterized and result in the suppression of PRL in the dam’s circulation as well as in her milk, which, in turn, causes the offspring to have abnormal PRL regulation and eventually leads to hyperprolactinemia as they mature (Shyr et al., 1986Go). For this study, BCR was given at the lowest observable adverse effects level (LOAEL) of 0.2 mg/kg given twice a day for developmental induction of prostatitis (Stoker et al., 1999bGo). The TH blocking effects of PTU given to the dam are also well characterized, and PTU leads to small, hypothyroid offspring that slowly obtain normal TH concentrations by adulthood (Kirby et al., 1992Go). The dose of PTU, 2 mg/kg/day, was selected as a dose that would cause hypothyroidism without being overtly toxic to either dams or pups (Crofton, personal communication) and was verified for efficacy by testing TH concentrations in the offspring on PND 14.

Experimental Design
Chemical exposure was initiated on GD 10, to avoid the possible interference of ATR with embryo implantation (Cummings et al., 2000Go). Pregnant SD dams were weighed daily during dosing, and dosing was done between 0700 and 1000 h in a volume of 0.1 ml/100 g body weight. Each of four separate groups of dams were subjected to a treatment regime from GD 10 through weaning of pups on postnatal day (PND) 23: (1) ATR (35 mg/kg/day) in 1% methylcellulose via gavage; (2) 1% methylcellulose via gavage for control dams; (3) BCR (0.2 mg/kg twice a day) in saline via subcutaneous (sc) injection at the nape of the neck; or (4) PTU (2 mg/kg/day) in saline sc. The second dose of BCR was given at approximately 1600 h. The pups were not treated directly and therefore were only exposed by transplacental transfer (ATR crosses placenta; Ugazio et al., 1991Go) or via lactation (ATR or its metabolites are present in milk of dams treated with ATR; Stoker, personal communication).

On PND 2, all pups from the dams of a given treatment regime were removed from their mothers, weighed, pooled, and reassigned (five each, males and females) to dams; no attempt was made to keep littermates together or apart or to return pups to their biological mother. This was done to (1) eliminate the effects of unequal litter size or sex ratio on development; (2) reduce the effect of variation in a dam’s response to treatment (e.g., twice daily doses of 25- or 12.5-mg/kg ATR inhibited suckling-induced PRL release in some dams and had no discernible effect in others; Stoker et al., 1999aGo); and (3) utilize dams with litters of fewer than five each of male and female pups, by the addition of pups from other litters. Excess pups were euthanized by decapitation. At PND 7, the pups were ear-punched for identification and tracked individually for the remainder of the experiment (prior to PND 7, the ears were too small to allow ear-punching). On PND 14, one arbitrarily selected pup of each sex was euthanized with CO2, a blood sample was taken via cardiac puncture, and the spleen and thymus were removed and weighed. Serum was collected and assayed for TH concentrations via ELISA to determine the efficacy of the TH antagonist (PTU). If more than four pups from a given dam died by PND 14, all pups from that litter were excluded from the experiment. On PND 23, the dams were euthanized with CO2, and the pups were weaned. After weaning, the offspring were housed three per cage, segregated by gender and treatment. The offspring were tested for immune function and innate immunity after reaching full immunological maturity (PND 49 or later).

Thyroxine (T4) and Triidothyronine (T3) ELISAs
Total T4 and T3 concentrations were measured using commercial kits (Endocrinetech, Newark, CA) designed for use with rat serum. Serum was added to duplicate wells of a microtiter plate and assayed according to the manufacturer’s instructions. The absorbance of each well was determined with a SpectraMax 250-plate reader (Molecular Devices, Sunnyvale, CA) at 450 nm. The detection limit of the ELISAs was 2 ng/ml for T4 and 0.2 ng/ml for T3. Using SOFTmax PRO software (Molecular Devices), the average absorbance reading of the replicate wells was transformed into a hormone concentration by interpolation from the standard curve.

Evaluation of Immune Function
In the first experiment, 7- to 9-week-old offspring from each treatment group and sex were divided into four subsets (n = 8 or 9/treatment/sex) for evaluation of specific and innate immunity. In the second, repeat, experiment three additional subsets of ATR and control offspring were added to the four groups used to measure immunity in 7- to 9-week-old offspring. These three additional subsets were used to evaluate the persistence of immunosuppression found in the first experiment. Therefore, in the repeat experiment, subsets were included to evaluate the DTH response in 12-week-old rats, the DTH response in 6-month-old rats, and the antibody response in 6-month-old rats.

Antibody response.
In each experiment, one subset of offspring was used to measure the primary (IgM) and secondary (IgG) T-dependent antibody responses. Eight-week-old rats immunized by intravenous injection of 2.5 x 108 SRBC were bled from the tail after 5 days, and the serum collected from the coagulated blood was frozen for later analysis of SRBC-specific IgM as described by Smialowicz et al.(2001)Go. Four weeks after the primary immunization, the secondary response to SRBC was evaluated by repeating the immunization and sampling process in the same animals. The serum concentrations of SRBC-specific IgM and IgG antibodies were measured by ELISA. The wells were coated with solubilized SRBC membranes prepared as described by Temple et al.(1993)Go. An internal standard (a serum pool from 20 immunized rats) was evaluated on every microtiter plate, and all ELISA procedures were optimized for differences in SRBC membrane preparations and each new lot of secondary antibody.

IgM assay.
Flat-bottom 96-well Immunolon-2 ELISA microtiter plates (Dynatech Labs, Chantilly, VA) were coated with 125-µl SRBC membrane (0.0015 mg/ml) in PBS by incubation at 4°C for at least 16 h. Each plate included 16 wells coated with 100 µl of goat anti-rat IgM (0.01 mg/ml, mchain specific; Bethyl Laboratories, Inc., Montgomery, TX) in 10.05-M Na2CO3 (Sigma) rather than SRBC membrane, to serve as a standard curve. After coating, the plates were washed with 0.05% TWEEN 20 (Sigma) in distilled water (dH2O). Nonspecific binding in all wells was blocked by adding 4% powdered milk in dH2O and incubating for 1 h at room temperature, followed by another wash. One hundred µl of serum samples were added in duplicate to SRBC membrane–coated wells at an initial concentration of 1:64 in diluent (50-mM tris; Sigma, 0.14-M NaCl, 0.05% TWEEN 20, 1% BSA; Sigma), and 100 µl of rat IgM (2000 ng/ml in diluent; Bethyl Laboratories, Inc.) were placed in the standard curve portion of the plate. Initial sample and standard curve wells were serially diluted 1:2 and the plates were incubated 1 h at room temperature, followed by washing three times with 0.05% TWEEN 20 in dH2O. One hundred µl of secondary antibody (0.0003-mg/ml goat anti-rat IgM horseradish peroxidase conjugate in 4% powdered milk in dH2O; Accurate Chemical & Scientific Corp., Westbury, NY) were added to all wells and the plates were incubated for 1 h at room temperature, followed by washing three times with 0.05% TWEEN 20 in dH2O. As the final, color-generating step, peroxidase substrate, 2,2' azino-bis (3-ethylbenzthiazoline-6-sulfonic acid) diammonium salt (Sigma) was added to all wells. The plates were incubated at room temperature for a maximum of 2 h.

IgG assay.
The concentration of SRBC-specific IgG was evaluated using the same quantitative assay as that for IgM with the following exceptions: The standard curve wells were coated with 100 µl of rabbit anti-rat IgG Fc (0.01 mg/ml; Bethyl) in 10.05-M Na2CO3, nonspecific binding was blocked with 50-mM tris/0.14 M NaCl containing 1% BSA in place of 4% milk, and the standard curve for IgG was generated by serial (1:2) dilution of rat IgG (Bethyl), beginning with 100 µl at a concentration of 1000 ng IgG/ml); 100 µl of rabbit anti-rat IgG horseradish peroxidase conjugate (0.00016-mg/ml Sigma) in 50-mM tris/0.14 M NaCl, containing 1% BSA were used as the secondary antibody.

IgM and IgG data collection and analysis.
Absorbance was read on a SpectraMax 250-plate reader at 410 nm and processed using SOFTmax PRO software. The dilution curves of each serum sample were examined for portions of the curve where absorbance values of successive dilutions were closest to halve the value of the previous dilution. The average of these two absorbance readings was then used to determine the IgM or IgG concentration from the standard curve.

DTH response.
The second subgroup of rats was used for assaying the DTH response to purified (fraction V) bovine serum albumin (BSA; Sigma) by injecting 0.1-ml BSA (1-mg/ml PBS) emulsified in Freunds complete adjuvant (FCA; Difco, Detroit, MI) at a ratio of 1:1 by sc injection in the caudal tail fold of 8-week-old rats anesthetized with inhaled isoflurane. Seven days later, heat-aggregated BSA was prepared by heating BSA (10 mg/ml) in saline to 75°C for 1 h and centrifuging the resulting BSA gel at 450 RCF to remove excess saline. The rats were anesthetized with isoflurane, and heat-aggregated BSA was injected into the right rear footpad and normal saline into the left rear footpad. After 24 h, footpad thickness was measured with Electronic Thickness Calipers (Battelle’s Environmental Technology Commercialization Center, Research Triangle Park, NC). Swelling was calculated by subtracting the saline-injected, left footpad thickness from the BSA-injected right footpad thickness. The swelling was then plotted as a percentage of left foot thickness.

NK cell activity.
The third subgroup was used to determine NK cell activity in 9-week-old rats. While removing the spleen for the NK assay, the thymus was also removed, and the body weight and the weight of both organs were recorded. Splenocyte cell suspensions were prepared and cultured in round-bottomed microtiter plates (Corning Costar, Acton, MA) following the in vitro 51Cr-release assay described by Smialowicz et al.(1991)Go and using an alternate method to measure 51Cr release by target cells. In the current assay, microtiter plates were centrifuged, and 25 µl of supernatant were transferred to a 96-well LumaPlate (Packard, Downers Grove, IL), covered, and read on a TopCount NXT (Packard) microplate scintillation counter. The data are presented as (mean counts per min - spontaneous release of 51Cr)/(average radioactivity of target cells - spontaneous release of 51Cr) x 100.

Macrophage phagocytosis.
The fourth subgroup was used to assay the phagocytic function of nonelicited peritoneal macrophages, using an in vitro protocol wherein latex bead phagocytosis is measured with a flow cytometer in a modification of the method of Brouseau et al.(1999). Seven-week-old rats were euthanized with an excess of CO2, the skin was cut with a scissors and pulled back from the abdomen, and Ca2+ and Mg2+ free Hanks’ Balanced Salt Solution (Invitrogen/Gibco, Carlsbad, CA) containing 10 I.-U./ml sodium heparin (Elkins-Sinn, Cherry Hill, NJ) was injected intraperitoneally. The abdomen was gently massaged and the peritoneal wash was removed through a small incision in the lateral peritoneal wall. The acrophages were pelleted by centrifugation at 4°C and resuspended in ice-cold RPMI, 10% FBS, and 50-ng/ml gentamicin. Cell viability was determined by trypan blue dye exclusion (at least 98% in all samples) and cells suspensions were adjusted to 2 x 106 cells/ml. For each animal, 500 µl of cell suspension was diluted to 1x106 cells/ml in RPMI and held at 37°C while 500 µl of cell suspension was diluted to 1 x 106 cells/ml in RPMI + 0.4 % sodium azide (Sigma) and kept at 4°C. After 30 min, yellow-green carboxylate modified microspheres (1.0 mm; Molecular Probes, Eugene, OR) were added to each sample at 100:1 (beads:cell), vortexed, and returned to the previous incubation temperature (37°C or 4°C) and subjected to gentle agitation. After 120 min, 500 µl of solution were layered over a 3% BSA/RPMI density gradient and centrifuged at 120 RCF and 4°C for 7 min to remove any free beads. The supernatants were discarded, and the cell pellets were then resuspended in 100-µl PBS 0.05% formalin (Sigma) and analyzed within 24 h using a Coulter Epics XL-MCL flow cytometer equipped with a 488-nm argon laser (Coulter, Miami, FL). Cell debris were excluded using a minimum forward scatter (FSC) gate, and data from a minimum of 10,000 cells were acquired in FSC and Log side scatter (SSC) from the predominantly macrophage population. Data from FL1 (the channel of peak fluorescence of the latex beads) were also recorded, and all data were analyzed with WinMDI software (Ver 2.8) kindly donated by J. Trotter (2000)Go.

In WinMDI, the data from each sample were transformed into the number of beads engulfed by each macrophage. FL1 data were displayed on a histogram with the first fluorescent peak corresponding to a single fluorescent bead encountered by the flow cytometry laser; each successive fluorescent peak corresponded to the phagocytosis of an additional bead. Markers were created to define cells with fluorescence >= the first FL1 peak (cells engulfing 1 or more beads) and fluorescence >= the third FL1 peak (cells with 3 or more beads). From these two markers in the FL1 histograms, two types of data were generated. The first data type was a measure of how many macrophages were active, and for this purpose the percent of cells that phagocytized one or more and three or more beads was calculated. The second data type was a measure of how active the macrophages were. The mean number of beads phagocytized per active macrophage (cells with one or more beads) was calculated as the indicator of the level of phagocytosis carried out by active macrophages. For each data type, the number of beads detected in cells incubated at 4°C was subtracted from the number of beads detected in pair-matched cells incubated at 37°C to control for simple adherence of beads to cells.

Statistics
All data are presented as the mean ± SEM. The data were analyzed with the Equality of Variance F Test using Statview 5.1 (1998, SAS Institute, Inc., Cary, NC). Homoscedastic data were analyzed using analysis of variance (ANOVA) for treatment and treatment x gender interactions. When heteroscedastic data were encountered, nonparametric tests were utilized (the Mann Whitney U test for comparisons between two groups and Kruskal Wallis for comparison among three or more groups). Nonparametric data (e.g., ranked data on T4 and T3 concentrations) were also analyzed with the Kruskal Wallis test for an effect of treatment and the Mann Whitney U test for specific comparisons to control or between sexes. In all cases, a p-value of less than 0.05 was considered significant.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Offspring Mortality
Offspring mortality measured on a per-pup basis in offspring of ATR-exposed dams was greater (p = 0.0007) than that of offspring from control dams and did not differ between controls and other treatment groups (Table 1Go). When mortality is analyzed on a per-litter basis, one or more pups died in a greater percentage of the ATR (p < 0.0001) and BCR (p = 0.012) litters than in the control or PTU litters. These numbers do not include litters from dams that died during the experiment (n = 1 BCR-exposed dam), cannibalized pups (n = 1 control dam), or abandoned their litters (n = 2 ATR-exposed dams).


View this table:
[in this window]
[in a new window]
 
TABLE 1 Early Offspring Survival by Litter
 
Body, Spleen, and Thymus Weights
There were no differences in the mean body weights of dams before treatment started (GD 10), dam weight at the end of gestation (weight on GD 21), or dam weight on the day pups were weaned (PND 22) among treatment groups (data not shown). Pups in the first experiment were not weighed before PND 7. However, in the second experiment, the pups were weighed within 36 h of birth (PND 2); PND 2 pups of each sex from BCR dams were smaller than the pups from the control dams (Table 2Go). Offspring in all treatment groups displayed the expected sexual dimorphism in body weight, with males larger than females at PND 2. By PND 7, male ATR pups were smaller than the male control pups, and BCR and PTU pups of each sex were smaller than the control pups (Table 2Go). On PND 7, ATR and BCR pups displayed no sexual dimorphism in body size, although female pups were smaller than treatment-matched males in both the control and PTU groups.


View this table:
[in this window]
[in a new window]
 
TABLE 2 Early Body Weight of Offspring
 
At 2 weeks of age, BCR and PTU pups of each sex were smaller than the respective control pups (Fig. 1AGo). The spleens and thymi were therefore analyzed relative to body weight. Only the spleen to body weight ratio of BCR pups (p = 0.0350 males and p = 0.0302 females) and the thymus to body weight ratio of PTU pups (p = 0.0050 males and p = 0.0028 females) were smaller than those of the control pups on PND 14. At 7 weeks of age there was no difference in body, spleen, or thymus weight among treatments (Fig. 1BGo), and offspring from all treatment groups displayed an expected sexually dimorphic body weight.



View larger version (37K):
[in this window]
[in a new window]
 
FIG. 1. Body, spleen, and thymus weights of offspring from ATR-, BCR-, and PTU-exposed dams on (A) PND 14 and (B) PND 62. Data are presented as means of two separate experiments ± SEM. *Difference from same-sex control (p < 0.05); **difference from same-sex control (p < 0.001).

 
T4 and T3
Serum T4 and T3 concentrations were below detection limits (2 ng/ml and 0.1 ng/ml, respectively) in 14-day-old offspring of PTU-treated dams. Therefore, treatment differences in T4 and T3 concentrations were tested using Kruskal-Wallis nonparametric analyses of ranks substituting a value of zero for nondetectable values. The offspring of PTU-exposed dams had lower T4 (p = 0.0033) and lower T3 (p = 0.0480) than the offspring of the control animals (Table 3Go); there were no differences in T4 and T3 concentrations among other treatments or between sexes within a treatment group. The intraassay coefficients of variation were 7.7% and 1.8% for the T4 and T3 assays, respectively.


View this table:
[in this window]
[in a new window]
 
TABLE 3 Serum T4 or T3 Concentration of Offspring on Postnatal Day 14
 
Antibody Production
The primary (IgM) antibody response to SRBC was suppressed in ATR-exposed male (p = 0.0149), but not female, offspring (Fig. 2AGo) 8 weeks of age. The female offspring from BCR-exposed dams had an increased primary antibody response relative to the control females. No other treatment group differed from the same-sex control animals. There was no difference among treatment groups in the IgG antibody response to SRBC (Fig. 2BGo). There was no difference between the offspring of ATR-exposed dams and the control offspring in the IgM antibody response to SRBC in 6-month-old animals (data not shown).



View larger version (21K):
[in this window]
[in a new window]
 
FIG. 2. (A) Serum concentration of SRBC-specific IgM in 8-week-old SD rats exposed perinatally to ATR, BCR, or PTU. (B) Serum concentration of SRBC-specific IgG in perinatally exposed rats given a second SRBC immunization 4 weeks after the primary immunization. *Difference from same-sex control (p < 0.05). Data are presented as means of two separate experiments ± SEM from males ({blacksquare}) and females ({square}).

 
DTH Response
The DTH response of male, but not female, offspring of ATR dams was reduced relative to the control animals (p = 0.0308) when the animals were sensitized to BSA at 8 weeks of age and measured at 9 weeks of age (Fig. 3AGo). The decrease in DTH response of male ATR offspring persisted until at least 12 weeks of age (p = 0.0496; Fig. 3BGo). The data shown in Figure 3BGo were from animals in a single experiment sensitized to BSA at 11 weeks of age and measured at 12 weeks of age. There was no difference in the DTH response between the offspring of ATR-exposed dams and the control offspring in 6-month-old animals (data not shown).



View larger version (20K):
[in this window]
[in a new window]
 
FIG. 3. (A) DTH response to BSA in SD rats exposed perinatally to ATR, BCR, or PTU and sensitized at 8 weeks of age and measured at 9 weeks of age. (B) DTH response to BSA in SD rats exposed perinatally to ATR and sensitized to BSA at 11 weeks of age and measured at 12 weeks of age. *Difference from same-sex control (p < 0.05). Data from (A) are presented as means of two separate experiments ± SEM and data from (B) are presented as the means from a single experiment with males ({blacksquare}) and females ({square}).

 
NK Cell Activity
There was no difference in 51Cr release from YAC-1 target cells (100:1, 50:1, and 25:1 effecter target cell ratios) among any of the treatment groups or between sexes within treatments in animals 9 weeks of age (data not shown). The average percent spontaneous Cr51 release of the two experiments was 8.5%.

In Vitro Macrophage Phagocytic Ability
A greater percentage of the peritoneal macrophage cell population was actively phagocytic (i.e., had a higher percentage of cells that phagocytized >= 3 three beads, p = 0.0012, and the percent of cells that phagocytized >= 1 bead was on the boarder of being significantly higher, p = 0.0531; Fig. 4AGo) in 7-week-old offspring of PTU dams when sex was not considered in the analyses. Actively phagocytic macrophages from PTU offspring also phagocytized a greater mean number of beads than did macrophages from other treatment groups (i.e., there were more beads per macrophage among macrophages that phagocytized >= one bead, p = 0.0037; Fig. 4BGo). Although there were no differences in any measure of phagocytic activity between sexes within the treatment groups, when each sex was analyzed separately only the female offspring of PTU dams had a greater percentage of actively phagocytic macrophages than the control offspring (as measured by the percentage of cells that phagocytized >= 3 beads, p = 0.0176) and a greater mean number of beads per active macrophage (p = 0.0100). No measure of phagocytosis assayed in animals from other treatments (ATR, BCR) differed from that of the control animals.



View larger version (29K):
[in this window]
[in a new window]
 
FIG. 4. In vitro phagocytic response to latex beads by peritoneal macrophages of SD rats perinatally exposed to ATR, BCR, or PTU and measured at 7 weeks of age. (A) The proportion of the macrophages in the population that was active with active cells defined as cells that phagocytized >= 1 bead and >= 3 beads. (B) The mean number of beads phagocytized per active macrophage is a measure of the level of activity of the macrophages. *Difference from same-sex control (p < 0.05). Data are presented as means of two separate experiments ± SEM.

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
This is the first report demonstrating that developmental ATR exposure causes gender-dependent immunosuppression of cellular and humoral immune function. Much attention has recently been focused on defining critical windows in the development of immune function and periods of increased susceptibility to immunotoxicants (West, 2002Go). The immunotoxic effects of chemical exposure during the fetal and early neonatal period are often more dramatic and persistent than those after exposure during adult life (Holladay and Smialowicz, 2000Go). The suppression of primary antibody production and DTH response was observed after developmental exposure to 35-mg/kg ATR in the current study. Previous studies in adult rats and mice found no effects on immune function at less than overtly toxic doses (500–900 mg/kg; Vos et al., Munson et al.) or when ATR exposure occurred concomitantly with other chemicals (Porter et al., 1999Go).

A number of immunotoxicants cause persistent suppression of immune function when exposure occurs during development. For example, Kalland (1984)Go reported the suppression of DTH responses and reduced NK activity in 17-month-old mice exposed to DES for the first 5 days of life. Similarly, DTH responses in female BALB/c mice were suppressed at 100 and 200 days of age after prenatal exposure to chlordane (4 mg/kg/day for 18 days) at doses that produce no apparent immunotoxicity when adults are exposed (Barnett et al., 1990Go). In the current studies, the suppression of the primary antibody response and DTH persisted until the male offspring were 8 or 12 weeks old, respectively. To fully appreciate the developmental immunotoxicity of ATR, we need to establish whether immunotoxicity is present from birth, and determine more precisely when function returns to normal (between 12 weeks and 6 months) in offspring.

Long-term reproductive dysfunction and reproductive cancers resulting from neonatal exposure to endocrine disrupting chemicals (EDCs) have garnered much of the attention given to EDCs, but it is also known that developmental exposure to EDCs, such as DES, have been found to result in long-term immune suppression (Kalland, 1984Go). In the current study, the dose of ATR (35 mg/kg/day) was set just above the NOEL for ATR-induced developmental endocrine dysfunction to examine the potential for ATR to indirectly induce immunotoxicity. Previous studies have shown that ATR disrupts the endocrine system in rodents with effects ranging from reduced serum PRL (Cooper et al., 2000Go) to increased occurrence of mammary tumors (Stevens et al., 1999Go; Wetzel et al., 1994Go). Hypoprolactinemic and hypothyroid dams were included to determine whether pharmacological suppression of either hormone during development caused effects that were similar to those observed with ATR exposure. The results shown here discount early developmental suppression of either PRL or THs as likely endocrine mechanisms of ATR-mediated immunosuppression because of the failure to replicate ATR-associated immunotoxicity through the inhibition of either PRL or THs early in development. Additionally, manipulation of PRL or THs produced two responses not found in ATR-treated animals: (1) PTU exposure increased the phagocytic response in the offspring of both sexes and (2) BCR increased the primary antibody response of the female offspring. The experiment does not address circulating PRL or TH concentrations at the time immune function was tested; concentrations at that time may still correlate to immune function and/or developmental exposure to ATR. Future studies of the mechanisms by which ATR produces developmental immunotoxicity should include concurrent measures of circulating PRL and THs to more definitively discount PRL and THs as proximate mechanisms of ATR immunotoxicity.

Furthermore, the current data indicate that immunosuppression by ATR was restricted to the male offspring. The data do not explain why only males were affected; however, there are other chemicals that produce gender-restricted immunotoxicity (e.g., heptachlor; Smialowicz et al., 2001Go). The gender specificity of ATR’s immunotoxic effects in rats may be related to the ability of ATR to dampen the pulsatile gonadotropin releasing hormone (GnRH) secretion from the hypothalamus (Cooper et al., 2000Go). Gender differences are well established in the neuroendocrine events resulting from GnRH release in rodents and humans, including differential secretion of gonadotropins (luteinizing hormone and follicle-stimulating hormone), which in turn lead to differences in gonadal steroids (estradiol and testosterone) (Fallest et al., 1995Go; Potau et al., 1999Go). Similar gender differences in paracrine GnRH secretion and responses to GnRH have recently been demonstrated within immune tissue (Jacobson, 2000Go).

The increased percent mortality of ATR litters during PNDs 2–14 was unexpected, as this has not been reported for this dose of ATR. ATR and BCR treatment also decreased the weight of male pups such that, for a brief age group (PND 7), there was no difference between male and female body weights in either the ATR or BCR treatment groups when there was the expected sexual dimorphism in the control pups. It is not known whether the weight loss or death of pups was due to an effect on the physiology or behavior of the dams or the pups themselves. SD dams within 2 days of giving birth accept pups from other dams, and cross-fostering is a common tool in developmental toxicology studies (Mahle et al., 2003Go). Although PRL was not measured in the dams, previous studies have demonstrated that both ATR and BCR reduce PRL and the suckling induced PRL surge in dams. Furthermore, PRL is an important endocrine regulator of maternal behavior and stimulates dams to stay near, groom, and feed pups, in addition to the role that PRL plays in the physiological process of milk letdown (Sakaguchi et al., 1996Go). This effect bears further investigation to determine if it can be reproduced in litters that are maintained intact, if it is present in litters cross-fostered intact, or if it is exacerbated by stress.

The endocrine disrupting properties of ATR on PRL and THs led us to hypothesize that ATR is a developmental immunotoxicant; however, the current data do not support nor completely refute a relationship between ATR’s immunotoxic potential and its endocrine disrupting properties. The potential for EDCs to act as immunotoxicants suggests that it may be advantageous to examine immunotoxicity after finding endocrine disruption in the Reproduction/Developmental Toxicity Screening Test or at least consider developmental immunotoxicity in EDCs even if there is no apparent adult immunotoxicity. Current EPA Health Effects Test Guidelines for immunotoxicity do not suggest scrutiny of susceptible populations (e.g., young or old animals) (U.S. EPA, 1998Go), while Neurotoxicity and Reproductive Toxicity Guidelines (which comprise much of endocrine toxicity guiding principles) do a much better job of considering susceptible populations by advocating developmental exposure studies in addition to general assays.


    ACKNOWLEDGMENTS
 
We gratefully acknowledge the critical technical assistance of C.B. Copeland and W.C. Williams throughout the study. We thankfully acknowledge the contributions of Dr. R. J. Smialowicz on assay design and optimization as well as his comments on the manuscript. We also thank Dr. S.D. Holladay and Dr. L.S. Birnbaum for providing helpful suggestions on the manuscript. Support was provided in part by the NCSU/EPA Cooperative Training Program in Environmental Sciences Research, Training Agreement CT826512010 with North Carolina State University under the helpful administration of Dr. K. Adler. Support was also provided in part by the American Chemistry Council under CRADA 0215–02.


    NOTES
 
This paper has been reviewed by the Environmental Protection Agency’s Office of Research and Development 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.

1 To whom correspondence should be addressed at U.S. Environmental Protection Agency, Division of Experimental Toxicology, Immunotoxicology Branch, B 143-01, Research Triangle Park, NC 27711. Fax: (919) 541-3538. E-mail: rooney.andrew{at}epa.gov. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Barnett, J. B., Blaylock, B. L., Gandy, J., Menna, J. H., Denton, R., and Soderberg, L. S. (1990). Long-term alteration of adult bone marrow colony formation by prenatal chlordane exposure. Fundam. Appl. Toxicol. 14, 688–695.[ISI][Medline]

Brousseau, P., Payette, Y., Tryphonas, H., Blakley, B., Boermans, H., Flipo, D., and Fournier, M. (1999). Manual of Immunological Methods. CRC Press, Boca Raton, FL.

Clark, G. M., and Goolsby, D. A. (2000). Occurrence and load of selected herbicides and metabolites in the lower Mississippi River. Sci. Total Environ. 248, 101–113.[CrossRef][ISI][Medline]

Cohen, S. Z., Creeger, S. M., Carsell, R. F., and Enfield, C. G. (1984). Potential pesticide contamination of groundwater from agricultural uses. In Treatment and Disposal of Pesticide Wastes (R. F. Krueger and J. N. Sieber, eds.), Vol. 259, p. 297, ACS Symposium Series. American Chemical Society.

Cooper, R. L., Stoker, T. E., Tyrey, L., Goldman, J. M., and McElroy, W. K. (2000). Atrazine disrupts the hypothalamic control of pituitary-ovarian function. Toxicol. Sci. 53, 297–307.[Abstract/Free Full Text]

Cummings, A. M., Rhodes, B. E., and Cooper, R. L. (2000). Effect of atrazine on implantation and early pregnancy in 4 strains of rats. Toxicol. Sci. 58, 135–143.[Abstract/Free Full Text]

Fabris, N., Mocchegiani, E., and Provinciali, M. (1995). Pituitary-thyroid axis and immune system: A reciprocal neuroendocrine- immune interaction. Horm. Res. 43, 29–38.[ISI][Medline]

Fallest, P. C., Trader, G. L., Darrow, J. M., and Shupnik, M. A. (1995). Regulation of rat luteinizing hormone beta gene expression in transgenic mice by steroids and a gonadotropin-releasing hormone antagonist. Biol. Reprod. 53, 103–109.[Abstract]

Fournier, M., Friborg, J., Girard, D., Mansour, S., and Krzystyniak, K. (1992). Limited immunotoxic potential of technical formulation of the herbicide atrazine (AAtrex) in mice. Toxicol. Lett. 60, 263–274.[CrossRef][ISI][Medline]

Gala, R. R. (1991). Prolactin and growth hormone in the regulation of the immune system. Proc. Soc. Exp. Biol. Med. 198, 513–527.[Abstract]

Holladay, S. D., and Smialowicz, R. J. (2000). Development of the murine and human immune system: differential effects of immunotoxicants depend on time of exposure. Environ. Health Perspect. 108(Suppl. 3), 463–473.

Jacobson, J. D. (2000). Gonadotropin-releasing hormone and G proteins: Potential roles in autoimmunity. Ann. N.Y. Acad. Sci. 917, 809–818.[Abstract/Free Full Text]

Kalland, T. (1980). Alterations of antibody response in female mice after neonatal exposure to diethylstilbestrol. J. Immunol. 124, 194–198.[Free Full Text]

Kalland, T. (1984). Exposure of neonatal female mice to diethylstilbestrol persistently impairs NK activity through reduction of effector cells at the bone marrow level. Immunopharmacology 7, 127–134.[CrossRef][ISI][Medline]

Kalland, T., and Forsberg, J.-G. (1978). Delayed hypersensitivity response to oxazolone in neonatally estrogenized mice. Cancer Lett. 4, 141–146.[ISI][Medline]

Kirby, J. D., Jetton, A. E., Cooke, P. S., Hess, R. A., Bunick, D., Ackland, J. F., Turek, F. W., and Schwartz, N. B. (1992). Developmental hormonal profiles accompanying the neonatal hypothyroidism-induced increase in adult testicular size and sperm production in the rat. Endocrinology 131, 559–565.[Abstract]

Kornilovskaya, I. N., Gorelaya, M. V., Usenko, V. S., Gerbilsky, L. V., and Berezin, V. A. (1996). Histological studies of atrazine toxicity on the thyroid gland in rats. Biomed. Environ. Sci. 9, 60–66.[Medline]

Laws, S. C., Ferrell, J. M., Stoker, T. E., Schmid, J., and Cooper, R. L. (2000). The effects of atrazine on female wistar rats: An evaluation of the protocol for assessing pubertal development and thyroid function. Toxicol. Sci. 58, 366–376.[Abstract/Free Full Text]

Mahle, D. A., Yu, K. O., Narayanan, L., Mattie, D. R., and Fisher, J. W. (2003). Changes in cross-fostered Sprague-Dawley rat litters exposed to perchlorate. Int. J. Toxicol. 22, 87–94.[CrossRef][ISI][Medline]

Munson, A. E., White, K. L., and McCay, J. A. (1994). Immunotoxicity of atrazine in female B6C3F1 mice: Report to National Toxicology Program, pp. 1–54. Immunotoxicology Program Medical College of Virginia, Richmond.

Nagy, E., Berczi, I., Wren, G. E., Asa, S. L., and Kovacs, K. (1983). Immunomodulation by bromocriptine. Immunopharmacology 6, 231–243.[CrossRef][ISI][Medline]

Nagy, G. M., Gorcs, T. J., and Halasz, B. (1991). Attenuation of the suckling-induced prolactin release and the high afternoon oscillations of plasma prolactin secretion of lactating rats by antiserum to vasopressin. Neuroendocrinology 54, 566–570.[ISI][Medline]

Porter, W. P., Jaeger, J. W., and Carlson, I. H. (1999). Endocrine, immune, and behavioral effects of aldicarb (carbamate), atrazine (triazine) and nitrate (fertilizer) mixtures at groundwater concentrations. Toxicol. Ind. Health 15, 133–150.[CrossRef][ISI][Medline]

Potau, N., Ibanez, L., Sentis, M., and Carrascosa, A. (1999). Sexual dimorphism in the maturation of the pituitary-gonadal axis, assessed by GnRH agonist challenge. Eur. J. Endocrinol. 141, 27–34.[ISI][Medline]

Sakaguchi, K., Tanaka, M., Ohkubo, T., Doh-ura, K., Fujikawa, T., Sudo, S., and Nakashima, K. (1996). Induction of brain prolactin receptor long-form mRNA expression and maternal behavior in pup-contacted male rats: Promotion by prolactin administration and suppression by female contact. Neuroendocrinology 63, 559–568.[ISI][Medline]

Short, P., and Colborn, T. (1999). Pesticide use in the U.S. and policy implications: A focus on herbicides. Toxicol. Ind. Health 15, 240–275.[CrossRef][ISI][Medline]

Shyr, S. W., Crowley, W. R., and Grosvenor, C. E. (1986). Effect of neonatal prolactin deficiency on prepubertal tuberoinfundibular and tuberohypophyseal dopaminergic neuronal activity. Endocrinology 119, 1217–1221.[Abstract]

Smialowicz, R. J., Simmons, J. E., Luebke, R. W., and Allis, J. W. (1991). Immunotoxicologic assessment of subacute exposure of rats to carbon tetrachloride with comparison to hepatotoxicity and nephrotoxicity. Fundam. Appl. Toxicol. 17, 186–196.[ISI][Medline]

Smialowicz, R. J., Williams, W. C., Copeland, C. B., Harris, M. W., Overstreet, D., Davis, B. J., and Chapin, R. E. (2001). The effects of perinatal/juvenile heptachlor exposure on adult immune and reproductive system function in rats. Toxicol. Sci. 61, 164–175.[Abstract/Free Full Text]

Stevens, J. T., Breckenridge, C. B., and Wetzel, L. (1999). A risk characterization for atrazine: Oncogenicity profile. J. Toxicol. Environ. Health 56, 69–109.[CrossRef][ISI]

Stoker, T. E., Laws, S. C., Guidici, D. L., and Cooper, R. L. (2000). The effect of atrazine on puberty in male wistar rats: An evaluation in the protocol for the assessment of pubertal development and thyroid function. Toxicol. Sci. 58, 50–59.[Abstract/Free Full Text]

Stoker, T. E., Robinette, C. L., and Cooper, R. L. (1999a). Maternal exposure to atrazine during lactation suppresses suckling-induced prolactin release and results in prostatitis in the adult offspring. Toxicol. Sci. 52, 68–79.[Abstract]

Stoker, T. E., Robinette, C. L., and Cooper, R. L. (1999b). Perinatal exposure to estrogenic compounds and the subsequent effects on the prostate of the adult rat: Evaluation of inflammation in the ventral and lateral lobes. Reprod. Toxicol. 13, 463–472.[CrossRef][ISI][Medline]

Temple, L., Kawabata, T. T., Munson, A. E., and White, K. L., Jr. (1993). Comparison of ELISA and plaque-forming cell assays for measuring the humoral immune response to SRBC in rats and mice treated with benzo[a]pyrene or cyclophosphamide. Fundam. Appl. Toxicol. 21, 412–419.[CrossRef][ISI][Medline]

Trotter, J. (2000). WinMDI. 2.8. (available at http://facs.scripps.edu/software.html).

U.S. EPA (1991). Spring sampling finds herbicides throughout Mississippi River and tributaries. In Report, p. 25. U.S. Department of Interior, U.S. Geological Survey, Reston, VA.

U.S. EPA (1998). Health effects test guildelines: OPPTS 870.7800 Immunotoxicity. In Report, EPA-712-C-98-351, p. 11. Office of Prevention, Pesticides and Toxic Substances, U.S. Environmental Protection Agency, Washington, D.C.

U.S. EPA (2001). Revised preliminary human health risk assessment: Atrazine. In Report Case No. 0062, p. 67. Health Effects Division, Office of Pesticide Programs, U.S. Environmental Protection Agency, Washington, D.C.

Ugazio, G., Bosio, A., Nebbia, C., and Soffietti, M. G. (1991). Age- and sex-related effects on hepatic drug metabolism in rats chronically exposed to dietary atrazine. Res Commun Chem Pathol Pharmacol 73, 231–243.[ISI][Medline]

Vos, J. G., Krajnc, E. I., Beekof, P. K., and van Logten, M. J. (1983). Methods for testing immune effects of toxic chemicals: evaluation of the immunotoxicity of various pesticides in the rat. In IUPAC Pesticide Chemistry, Human Welfare and the Environment. Vol. III. Mode of Action, Metabolisms and Toxicology (J. Miyamoto, ed.), pp. 497–504. Pergamon Press, Oxford.

West, L. J. (2002). Defining critical windows in the development of the human immune system. Hum. Exp. Toxicol. 21, 499–505.[CrossRef][ISI][Medline]

Wetzel, L. T., Luempert, L. G., 3rd, Breckenridge, C. B., Tisdel, M. O., Stevens, J. T., Thakur, A. K., Extrom, P. J., and Eldridge, J. C. (1994). Chronic effects of atrazine on estrus and mammary tumor formation in female Sprague-Dawley and Fischer 344 rats. J. Toxicol. Environ. Health 43, 169–182.[ISI][Medline]