A Mixture of Five Phthalate Esters Inhibits Fetal Testicular Testosterone Production in the Sprague-Dawley Rat in a Cumulative, Dose-Additive Manner

Abstract

Phthalate diesters are chemicals to which humans are ubiquitously exposed. Exposure to certain phthalates during sexual differentiation causes reproductive tract malformations in male rats. In the fetal rat, exposure to the phthalates benzylbutyl phthalate (BBP), di(n)butyl phthalate (DBP), and diethylhexyl phthalate (DEHP) decreases testicular testosterone production and insulin-like 3 hormone mRNA levels. We characterized the dose-response effects of six individual phthalates (BBP, DBP, DEHP, diethyl phthalate [DEP], diisobutyl phthalate [DiBP], and dipentyl phthalate [DPP]) on gestation day (GD) 18 testicular testosterone production following exposure of Sprague-Dawley rats on GD 8–18. BBP, DBP, DEHP, and DiBP were equipotent (ED50 of 440 ± 16 mg/kg/day), DPP was about threefold more potent (ED50 = 130 mg/kg/day) and DEP had no effect on fetal testosterone production. We hypothesized that coadministration of these five antiandrogenic phthalates would reduce testosterone production in a dose-additive fashion because they act via a common mode of toxicity. In a second study, dams were dosed at 100, 80, 60, 40, 20, 10, 5, or 0% of the mixture. The top dose contained 1300 mg of total phthalates/kg/day including BBP, DBP, DEHP, DiBP (300 mg/kg/day per chemical), and DPP (100 mg DPP/kg/day). This mixture ratio was selected such that each phthalate would contribute equally to the reduction in testosterone. As hypothesized, testosterone production was reduced in a dose-additive manner. Several of the individual phthalates and the mixture also induced fetal mortality, due to pregnancy loss. These data demonstrate that individual phthalates with a similar mechanism of action can elicit cumulative, dose additive effects on fetal testosterone production and pregnancy when administered as a mixture.

androgens, prenatal, reproductive tract, phthalates, cumulative risk, structure activity relationship

Phthalate esters are commonly used as plasticizers, solvents, and emulsifiers in polyvinyl chloride plastics, medical tubing, intravenous bags, pharmaceuticals, pesticides, health, and beauty products, and children's toys. Furthermore, phthalate monitoring data demonstrate that humans are exposed to multiple phthalates and their metabolites. Phthalates and/or their metabolites have been detected in the saliva and urine of children and adults (Silva et al. 2004a, 2005), amniotic fluid of second trimester fetuses (Silva et al., 2004b), cord blood of newborns (Latini et al., 2003), human breast milk (Mortensen et al., 2005), and urine of neonates in intensive care units (Green et al., 2005).

Prenatal exposure to some phthalate esters during the period of sexual differentiation disrupts male reproductive development in the rat and rabbit (Higuchi et al., 2003). In utero exposure to benzyl butyl phthalate (BBP), dibutyl phthalate (DBP), or diethylhexyl phthalate (DEHP) suppresses fetal testosterone production and insulin-like 3 hormone (insl3) mRNA levels (Parks et al., 2000; Mylchreest et al., 2002; Wilson et al., 2004) inducing a suite of reproductive malformations in the male rat termed the “phthalate syndrome” (Gray et al., 2000; Mylchreest et al., 1998). Therefore, several of the phthalate diesters disrupt the reproductive development of the fetal male rat via a common mode of toxicity.

Because testosterone is critical for normal sexual differentiation in male rodents as well as humans, an ongoing United States Environmental Protection Agency (USEPA) risk assessment on DBP (USEPA, 2006) proposed fetal testicular testosterone levels (Lehmann et al., 2004) as the “critical” endpoint to set the no observable adverse effect level (NOAEL). The proposed EPA risk assessment on DBP has been criticized by several environmental and public health advocacy groups for failure to consider the cumulative effects of other phthalates with DBP, among other reasons (Janssen, 2006). In addition, there have been congressional inquiries regarding the need to account for the cumulative effects of phthalates because people are exposed to multiple phthalates that are known to affect testosterone levels in fetal rats. In response, a National Research Council committee was established at the request of the USEPA to evaluate the published scientific data on phthalate mixtures in laboratory animals, address questions related to human relevance of experimental data, modes of action, exposure information, dose-response assessment, and the potential for cumulative effects. The committee will provide recommendations to the USEPA on conducting a cumulative risk assessment on the phthalates, including conducting cumulative risk assessment for other chemical classes. (http://www8.nationalacademies.org/cp/projectview.aspx?key=48860).

To date, there are only a few published studies that have tested the effects of prenatal exposure to combinations of phthalates with each other or in combination with other antiandrogenic chemicals. Although phthalates do not act as classical antiandrogenic chemicals by binding to the androgen receptor, they do have the same effect of blocking androgen-action at the target tissue and thus we refer to them herein as “antiandrogenic”.

Howdeshell et al. (2007) demonstrated that a mixture of DBP and DEHP, two phthalates with a common mode of action but different active metabolites, altered fetal testosterone production and insl3 gene expression in a manner that resulted in cumulative dose-additive increases in reproductive tract malformations. Hotchkiss et al. (2004) reported that a mixture of the phthalate BBP and the antiandrogenic pesticide linuron also produced dose-additive effects on reproductive tract development, even though these chemicals do not share a common mechanism of toxicity. In addition, a recent study from our laboratory dosed rat dams with a combination of seven antiandrogenic chemicals with different mechanisms of action, including three phthalates (BBP, DBP, and DEHP), and observed cumulative, dose-additive effects on androgen-dependent reproductive development (Rider et al., 2008).

The current study was designed to provide dose-response information on the effects of several individual phthalate diesters on fetal testosterone production, and to determine if these data could be used to predict the effects of a mixture of five antiandrogenic phthalates on fetal rat testosterone production. Additional studies are currently examining the effects of the phthalates and the mixture on fetal testis gene expression, and we are evaluating the effects of these fetal genomic changes on the postnatal development of male rats after in utero exposure to the mixture. We hypothesized that prenatal exposure of male rats to a mixture of five phthalates with similar modes of action, some with different active metabolites, would act in a cumulative, dose-additive fashion to decrease fetal testosterone production on gestational day (GD) 18. This fetal stage was selected because it is near the end of the hormone-dependent stage of differentiation of the male reproductive tract in the rat, yet before the maturation of hypothalamic-pituitary-gonadal feedback system for regulating testosterone production (El-Gehani et al. 1998).

In the first study, pregnant dams were dosed by gavage from GD 8 to GD 18 with graded doses of BBP, DBP, DEHP, diethyl phthalate (DEP), DiBP, or dipentyl phthalate (DPP) to determine the effective dose which inhibited fetal testosterone production by 50% (ED50) and the slope of the dose-response curve for each phthalate. Based on the structure activity relationship of phthalates to induce testicular atrophy in immature male rats (Foster et al., 1980), we predicted that DiBP and DPP would act similarly to BBP, DBP, and DEHP (Wilson, et al., 2004) to inhibit fetal testosterone production, and that DEP would not influence fetal steroidogenesis. In addition to measuring testosterone production of the DBP-exposed male fetuses, we extracted testosterone from the testes in order to compare the sensitivity and precision of these two endpoints, because the USEPA draft risk assessment on DBP was based upon extracted testicular testosterone levels rather than testosterone production. In the second study, pregnant rats were dosed similarly with a mixture of the five phthalates that reduced testosterone production (BBP, DBP, DEHP, DiBP, and DPP), and fetal testicular testosterone production was measured on GD 18. The ratio of the phthalates in the mixture was based upon their relative potencies such that each of the five phthalates would contribute equally to the reduction in testosterone if they behaved in a cumulative dose-additive manner. Dose addition has been validated extensively for use with mixtures of chemicals that have the same mechanism of toxicity (Altenburger et al., 2005; Berenbaum, 1985; Cassee et al., 1998; Konemann, 1981). According to the dose addition model, chemical doses are added together after first accounting for their individual potencies. The overall observed mixture response was then compared with the predictions of the dose addition model for these five phthalates.

MATERIALS AND METHODS

Animals

Adult female Sprague-Dawley (SD) rats (Charles River, Raleigh, NC) were mated by the supplier and shipped on gestation day 2 (GD 2). Mating was confirmed by sperm presence in vaginal smears (day of sperm plug positive = GD 1). Animals were housed individually in 20 × 25 × 47 cm clear polycarbonate cages with laboratory-grade heat-treated pine shavings (Northeastern Products, Warrensburg NY) with a 14:10 light/dark photoperiod (lights off at 11:00 A.M.) at 20–24°C. Pregnant and lactating females were fed Purina Rat Chow 5008 ad libitum. Animals were provided access to filtered (5 μm filter) municipal drinking water (Durham, NC) ad libitum. Water was tested monthly for Pseudomonas and every 4 months for a suite of chemicals including pesticides and heavy metals. The current study was conducted under protocols approved by the National Health and Environmental Effects Research Laboratory Institutional Animal Care and Use Committee in an Association for Assessment and Accreditation of Laboratory Animal Care.

Doses and Administration of Chemicals

Individual chemical study.

Pregnant rat dams were assigned to treatment groups on GD 8 in a manner that provided similar mean (± SE) body weight per treatment group prior to dosing. In the first experiment, the dams were dosed by gavage on GD 8–18 with either 0 (vehicle control) or a dose range of an individual phthalate. The individual phthalate dose ranges were BBP, DEHP, DEP, or DiBP at 100, 300, 600, and 900 mg/kg/day; or DBP at 33, 50, 100, 300, and 600 mg/kg/day; or DPP at 25, 50, 100, 200, 300, 600, and 900 mg/kg/day. The individual phthalate dose-response curves were conducted in two to four blocks with one to two dams per treatment per block for a total of 10–12 dams per block with the exception of the DEHP individual dose response. For the DEHP dose-response curve, the experiment was run in four blocks with one dam per treatment in the 0 (control), 100, 300, 600, or 900 mg DEHP/kg/day dose groups (n = 4 per dose). The total dams per treatment group in the BBP and DEP individual dose response were 0 (control), n = 9; 100 mg BBP/kg/day and DEP at 100 and 900 mg/kg/day, n = 4; and BBP at 300, 600, and 900 mg/kg/day and DEP at 300 and 600, n = 5. The total dams per treatment group in the DBP individual dose response were 0 (control), and DBP at 33, 50, 100, 300, and 600 mg/kg/day, n = 4 (incomplete block design). The total dams per dose group in the DiBP individual dose response were 0 (control), n = 5; 100 and 300 mg/kg/day, n = 8; and 600 and 900 mg/kg/day, n = 5. The total dams per dose group in the DPP individual dose response were 0 (control), n = 6; 25 mg/kg/day, n = 5; 50 and 200 mg/kg/day, n = 4; 300 mg/kg/day, n = 3; and 600 and 900 mg DPP/kg/day, n = 2. In this discussion, we refer to the phthalates that inhibit testosterone synthesis and pesticides that act as androgen receptor antagonists as antiandrogens, because both classes of chemicals disrupt the androgen-signaling pathway in the target tissues, either indirectly or directly.

Mixture study.

In the second experiment, dams were dosed via gavage from GD 8 to GD 18 with either 0 (vehicle control) or seven dose levels of the mixture. The top dose (100%) included a total of 1300 mg of the five phthalates (BBP, DBP, DEHP, and DiBP each at 300 mg/kg/day, plus 100 mg/kg DPP/day) and was administered at 100, 80, 60, 40, 20, 10, and 5% of the top dose. This dose ratio (3:3:3:3:1) was selected based upon the ED50 values (mg/kg/day) from the dose-response studies of each phthalate in order that each phthalate would contribute equally to the mixture effects on testosterone production if the mixture behaved in a dose additive manner. The top dose was expected to produce a dramatic reduction in fetal testosterone production without causing severe maternal toxicity (death, extreme weight loss, tremors, salivation, lacrimation, etc.) if the phthalates behaved in an additive manner as hypothesized. The mixture experiment was conducted in four blocks (incomplete block design). Each block had six treatments with two dams per treatment (12 dams per block). Two control dams were included in every block. The total dams per dose in the phthalate mixture experiment was n = 8 per group for 0, 260, 520, and 780 mg total phthalates/kg/day treatments and n = 4 per group in the 65, 130, 1040, and 1300 mg total phthalates/kg/day groups. The larger sample sizes were used in the control group and the 20–80% affected area of the dose-response curve to enhance the precision of the estimate of the ED50 and slope.

Chemicals.

The vehicle control, laboratory-grade corn oil (CAS 8001-30-7, Cat# C-8627), and the phthalates used in this study were purchased from Sigma-Aldrich (St Louis, MO): BBP (CAS 85-68-7, Cat# 308501, lot# 08523JQ, purity 98%); DBP (CAS 84-74-2, Cat# D-2270, lot# 109F0386, purity = 99%); DEHP (CAS 117-81-7, Cat# P-6699, lot# 101K3696 for individual dose response and lot# 106H3487 for combination dose, purity = 99%); DEP (CAS #84-66-2, Cat# P-5787, lot #48H3537, purity 99%); DiBP (CAS 84-69-5, Cat# 152641, lot# 103141C, purity =99%), and DPP (CAS 131-18-0, Cat# 80154, lot# 1151652, purity = 99%). The purity of each chemical was provided by the vendor. The doses were dissolved in corn oil and delivered in 2.5 ml of corn oil per kg body weight. The rat dams were weighed daily during the dosing period to administer the dose per kg body weight and to observe the health of the dams. Following the last dosage on the morning of GD 18, the rat dams were anesthetized with CO₂ and killed by decapitation. The uterus was removed and the number of fetuses (live and dead) and resorptions were counted and recorded. The total number of implantations was calculated by adding together the number of live and dead fetuses with the total number of resorptions. Fetal mortality was calculated by adding together the number of resorptions and dead fetuses, then dividing by the total implantations.

Fetal necropsy

Following the last dosage on the morning of GD 18, the rat dams were anesthetized with CO₂ and killed by decapitation. Once removed from the uterus, fetuses were anesthetized and killed on ice, and testes were removed using a dissecting microscope. Testes from the first three males were immediately transferred to M199 media without phenol red for ex vivo testis hormone production as per Wilson et al. (2004). Each individual testis was placed in a separate well for a total of two hormone production measurements per fetus, which resulted in six measurements per litter. Remaining testes were quickly transferred to TRI-Reagent (Sigma) in sterile 1.5 ml microcentrifuge tubes, homogenized with a Kontes pestle homogenizer on ice, and stored at −80°C until RNA isolation; testes were pooled per litter. Dissections were conducted within a 2-h period between 830 and 1030 Eastern Standard Time. For the DEHP individual dose response, testes from all the males were incubated for fetal testicular hormone production.

Ex Vivo Testis Testosterone Production

Fetal testicular hormone production was evaluated as per Wilson et al. (2004). The majority of the individual phthalate dose-response studies used a 3-h testes incubation period with the exception of the BBP and DEP study. The BBP and DEP studies incubated the testes for 2 h, thus resulting in lower total levels of testosterone production. Following incubation, the media was stored in siliconized eppendorf tubes and stored at −80°C until hormone radioimmunoassay (RIA). The incubated testes were immediately transferred to TRI-Reagent (Sigma), pooled by litter, and stored at −80°C for subsequent RNA analysis. Testosterone levels in the media were measured by RIA using Coat-a-Count kits according to manufacturer's protocols (Diagnostic Products Corporation, Los Angeles, CA; Coat-a-Count Kit Total Testosterone Manual, #PITKTT-4, 2005-03-18). The intra-assay coefficient of variation was 3.1%, based upon the variability of the standard curve and the interassay coefficient of variation was 13.7%. Cross reactivity with DHT was 3.2%. The limits of detection of the RIAs were 0.2 ng/ml testosterone. Data are presented as litter means.

Fetal Testicular Testosterone Extraction

Testosterone was extracted from the testes on GD 18 of SD rats exposed to DBP at 0 (10 fetuses, two litters), and 33 (2, 2), 50 (5, 2), 100 (6, 2), 300 (3, 2), or 600 (6, 2) mg/kg/day on GD 8–18. Both testes were collected from the males and placed in a microcentrifuge tube (one tube per male), frozen quickly on dry ice, and stored at −80°C until extraction. Fetal testes pairs were placed in a 12- × 75-mm borosilicate glass tube and 200 μl of the 0 standard from the Coat-A-Count testosterone RIA kit (Diagnostic Products Corporation) was added. Testes were then homogenized with a Teflon rod for about 30 s. One milliliter of ethyl ether was subsequently added to the testes homogenate. The testes/ethyl ether mixture was vortexed lightly, and then placed in a dry ice/ethanol bath for 10 s to freeze the aqueous portion of the mixture. The ether-extracted hormone, the unfrozen portion of the mixture, was poured into a clean 12- × 75-mm glass tube. Extraction was repeated with an additional 1 ml of ether, which was combined with the ether-extracted hormone solution from the first extraction. The ether-extracted hormone samples were allowed to evaporate overnight at room temperature. Following evaporation, the glass tubes containing the samples were sealed with parafilm and stored at 4°C until time of assay. For the testosterone RIA, the tubes were allowed to acclimate to room temperature and 100 μl of the 0 standard from the testosterone RIA kit was added to each tube. The samples were then heated for five minutes in a 40°C water bath, then vortexed. Testosterone levels in a 50-μl aliquot of each extracted testes sample was measured by RIA using Coat-a-Count kits according to manufacturer's protocols (Diagnostic Products Corporation); an additional standard of 0.1 ng/ml was added in duplicate to the standard curve.

Statistics

The data for each individual phthalate were analyzed using two-way ANOVA using the general linear measures procedures from the Statistical Analysis Systems (SAS, Inc., Cary, NC); the two factors used in this analysis were block and dose. Post hoc comparisons were made using the Least Squared Means procedure on SAS, which is appropriate for a priori hypotheses. We expected treatments to reduce testosterone production and to increase fetal mortality; the mortality being associated with difficulties with pregnancy maintenance. For analysis of treatment effects, litter means were used as the sample size. Differences were considered significant at p < 0.05.

We also analyzed the testosterone production data using a clustered model on PROC MIXED, which computes the variances within and between litters. These data can be used to determine how much statistical power is gained by measuring testosterone production from more than one pup per litter, which is useful for designing future studies.

Estimation of Dose Addition

The reduction in testosterone production and increase in fetal mortality were analyzed to determine if the effects were dose additive using methods described by Rider and LeBlanc (2005). The litter means, standard errors and samples sizes of the testosterone production (ng/testis) and fetal mortality (% mortality) data were fit to the four parameter logistic regression equation using Prizm 5 for Windows software, version 5.01 (GraphPad Software, San Diego, CA), which represents the sigmoidal fit of the data: 

(1)

where R is the response, D is the daily dose, the Hill slope is the slope of the curve and ED50 is the dose resulting in a 50% effect. For testosterone production, the bottom of the curve was constrained at 0 and the top of the curve was the testosterone production (ng/testis) of the concurrent control group. For fetal mortality, the bottom of the curve was constrained at 0% and the top at 100%. We used these methods to fit the data from the first experiment to calculate the ED50 and Hill slope parameters for each of the six phthalates. In addition, the analyses also produced goodness of fit values (R²), which describe how well each phthalate fit the logistic regression model, confidence limits and standard errors of the Hill slopes and ED50s.

The Hill slope and ED50 values from these calculations were used to predict the effects of the mixture of five of the above six phthalates on testosterone production and fetal mortality and to determine if the phthalates in the mixture behaved in a dose-additive manner.

Dose addition, referred to as concentration addition in Rider and LeBlanc (2005), is predicted to occur if chemicals work through a similar mechanism of action. We calculated dose addition for the mixture of five phthalates using the ED50 and the slopes for the individual phthalates with the following Equation (2): 

(2)

where R is the response of the mixture, D_i is the dose of individual chemical i in the mixture, ED50i is the dose of chemical causing a 50% response and ρ′ is the average Hill slope associated with the five phthalates in the mixture. We compared the predicted mixture effects on fetal mortality and testosterone production to the observed data using the 95% confidence limits of the observed data at each dosage level. In addition, we fit with the observed mixture results to the dose addition logistic regression model (the four parameters of the model were constrained to dose addition model values), to determine how well the dose addition model predicted the observed results.

RESULTS

Individual Phthalate Study

Although three of the phthalate diesters reduced maternal body weight gain (Tables 1–5), none of the treatments induced any signs of overt toxicity or maternal mortality. Dams treated with DPP at 300, 600, or 900 gained little or no weight from GD 8 to GD 18, whereas controls gained about 70 g. None of the dams in these dose groups had viable fetuses. Exposure to DiBP reduced maternal body weight gain from 73 g in controls to 48 and 43 g in the 600 and 900 mg/kg/day dose groups, respectively. Exposure to BBP reduced maternal body weight gain from 80 g in controls to 59, 44, and 43 g in the 300, 600, and 900 mg/kg/day dose groups, respectively. We did not observe any decrease in maternal body weight gain for DBP at the doses tested (33–600 mg/kg/day; Table 4) or DEHP (Table 5).

TABLE 1

Maternal Body Weight Gain, Litter Size, Resorptions, and Fetal Mortality of SD Rat Dams Administered a Dose Range of DPP from GD 8 to GD 18

	0 mg/kg/day	DPP25 (mg/kg/day)	DPP50 (mg/kg/day)	DPP100 (mg/kg/day)	DPP200 (mg/kg/day)	DPP300 (mg/kg/day)	DPP600 (mg/kg/day)	DPP900 (mg/kg/day)
No. of dams on GD 18	5	5	4	6	4	2	1	2
No. of dams with whole litter loss	0	0	0	0	0	2	1	2
Maternal body wt. at GD 8 (g)	269.8 ± 10.3	269.2 ± 10.3	266.3 ± 8.7	276.3 ± 7.4	277.3 ± 8.5	300.1 ± 6.2	292.3	295.1 ± 7.3
Maternal body wt. at GD 18 (g)	341.8 ± 9.8	341.3 ± 12.1	336.8 ± 9.2	339.7 ± 9.1	350.8 ± 2.4	323.0 ± 13.1	301.7	275.6 ± 5.5*
Maternal body wt. gain (g)	72.0 ± 5.9	72.1 ± 5.7	70.5 ± 3.9	63.4 ± 11.2	73.5 ± 7.9	19.9 ± 19.9*	9.4*	−19.5 ± 1.8*
No. of implantationsa	12.2 ± 1.2	12.2 ± 1.8	13.3 ± 0.9	13.5 ± 0.6	13.0 ± 0.4	16.0 (1)	13.0	10.0 (1)
No. of live fetuses	11.8 ± 1.4	12.0 ± 2.0	12.3 ± 0.9	12.8 ± 0.6	12.3 ± 0.5	0.0 ± 0.0*	0*	0.0 ±0.0*
Total resorptions	0.4 ± 0.3	0.2 ± 0.2	1.0 ± 0.4	0.7 ± 0.3	0.8 ± 0.5	16* (1)	13*	10* (1)
Fetal mortality (%)b	3.9 ± 2.4	4.0 ± 4.0	7.5 ± 3.0	4.9 ± 2.4	5.6 ± 3.7	100.0 ± 0.0*	100*	100 ± 0.0*

Note. Number in parentheses indicates number of litters in the analysis when different from number of dams on GD 18. Values in bold are significantly different than controls due to treatment.

*p < 0.05 versus control. Values are means ± standard errors.

a

Implantations = total live fetuses + total resorptions.

b

Fetal mortality = resorptions/implantations.

TABLE 2

Maternal Body Weight Gain, Litter Size, Resorptions, and Fetal Mortality of SD Rat Dams Administered a Dose Range of DiBP from GD 8 to GD 18

	0 mg/kg/day	DiBP100 (mg/kg/day)	DiBP300 (mg/kg/day)	DiBP600 (mg/kg/day)	DiBP900 (mg/kg/day)
No. of dams on GD 18	5	8	5	5	5
No. of dams with whole litter loss	0	0	0	0	1
Maternal body wt. at GD 8 (g)	300.1 ± 1.7	294.8 ± 12.0	290.6 ± 16.6	299.3 ± 2.2	301.7 ± 3.1
Maternal body wt. at GD 18 (g)	373.5 ± 4.3	374.7 ± 6.7	366.8 ± 10.0	347.2 ± 6.5*	344.6 ± 6.2*
Maternal body wt. gain (g)	73.4 ± 4.3	79.9 ± 10.2	76.2 ± 14.8	47.9 ± 5.6	42.9 ± 5.5*
No. of implantationsa	13.7 ± 0.9 (3)	14.8 ± 0.8 (4)	16.0 ± 1.0 (3)	12.7 ± 1.2 (3)	13.3 ± 0.9
No. of live fetuses	13.3 ± 0.7 (3)	13.5 ± 0.5 (4)	15.3 ± 1.5 (3)	9.3 ± 2.6 (3)	5.0 ± 3.6* (3)
Total resorptions	0.2 ± 0.2	1.0 ± 0.5	0.4 ± 0.4	2.0 ± 1.1	7.8 ± 2.5*
Fetal mortality (%)b	1.3 ± 1.3 (3)	4.6 ± 2.6 (4)	2.7 ± 2.7 (3)	17.2 ± 10.4	59.0 ± 30.2*

Note. Number in parentheses indicates number of litters in the analysis when different from number of dams on GD 18. Values in bold are significantly different than controls due to treatment.

*p < 0.05 versus control. Values are means ± standard errors.

a

Implantations = live fetuses + total resorptions.

b

Fetal mortality = resorptions/implantations.

TABLE 3

Maternal Body Weight Gain, Litter Size, Resorptions, and Fetal Mortality of SD Rat Dams Administered a Dose Range of BBP and DEP from GD 8 to GD 18

	0 mg/kg/day	BBP100 (mg/kg/day)	BBP300 (mg/kg/day)	BBP600 (mg/kg/day)	BBP900 (mg/kg/day)	DEP100 (mg/kg/day)	DEP300 (mg/kg/day)	DEP600 (mg/kg/day)	DEP900 (mg/kg/day)
No. of dams on GD 18	9	4	5	2	2	4	5	3	3
No. of dams with whole litter loss	0	0	0	0	0	0	0	0	0
Maternal body wt. at GD 8 (g)	293.2 ± 9.1	287.9 ± 7.4	293.1 ± 3.1	294.1 ± 6.3	305.9 ± 4.8	290.3 ± 7.6	295.8 ± 7.7	301.1 ± 9.2	290.6 ± 4.9
Maternal body wt. at GD 18 (g)	372.8 ± 8.7	358.0 ± 11.3	351.7 ± 7.0	337.9 ± 7.9	349.2 ± 3.4	376.1 ± 6.2	363.9 ± 6.9	359.3 ± 12.2	365.6 ± 3.6
Maternal body wt. gain (g)	79.6 ± 8.5	70.2 ± 4.7	58.6 ± 4.9*	43.9 ± 1.7*	43.4 ± 1.4*	85.8 ± 4.1	68.2 ± 6.5	58.2 ± 3.3a	75.0 ± 4.7
No. of implantationsb	14.1 ± 0.7 (8)	15.3 ± 0.5	15.3 ± 0.3 (3)	12.0 ± 1.0	12.0 (1)	14.8 ± 0.6	14.5 ± 0.3 (4)	14.7 ± 1.2	16.0 ± 1.0
No. of live fetuses	13.8 ± 0.8 (8)	15.3 ± 1.5	14.7 ± 0.3 (3)	10.5 ± 0.5*	5.0 ± 3.0*	14.8 ± 6.2	14.5 ± 0.3 (4)	13.0 ± 1.2	16.0 ± 1.0a
Total resorptions	0.4 ± 0.3 (8)	0.0 ± 0.0	0.4 ± 0.2	1.5 ± 0.5*	4.0* (1)	0.0 ± 0.0	0.0 ± 0.0	1.7 ± 0.9*	0.0 ± 0.0
Fetal mortality (%)c	2.9 ± 2.1 (8)	0.0 ± 0.0	2.2 ± 1.6	12.2 ± 3.1*	33.3* (1)	0.0 ± 0.0	0.0 ± 0.0 (4)	11.1 ± 6.2*	0.0 ± 0.0

Note. Number in parentheses indicates number of litters in the analysis when different from number of dams on GD 18. Values in bold are significantly different than controls due to treatment.

*p < 0.05 versus control. Values are means ± standard errors.

a

Implantations = total live fetuses + total resorptions.

b

Fetal mortality = resorptions/implantations.

TABLE 4

Maternal Body Weight Gain, Litter Size, Resorptions, and Fetal Mortality of SD Rat Dams Administered a Dose Range of DBP from GD 8 to GD 18

	Control	DBP33	DBP50	DBP100	DBP300	DBP600
No. of dams on GD 18	3	4	4	4	4	4
No. of dams with whole litter loss	0	0	0	0	0	0
Maternal body wt. at GD 8 (g)	246.1 ± 15.9	245.4 ± 12.9	256.7 ± 12.2	260.4 ± 13.0	263.9 ± 11.4	267.6 ± 13.5
Maternal body wt. at GD 18 (g)	313.0 ± 11.0	316.1 ± 9.5	317.6 ± 8.5	331.1 ± 10.5	330.9 ± 13.4	327.3 ± 7.9
Maternal body wt. gain (g)	66.9 ± 5.6	70.7 ± 4.9	60.9 ± 4.5	70.8 ± 3.3	67.0 ± 6.2	59.7 ± 7.9
No. of implantationsa	14.7 ± 0.7	12.0 ± 0.7	14.5 ± 0.3	14.0 ± 0.4	12.0 ± 3.5 (3)	15.3 ± 0.6
No. of live fetuses	14.3 ± 0.9	12.3 ± 0.3	14.0 ± 0.4	13.5 ± 0.6	11.3 ± 3.2 (3)	13.8 ± 0.5
No. of dead fetuses	0.3 ± 0.3	0.0 ± 0.0	0.0 ± 0.0	0.0 ± 0.0	0.0 ± 0.0 (3)	0.0 ± 0.0
Total resorptions	0.0 ± 0.0	0.3 ± 0.3	0.5 ± 0.3	0.5 ± 0.5	0.7 ± 0.7 (3)	1.5 ± 0.5
Fetal mortality (%)b	2.4 ± 2.4	2.6 ± 2.6	3.5 ± 2.0	3.6 ± 3.6	4.2 ± 4.2 (3)	9.6 ± 3.2

Note. Number in parentheses indicates number of litters in the analysis when different from number of dams on GD 18. Values in bold are significantly different than controls due to treatment.

a

Implantations = total live fetuses + total resorptions.

b

Fetal mortality = resorptions/implantations. Values are means ± standard errors.

TABLE 5

Maternal Body Weight Gain, Litter Size, Resorptions, and Fetal Mortality of SD Rat Dams Administered a Dose Range of DEHP from GD 8 to GD 18

	0 mg/kg/day	DEHP100 (mg/kg/day)	DEHP300 (mg/kg/day)	DEHP600 (mg/kg/day)	DEHP900 (mg/kg/day)
No. of dams on GD 18	4	4	4	4	4
No. of dams with whole litter loss	0	0	0	0	0
Maternal body wt. at GD 8 (g)	274.2 ± 1.2	275.2 ± 5.0	278.3 ± 2.7	274.9 ± 3.7	275.3 ± 1.1
Maternal body wt. at GD 18 (g)	328.8 ± 15.0	329.5 ± 4.0	349.5 ± 8.5	321.4 ± 8.7	327.8 ± 2.7
Maternal body wt. gain (g)	54.7 ± 14.1	54.3 ± 0.9	71.2 ± 7.3	46. 5 ± 5.6	52.6 ± 2.1

Note. Data were not recorded on implantation number, resorptions, and fetal mortality for these dams. Values are means ± SE.

One dam in each of the following dose groups was not pregnant: 600 and 900 mg BBP/kg/day, Control from DBP study, 300 mg DiBP/kg/day, and 300 and 600 mg DPP/kg/day. The following dams died or were removed from the study due to dosing errors: two dams each from 600 and 900 mg BBP/kg/day, and one dam from 900 mg DEP/kg/day, two dams from 600 mg DEP/kg/day, two dams from 300 mg DiBP/kg/day, and one Control dam from DPP study.

Several of the phthalates also caused midgestation whole or partial litter loss (Tables 1–5; Fig. 1A). DPP was the most potent of the six phthalates as evidenced by midgestation pregnancy loss leading to 100% fetal mortality at doses of 300, 600, and 900 mg/kg/day (Table 1). Of the dams administered 300–900 mg DPP/kg/day, four of five dams were observed to have vaginal bleeding during midpregnancy, indicative of aborted pregnancies.

FIG. 1.

Percent of mortality (A) and testicular testosterone (testosterone) production expressed as percent of control (B) on GD 18 SD rat fetuses exposed to a dose range of six individual phthalates administered to the dam via gavage on GD 8–18, including BBP, DBP, DEHP, DEP, DiBP, and DPP. Fetal mortality data on DEHP-treated litters were not recorded. Data are litter means ± SE.

DiBP was slightly more potent than BBP at inducing fetal mortality, with both producing effects at 600 and 900 mg/kg/day (Tables 2–3). DiBP induced complete litter loss in one dam at 900 mg/kg/day, and induced greater than 50% resorptions in two dams at 900 mg/kg/day and one dam at 600 mg/kg/day. BBP did not induce complete litter loss at any dose, but did induce 33% fetal mortality at a 900 mg/kg/day. DBP did not significantly increase the rate of fetal death at any dosage levels, however it was only administered up to 600 mg/kg/day (Table 4). The percent fetal mortality induced by DBP at 600 mg/kg/day was similar to the effect of BBP at 600 mg/kg/day (Fig. 1A). There was one dead fetus in a control litter in the DBP study. Although fetal mortality data were not recorded for the DEHP study, it was noted that few, if any, resorptions or dead fetuses were seen at doses up to 900 mg DEHP/kg/day. Given the lack of high dose fetal mortality data for DBP and DEHP, in order to model the effect of the mixture on fetal mortality, we assumed that DBP and DEHP were equipotent to BBP, however this assumption warrants further experimentation to test its validity. Therefore, we used the ED50 and slope parameters for fetal mortality calculated from the BBP data (Table 4) for DBP and DEHP as well as BBP. The ED50s (mg/kg/day) and Hill slopes for fetal mortality are shown in Table 8. It is evident from this table that the model for DPP was not able to provide an accurate determination of a slope because the slope was quite steep and the spacing of our dosing resulted in all-or-none effects. Furthermore for BBP, our ED50 and slope values are based upon a partial curve that only attained 33% mortality at the highest dose administered. Given these limitations, it is not surprising that our prediction of the slope for the mixture is not very accurate. However, we were able to accurately predict the ED50 of the mixture using these data and all the assumptions.

TABLE 6

Testicular Testosterone Production (ng/Testis/3 h) of GD 18 Male SD Rat Fetuses Exposed to a Dose Range of Six Individual Phthalates Administered to the Dam via Gavage on GD 8–18, including DEHP, BBP, DBP, DEP, DiBP, and DPP

0 mg/kg/day		25 mg/kg/day	33 mg/kg/day	50 mg/kg/day	100 mg/kg/day	200 mg/kg/day	300 mg/kg/day	600 mg/kg/day	900 mg/kg/day
BBPa	3.45 ± 0.15 (27, 9)b	na	na	na	3.66 ± 0.26 (12, 4)	na	2.68 ± 0.27 (15, 5)*	1.18 ± 0.24 (6, 2)*	0.34 ± 0.14 (4, 2)*
DBP	5.51 ± 0.37 (9, 3)	na	5.16 ± 0.46 (12, 4)	4.31 ± 0.30 (12, 4)	4.62 ± 0.37 (12, 4)	na	3.62 ± 0.60 (12, 4)*	1.81 ± 0.34 (12, 4)*	na
DEHP	6.23 ±1.28 (24, 4)	na	na	na	5.11 ± 0.33 (26, 4)	na	3.59 ± 0.12 (27, 4)*	2.53 ± 0.38 (29, 4)*	1.40 ± 0.25 (20, 4)*
DEPa	3.45 ± 0.15(27, 9)	na	na	na	3.71 ± 0.17 (12, 4)	na	3.03 ± 0.16 (15, 5)	3.73 ± 0.39 (9, 3)	3.22 ± 0.15 (9, 3)
DiBP	5.70 ± 0.13 (15, 5)	na	na	na	5.44 ± 0.19 (24, 8)	na	3.40 ± 0.28 (15, 5)*	2.31 ± 0.35 (15, 5)*	2.09 ± 0.91 (6, 2)*
DPP	5.89 ± 0.36 (10, 5)	5.30 ± 0.60 (12, 4)	na	5.00 ± 0.45 (12, 4)	3.26 ± 0.30 (18, 6)*	2.26 ± 0.37 (12, 4)*	No fetuses	No fetuses	No fetuses

Note. Values in bold are significantly different than controls due to treatment. na = no data available at these dosage levels.

*p < 0.01 versus control. Values are litter means ± standard errors.

a

BBP and DEP dose ranges were dosed in the same blocks, thus they share the same controls and the values are lower because the incubation period was 2 h rather than 3 h used for the other phthalate diesters.

b

Number of individual fetuses and litters are in parentheses.

TABLE 7

Phthalate Mixture Study

Dose as total phthalates (mg/kg/day)	0	65	130	260	520	780	1040	1300
Dose as % of top dose	0%	5%	10%	20%	40%	60%	80%	100%
No. of dams on GD 18	8	4	4	7	7	5	2	4
No. dams: whole litter loss	0	0	0	0	0	1	0	2
Maternal body wt. GD 8 (g)	250.4 ± 5.8	255.2 ± 1.7	267.2 ± 1.1	261.4 ± 5.0	267.2 ± 4.1*	274.9 ± 9.3*	259.1 ± 21.0**	273.5 ± 10.7*
Maternal body wt. GD 18 (g)	311.4 ± 10.7	311.4 ± 7.3	336.1 ± 2.6	326.4 ± 5.3	308.8 ± 6.8	317.5 ± 4.7	294.6 ± 29.4	290.0 ± 12.3
Maternal body wt. gain (g)	61.0 ± 8.3	56.2 ± 7.3	68.9 ± 3.7	65.0 ± 3.0	41.6 ± 7.0*	42.6 ± 6.9**	35.5 ± 8.4*	16.5 ± 3.7*
No. of implantationsa	11.6 ± 1.4	13.3 ± 0.3	13.5 ± 0.5	14.3 ± 0.4	13.0 ± 0.7	13.4 ± 0.3	14.0 ± 2.0	12.0 ± 0.8
No. of live fetuses	11.4 ± 1.4	13.0 ± 0.4	13.3 ± 0.5	13.6 ± 0.4	8.0 ± 1.8**	4.6 ± 2.2*	10.5 ± 3.5	3.0 ± 1.9*
Total resorptions	0.1 ± 0.1	0.3 ± 0.3	0.3 ± 0.3	0.7 ± 0.4	5.0 ± 1.8*	8.6 ± 2.4*	3.5 ± 1.5*	9.0 ± 1.7*
Fetal mortality (%)b	1.0 ± 1.0	1.9 ± 1.9	1.8 ± 1.8	4.8 ± 2.8	38.2 ± 12.5*	65.8 ± 16.5*	27.1 ± 14.6*	76.2 ± 15.8*
GD 18 testosterone production (ng/testis/3 h)	5.78 ± 0.38	5.82 ± 0.40	5.30 ± 0.60	4.32 ± 0.28*	2.55 ± 0.21*	1.57 ± 0.58*	1.65 ± 0.22*	1.40 ± 0.02*

Note. Maternal body weight gain, litter size, resorptions, and fetal mortality of SD rat dams administered a dose range of a combination dose of five phthalates from GD 8 to GD 18. The top dose of the combination dose was administered at 1300 mg of the five phthalates/kg/day, including 100 mg/kg/day of DPP and 300 mg/kg/day of each of the following: DEHP, DBP, BBP, and DiBP.

*p < 0.05 versus control. Values in bold are significantly different than controls due to treatment. Values are litter means ± standard errors.

**p = 0.07–0.08 versus control.

a

Implantations = total live fetuses + total resorptions.

b

Fetal mortality = resorptions/implantations.

TABLE 8

Logistic Regression Analysis Results of the Effects on Fetal Mortality by Three Phthalates Administered to the Dam via Gavage on GD 8–18, including BBP, DiBP, and DPP

		DPP	DiBP	BBP
Best-fit values	Bottom	= 0.0	= 0.0	= 0.0
	Top	= 100.0	= 100.0	= 100.0
	Hill slope	27.79	2.87	2.99
	Hill slope SE	615.3	1.629	0.7277
	Hill slope 95% CL	−1235 to 1290	−0.5528 to 6.293	1.462–4.519
	ED50	221.2	527.9	1145
	EC50 95% CL	2.276–21502	354.3–786.6	905.0–1449
Goodness of fit	df	27	18	18
	R2	96.1%	33.9%	71.8%

Note. Fetal mortality data for DBP and DEHP could not be modeled due to insufficient data, thus DBP and DEHP were assumed to have the same potency as BBP.

Administration of five of the six phthalate esters reduced fetal testicular testosterone production in the current study (Table 6). BBP, DBP, DEHP, and DiBP significantly reduced testosterone production at doses of 300 mg/kg/day or higher (p < 0.001), and DPP reduced testosterone production at doses as low as 100 mg/kg/day (p < 0.005) compared with controls. In contrast, DEP did not affect testosterone production even at dose levels up to 900 mg/kg/day. The phthalates used in the current study tended to reduced fetal testosterone production at dosage levels about one-half to one-third of that required to increase fetal mortality, however more data are needed to more accurately define the dose-response relationships among these phthalates and fetal mortality and pregnancy loss.

A comparison of the ED50s from the sigmoidal regression models indicates that BBP, DBP, DEHP, and DiBP were equipotent (the average and standard error of the ED50s of these four phthalates was 440 ± 16 mg/kg/day), DPP was about threefold more potent (ED50 = 130 mg/kg/day) and DEP had no effect on fetal testosterone production (Fig. 1B).

The ED50s and Hill slopes for fetal testosterone production are shown in table 9. The ED50 and Hill slope calculations for DEP are not included here because DEP did not significantly reduce fetal testosterone even at the highest doses tested in this study (Table 6). Many of the testes collected from the BBP or DiBP fetuses at doses of 600 and 900 mg/kg/day were smaller, mucilaginous, and/or located higher in the abdominal cavity.

TABLE 9

Logistic Regression Analysis Results of the Effects on Fetal Testicular Testosterone Production by the Five Antiandrogenic Individual Phthalate Esters Administered to the Dam via Gavage on GD 8–18, including BBP, DBP, DEHP, DEP, DiBP, and DPP

Logistic regression		BBP	DPP	DEHP	DBP	DiBP
Best-fit values	Bottom	= 0.0	= 0.0	= 0.0	= 0.0	= 0.0
	Top	= 3.450	= 5.890	= 6.230	= 5.510	= 5.700
	Hill slope	−3.016	−1.413	−1.204	−1.019	−1.465
	Hill slope SE	0.6638	0.3167	0.3396	0.2319	0.2364
	Hill slope 95% CI	−4.401 to −1.631	−2.072 to −0.7544	−1.918 to −0.4909	−1.501 to −0.5365	−1.954 to −0.9760
	ED50	464	130.7	383.4	398.5	465.6
	ED50 95% CI	384.1–560.5	99.80–171.2	258.0–569.7	258.4 to 614.4	381.8 to 567.9
Goodness of fit	R2%	85%	74%	71%	67%	84%
	df	20	21	18	21	23

Analysis of the components of variance of fetal testis testosterone production using PROC MIXED for each of the individual phthalates indicated that the ratio of the variance among pups within litter to the variance between litters was near unity. Applying Cox's Rule of Thumb (Bergerud, 1995) to these results, we determined that measuring testosterone production from three to four pups per litter provides optimal statistical power, and that beyond this number, sampling more pups per litter does not increase the power of the study to detect treatment-related effects on testosterone production.

Because it has been reported that DBP reduces extracted fetal testis testosterone levels, a subset of the fetal testes collected from DBP-treated dams were extracted for testicular testosterone levels and measured by RIA to compare to the testosterone production results of the incubated testes (data not shown; see Supplemental Fig.). Testosterone extracted from the testes is an assessment of the total hormone content of the testis at the time of tissue collection, whereas ex vivo testosterone production is a functional assay that assesses the ability of the testis to produce the hormone over the 3-h incubation period. Although the extracted testicular testosterone levels declined in a dose-dependent fashion with DBP exposure, the extracted testosterone data did not reach statistical significance suggesting that this method was less sensitive and less precise a measure of phthalate inhibition of steroidogenesis than testosterone production of incubated fetal testes at this gestational age.

Phthalate Mixture Study

The phthalate mixture significantly reduced maternal body weight gain at 40% of the top dose and greater (Table 7; Fig. 2A). The phthalate mixture treatment did not induce illness (i.e., the dams appeared healthy) or death in the dams. One dam in each of the following dose groups was not pregnant: 260 mg/kg/day (20%), 780 mg/kg/day (60%), and 1040 mg/kg/day (80% of top dose) in phthalate mixture study. The following dams died or were removed from the study due to dosing errors: one dam each from 520 mg/kg/day (40%) and 1040 mg/kg/day (80% top mixture dose) and two dams from 780 mg/kg/day (60% of top dose) dose groups. There was one dead fetus in one litter of the 780 mg/kg/day dose group (60% of top dose).

FIG. 2.

Observed fetal mortality (A) and testicular testosterone production (B) with their respective predicted dose addition response in GD 18 SD rat fetuses exposed to a combination of five phthalates administered to the dam via gavage on GD 8–18. The top dose of the mixture contained a total of 1300 mg phthalate/kg/day, including 100 mg DPP/kg/day and 300 mg/kg/day of each of the following: BBP, DBP, DEHP, and DiBP. Data are litter means ± SE.

Administration of the mixture of five antiandrogenic phthalates significantly increased fetal mortality at a total phthalate dose of 520 mg/kg/day (40% of top dose, which contains BBP, DBP, DEHP, and DiBP at 120 mg/kg/day per chemical and 40 mg DPP/kg/day) and above (Fig. 2A, Table 7), which were comparable to doses that reduced maternal body weight gain (Table 7). Whole litter loss occurred at midpregnancy in one of five dams in the 780 mg total phthalates/kg/day and two of four dams in the 1300 mg total phthalates/kg/day group.

None of the predicted values differed significantly from the fetal mortality observed at any dosage level of the mixture, as assessed using 95% confidence limits of the observed data. The ED50 value for the increase in fetal mortality observed with the phthalate mixture study was accurately predicted by the dose addition model (ED50, observed = 735 mg/kg/day [R² = 58%] vs. ED50, predicted = 720 mg/kg/day), whereas the predicted slope was steeper than that generated from the observed data (Hill slope, observed = 1.87 vs. predicted = 8.0; Fig. 2A). Additional individual phthalate and mixture data are needed to evaluate the utility of the dose addition model to predict the outcome of a mixture of phthalates on pregnancy loss and fetal mortality, especially because the slopes of the individual phthalates used to generate the model are not parallel with one another and two of the three individual phthalate models were quite variable, having large confidence limits (DPP slope) or low R² values (DiBP). In addition, when the observed fetal mortality data (41 df) were fit with the logistic regression parameters predicted by dose addition modeling (the four parameters of the model were constrained to dose addition model values), the R² for the model was only 28% as compared with an R² of 58% attained by logistic regression analyses of the data with no constraints on the slope or ED50 values.

Administration of the mixture of five antiandrogenic phthalates reduced fetal testicular testosterone production at doses of 260 mg/kg/day (20% of top dose, which contains BBP, DBP, DEHP, and DiBP at 60 mg/kg/day per chemical and 20 mg DPP/kg/day) and higher (Fig. 2B; Table 7). Dose addition modeling based upon the data from the five individual phthalates provided an accurate prediction of the observed effect of the mixture on fetal testosterone production; the dose addition predictions fell within the confidence limits calculated for the observed effects of the mixture at every dosage level. The ED50 of the observed data was 480 mg/kg of total phthalates versus dose addition ED50 predicted = 360 mg/kg total phthalates; Figure 2B. In addition, the Hill slope of the observed data closely matched that of the dose addition model (Hill slope: observed = −1.70 vs. predicted = −1.62). Furthermore, when the observed testosterone production values (38 df) were fit with the logistic regression parameters predicted by dose addition modeling (the four parameters of the model were constrained to dose addition model values), the R² for the model was 76% (38 df) which is very similar to the R² of 81% attained by logistic regression analyses of the data with no constraints on the slope or ED50 values (36 df). Thus, the phthalate mixture significantly reduced fetal testosterone production at one-half to one-third the dosage levels of those that significantly increased fetal mortality (Fig. 3). In addition, there was evidence of phthalate-induced underdevelopment of the testes in the highest mixture dose groups. In one litter of the 1300 mg total phthalates/kg/day (100% of top dose), the testes were located high in the abdominal cavity and appeared to be attached to the kidneys via a ligament. In one litter of the 1040 mg total phthalates/kg/day dose (80%), the testes were located high in the abdominal cavity and appeared underdeveloped, and one litter of the 789 mg total phthalates/kg/day group (60%) had several males with mucilaginous testes.

FIG. 3.

Observed percent fetal mortality and reduction in testicular testosterone production of GD 18 SD rat fetuses exposed to a combination of five phthalates administered to the dam via gavage on GD 8–18. The top dose of the mixture contained a total of 1300 mg phthalates/kg/day including 100 mg DPP/kg/day and 300 mg/kg/day of each of the following: BBP, DBP, DEHP, and DiBP. Data are litter means.

DISCUSSION

The current study confirmed that the ability of individual phthalates to inhibit fetal testosterone production is dependent on their chemical structure. Phthalate diesters with four to six carbons in length and in the ortho position (i.e., BBP, DBP, DEHP, DiBP, and DPP) inhibited fetal testosterone production, whereas DEP was without effect. This structure activity relationship was first identified as phthalate-induced testicular atrophy in a 4-day exposure of immature male rat (Foster et al., 1980) and has since been verified by gestational exposure studies in the rat (Gray et al., 2000; Liu et al., 2005). The phthalates BBP, DBP, and DiBP were of equivalent potency to DEHP at reducing fetal testosterone production, whereas DPP was three times as potent as DEHP. In vivo studies of gestationally exposed male rats have demonstrated that BBP, DBP, and DEHP are relatively equipotent at inducing reproductive malformations and decreasing androgen-dependent organ weights (Gray et al., 1999, 2000; Mylchreest et al., 1998). Although there are currently no published reports on the effects of gestation exposure to DiBP or DPP on male reproductive development in adult rat offspring, we predict that these phthalates will induce similar effects as BBP, DBP, and DEHP.

Using data for the individual phthalates, we designed a mixture of the five antiandrogenic phthalates and accurately predicted the effects of the combination dose to inhibit fetal testosterone production using the dose addition model. The dose addition model was selected because it assumes that the chemicals in the mixture share a common mechanism of toxicity. Other studies from our laboratory have demonstrated the cumulative, dose-additive effects of combinations of phthalates. In utero exposure to a combination dose of DBP and DEHP (500 mg/kg/day per chemical from GD 14 to GD 18), two antiandrogenic phthalates with different active metabolites, caused cumulative, largely dose-additive effects on reproductive tract development due to inhibition of fetal endocrinology (testosterone and insl3) (Howdeshell et al., 2007). Cumulative effects on male reproductive development have also been reported for a combination of two antiandrogenic phthalates with similar metabolites (Gray et al., 2006; Gray et al., in preparation). We have also reported cumulative dose-additive effects of phthalates in combination with other antiandrogenic chemicals, which disrupt the androgen signaling pathway via diverse mechanisms of toxicity (Hotchkiss et al., 2004; Rider et al., 2008). Some of the toxicants lower testosterone levels in the tissues, whereas others prevent testosterone from activating the androgen receptor, reduced expression of androgen-dependent genes and altered differentiation of the tissues. Therefore, antiandrogenic phthalates work in combination with one another to disrupt androgenic-dependent reproductive development when administered during sexual differentiation.

Due to the critical role that testosterone plays during sexual differentiation of the male fetus in all mammals including rodents and human, fetal testicular testosterone levels were selected as the critical endpoint for establishing a NOAEL in the current USEPA risk assessment on DBP (USEPA, 2006). Fetal testicular testosterone levels, obtained by extracting the testes, are reported to decrease with gestation exposure to DBP (Lehmann et al., 2004; Parks et al., 2000) and DiBP (Borch et al., 2006). We report here that fetal testosterone production was significantly inhibited by DBP at doses of 300 mg/kg/day and higher. However, one should keep in mind that the current study was designed to determine the slope and ED50 values of the individual phthalates and a mixture of phthalates and not to detect NOAELS or low observable adverse effect levels. Good estimates of the logistic regression parameters requires a focus on the dose groups producing a 20–80% effect, whereas a NOAEL study would emphasize the low dose range of the dose-response curve and would require much larger samples sizes in the low dose range than we used herein. It will be important that future gestational studies with phthalates determine how much of an inhibition of fetal testosterone production is necessary to induce reproductive malformations and permanently reduce androgen-dependent organ weights. By defining this “point of departure” for fetal testosterone production, scientists will be able to more accurately predict the permanent, latent effects of in utero exposure to phthalates (either individually or in combination) on male reproductive development from the fetal endocrine data and thus reduce animal use and the length of the study. Execution of a fetal study takes a few weeks, whereas a postnatal study that examines anogenital distance, areolae/nipples, organ weights, malformations and histopathology in the offspring through adulthood and takes considerable time and resources.

Decreased maternal body weight gain and increased fetal mortality were induced by gestation exposure to several of the individual phthalates and the phthalate mixture. Such effects are commonly reported for exposure to high doses of phthalate esters. In a developmental toxicology study in SD rats, DiBP (administered via gavage on GD 6–20) decreased maternal body weight gain (at 1000 mg/kg/day), increased postimplantation fetal loss (at 750 and 1000 mg/kg/day), and reduced number of live fetuses at birth (Saillenfait et al., 2006). Previous studies in our laboratory with SD rats observed that gestational exposure to DEHP (at 750 mg/kg/day from GD 14 through pregnancy) reduced maternal body weight gain; however it did not induce fetal mortality (Gray et al., 2000; Parks et al., 2000). Exposure to DEHP from GD 3 through pregnancy reduced maternal body weight gain and induced pregnancy loss in SD rats at doses of 750 and 1500 mg/kg/day leading to a reduction in litter size at birth with 1500 mg DEHP/kg/day (Moore et al., 2001). Gestational exposure to DBP induced postimplantation loss and reduced litter size at birth in SD and Wistar rat dams at 750–1500 mg DBP/kg/day (Ema et al., 2000; Mylchreest et al., 1998). In a study that administered DBP by oral gavage from weaning throughout life decreased the litter sizes of Long Evans rat dams at similar oral doses that caused a reduction in maternal serum levels of progesterone at GD 13 (at oral doses of 500–1000 mg/kg/day), indicating that DBP was disrupting maternal ovarian steroidogenesis at midpregnancy (Gray et al., 2006). Therefore, studies in rats suggest that pregnancy maintenance follows the same structure activity relationship of phthalate diesters on inhibition of fetal steroidogenesis, possibly by inhibition of maternal progesterone levels; however, this hypothesis needs to be rigorously tested. In addition, future experiments should examine the effect of the mode of action by which the phthalates disrupt pregnancy during midgestation in the rat to determine if this is a direct or indirect effect on ovarian steroidogenesis.

Humans also are exposed to multiple phthalates, thus human health could also be adversely impacted by the cumulative effects of antiandrogenic phthalates (Blount et al., 2000; Latini, et al., 2003; Mortensen, et al., 2005; Silva et al. 2004a, 2005). For example, the human fetus is exposed to combinations of monoester metabolites, which are known to disrupt reproductive tract development in rats; the majority of the exposures being below the levels seen in the amniotic fluid of affected rats. In a study of second trimester human pregnancies, monoester metabolites of the antiandrogenic phthalates DBP (monobutyl phthalate, MBP) and DEHP (monoethylhexyl phthalate, MEHP) were detected in 93 and 24%, respectively, of the amniotic fluid samples tested (n = 54; Silva et al., 2004b). In 2% of these samples, the levels of MBP and MEHP present in human amniotic fluid were only 5- and 24-fold less, respectively, than levels of the same metabolites detected in rat amniotic fluid following oral exposure of the dam to doses affecting reproductive development in male rat offspring (Calafat et al., 2006). Wolff et al. (2008) reported that 95% women (382/404) in their third trimester of pregnancy had urine samples detecting phthalate metabolites; a least 91% of the phthalate-positive women had detectable levels of monoester metabolites of antiandrogenic phthalates BBP, DBP, DEHP, and DiBP. In some countries, the estimated daily intake of DBP or DEHP actually exceed the tolerable daily intake or reference dose identified as safe by regulatory agencies in the European Union or the United States (Chen et al., 2008; Wittassek and Angerer, 2007). A study of 239 children (2–14 years of age) in Germany, reported that approximately one-fifth of the study population had a cumulative daily intake DBP and DEHP that exceeded the regulatory limit (Wittassek and Angerer, 2007). Finally, some epidemiological studies that suggest that cumulative phthalate exposure could be impacting human reproductive development or normal hormone balance. Swan et al. (2005) reported an association between increasing phthalate metabolite levels in pregnant women (including metabolites of BBP, DBP, DEHP, and DiBP) and a reduction in the anogenital distance of their male infants; a reduction in this androgen-dependent trait is reported in male offspring of rat dams treated during pregnancy and is correlated with an increase in abnormal reproductive development observed in adulthood in rats (Hotchkiss et al., 2004).

In conclusion, we demonstrated that antiandrogenic phthalates work together to exert cumulative, dose additive, inhibitory effects on fetal steroidogenesis in the rat. The reproductive toxicity of the phthalate esters is determined by their chemical structure, and this structure activity relationship is true for phthalate-induced inhibition of fetal endocrinology leading to disrupted reproductive tract development in male rats, and possibly for inhibition of maternal steroidogenesis leading to impaired pregnancy maintenance. As, the human fetus is exposed to multiple phthalates in the womb, and because sexual differentiation in male human fetus, like the rat, is critically dependent on appropriate levels of testosterone, there is a credible rationale for the use of fetal testicular testosterone inhibition rat data in the USEPA risk assessment for DBP (and other phthalates) as well as the inclusion of cumulative risk to multiple antiandrogenic phthalate exposure.