4.2 Experimental Animal Data
Lovekamp.and.Davis.(159), supported by NIEHS, evaluated the effects of MEHP on F344 rat granulosa cells in culture. Four-week-old animals were stimulated with diethylstilbestrol or with pregnant mare serum gonadotropin (PMSG) following which ovaries were removed and granulosa cells harvested.
Cells were cultured in the presence of 500 nM testosterone and 200 ng/mL FSH. MEHP [purity.not.
specified].was added to cultures at concentrations of 0 – 200 µM [0.–.55.6.mg/L] in a comparison with other phthalate monoesters and with Wy-14,643, another peroxisome proliferator. After 48 hours, estradiol and progesterone were measured in media using commercial RIA kits. RNA was extracted from cells and RT-PCR performed to quantify mRNA for aromatase and for cholesterol side-chain cleavage enzyme. Aromatase protein was quantified by Western blot. Statistical comparisons were made using the Student t test or ANOVA, followed by least significant difference test.
A comparison with other phthalate monoesters showed a decrease in estradiol in the medium (con-trolled for protein content) with MEHP 100 or 200 µM [27.8.or.55.6.mg/L] but not with comparable molar concentrations of monomethyl, -ethyl, -propyl, -butyl, -pentyl, or -hexyl phthalate. Monopentyl phthalate was associated with a decrease in estradiol production at 400 µM. mRNA for aromatase was estimated in media after culture with MEHP 0, 25, 50, or 100 µM [0,.7.0,.13.9,.or.27.8.mg/L].
Graphically, there appeared to be a concentration-dependent decrease in estradiol concentration and in aromatase mRNA; pair-wise comparisons with control were statistically significant for the 100 µM concentration for estradiol and for the 50 and 100 µM concentrations for aromatase mRNA. The peroxisome proliferator Wy-14,643 also decreased estradiol and aromatase mRNA. Cholesterol side-chain cleavage enzyme was not altered by MEHP, suggesting specificity of the effect on aromatase mRNA. Aromatase protein was decreased by MEHP at concentrations of 100 and 200 µM [27.8.and.
55.6.mg/L]. In a final experiment, granulosa cells were incubated for 48 hours with MEHP 0 or 200 µM [0.or.56.mg/L] and 8-bromo-cyclic adenosine monophosphate (8-Br-cAMP) added for the last 24 hours. mRNA for aromatase and cholesterol side-chain cleavage enzyme and medium progesterone levels were increased by 8-Br-cAMP in the absence of MEHP. In the presence of MEHP, mRNA for aromatase and medium estradiol levels were suppressed, but there was no suppression of P450 side-chain cleavage enzyme mRNA or progesterone. The authors interpreted their results as consistent with transcriptional suppression of aromatase by MEHP independent of the FSH-cAMP pathway.
They proposed a PPAR pathway as a candidate mechanism for MEHP suppression of granulosa cell steroidogenic function.
Strengths/Weaknesses: The strengths of this study include appropriate technical proficiency and appro-priate study controls and statistics, as well as examination of a set of structurally diverse phthalates.
The Panel has confidence in the veracity of these results. This study identified a probable main point of interference with steroid production, showed that it was most affected by MEHP and not other phthalates (thus identifying MEHP as the main concern), and proposed a possible mechanism by which this interference might occur (PPAR mediation). This in vitro study comes after several in vivo
studies that demonstrated low-estradiol-related changes in female rats, so relevance is another strength.
Weakness includes examining cells from only 1 species and using MEHP of uncertain purity.
Utility (Adequacy) for CERHR Evaluation Process:.The data point towards a subcellular site of action.
Their utility in risk assessment would come in helping to identify vulnerable species as those having these target enzymes. The in vitro data are presumed relevant for in vivo protection when circulating blood levels of MEHP (or even better, tissue levels) are known.
Lovekamp-Swan.et.al..(160), supported by NIEHS, evaluated the interaction of MEHP with PPAR pathways as a mechanism for modifying ovarian steroidogenic functions. Granulosa cells were harvested from 4-week-old Fisher rats 24 hours after injection of PMSG. Cells were cultured for 48 hours with 500 nM testosterone and 200 ng/mL FSH with or without MEHP [purity.not.specified] 50 µM [13.9.
mg/L]. Culture media were assayed for estradiol using a commercial RIA kit. RNA was extracted from cells and specific RT-PCR probes used to amplify mRNA for quantification. The mRNAs were chosen to reflect the activity of genes involved in the PPAR pathways, steroidogenesis, or phthalate toxicity. The mRNAs included aromatase, 17b-hydroxysteroid dehydrogenase IV (which metabolizes estradiol to estrone), cholesterol side-chain cleavage enzyme, the aryl hydrocarbon (Ah) receptor, cytochrome P450 1B1 (CYP1B1), epoxide hydrolase, and heart-fatty acid binding protein (H-FABP, which is associated with luteal transformation of granulosa cells). mRNA results were normalized to glyceraldehyde 3-phosphate dehydrogenase as an internal control.
To test the hypothesis that MEHP inhibition of granulosa cell aromatase is mediated through different PPAR isoforms, the experiments included the coadministration with MEHP of troglitazone (a PPARγ ligand), selective PPARα and γ agonists, a selective PPARγ antagonist, a selective retinoic acid X receptor (RXR) agonist (because a PPARγ:RXR heterodimer is believed to decrease aromatase), 9-cis-retinoic acid (which binds both RXR and the retinoic acid receptor, RAR), and 15-deoxy-Δ12,14-prostaglandin J2 (a PPAR activator). Statistical comparisons were made using the Student t test or ANOVA, followed by the least significant difference test.
MEHP in culture reduced aromatase mRNA by >40% compared to control. The addition of a PPARγ antagonist partially reversed the decrease in aromatase mRNA. PPARα and γ agonists reduced aromatase mRNA to an extent similar to that after treatment with MEHP. To test the hypothesis that MEHP-activated PPARγ plus MEHP-activated RXR results in aromatase suppression mediated by the PPARγ:RXR heterodimer, cells were treated with MEHP plus RXR or RAR ligands. All treatments decreased aromatase, with significant additional suppression by MEHP when it was added with 9-cis-retinoic acid. To demonstrate that MEHP activity could be mediated through PPARα, MEHP treatment was shown to increase mRNA for 17b-hydroxysteroid dehydrogenase IV, which is inducible by PPARα. The addition of a PPARγ antagonist did not alter the MEHP induction of 17b-hydroxysteroid dehydrogenase IV, suggesting MEHP activation of the α-isoform of PPAR. Both MEHP and a selective PPARα agonist, but not a selective PPARγ agonist, increased the expression of Ah receptor, CYP1B1, and epoxide hydrolase. Cholesterol side-chain cleavage enzyme was not induced by MEHP or by either PPAR isoform agonist. H-FABP was induced by MEHP and by each specific PPAR isoform agonist.
The authors concluded that MEHP effects on granulosa cell gene expression, which would serve to decrease estrogen production, were mediated by both PPAR pathways.
Strengths/Weaknesses:.These data further dissect the molecular pathways by which MEHP reduces estradiol production in granulosa cells. The experiments were well-conceived and executed, used appropriate statistics, and the Expert Panel has confidence in the veracity of these studies. Possible weaknesses for the evaluation process include limited species evaluation, and relying on the presump-tion that these in vitro data report processes that are relevant and active in vivo.
Utility (Adequacy) for CERHR Evaluation Process:.These in vitro data are adequate for the evaluation process and are presumed relevant for in vivo biology. An additional assumption implicit in these experi-ments is that immature stimulated rodent granulosa cells are appropriate models for human granulosa cells, and that the agonists and antagonists employed are specific for their intended targets. The data are uniquely valuable in describing a mechanism of toxicity that may be tested in other species, and that is presumed relevant for other species sharing a similar biochemistry.
Anas.et.al..(161), supported by the Japanese Ministry of Agriculture, Forestry, and Fisheries, evaluated the effect of MEHP on in vitro maturation of bovine oocytes. Cumulus-oocyte complexes were obtained by aspirating 2 – 5 mm follicles from slaughterhouse beef ovaries. Oocytes with unexpanded cumulus layers were selected for study. In experiment 1 (n = 91 – 99 oocytes/group), cumulus – oocyte complexes were cultured for 24 hours with MEHP (90% purity) at 0, 10, 25, 50, 75, or 100 µM [0,.2.8,.7.0,.13.9,.
20.9,.or.27.8.mg/L], after which they were evaluated for cumulus expansion and for stage of oocyte maturation. [Maturation.was.assessed.in.ethanol-fixed.orcein-stained.oocytes.using.unspecified.
criteria.] In experiment 2 (n = 123 – 131 oocytes/group), denuded oocytes were cultured with MEHP at the same concentrations with the addition of a 5 µM [1.4.mg/L].concentration level, followed by assessment of oocyte maturation stage. In experiment 3, cumulus-oocyte complexes were cultured for 24 hours with MEHP 0, 50, or 100 µM [0,.13.9,.or.27.8.mg/L], following which some oocytes were evaluated for maturational stage. Other MEHP-treated oocyte cultures were continued in MEHP-free culture medium for an additional 24 hours or were continued at their original MEHP concentration (50 or 100 µM) for an additional 24 hours. The treatment groups (n = 124 – 135 oocytes/group) consisted of MEHP exposure for 24 hours, MEHP exposure for 24 hours followed by 24-hour “recovery,” and MEHP exposure for 48 hours. Statistical analysis was by ANOVA and Fisher protected least significance test. Experiments 1 and 2 showed a concentration-dependent decrease in progression through oocyte maturation stages, with a significant decrease in oocytes reaching metaphase-II beginning at MEHP 25 µM [7.0.mg/L] for cumulus-oocyte complex culture and at 10 µM [2.8.mg/mL] for denuded oocytes (Figure 6). Cumulus expansion was not impaired in experiment 1 [suggesting.no.effect.of.MEHP.on.
granulosa.cell.division.and.differentiation]. In experiment 3, impairment of progression to metaphase II was seen with the initial 24-hour culture with MEHP, as expected. Culture in MEHP-free medium permitted the progression of oocyte maturation to metaphase II in 64.5 – 71.1% of oocytes, although the proportion of oocytes reaching metaphase II did not recover completely to control levels (83.2% at 24 hours). The recovery of maturation ability suggested that the MEHP-associated decrease in oocytes reaching metaphase II was not likely to be due to nonspecific cytotoxic or lethal effects. When oocytes were cultured in MEHP 50 µM for 48 hours, the proportion reaching metaphase II was higher at 48 hours (41.6%) than at 24 hours (26.9%), suggesting that MEHP delayed maturation rather than preventing maturation altogether.
Strengths/Weaknesses:.The strengths of this study include very large sample sizes, reasonable statistics, a useful study design that included recovery, and a novel approach. It would have been more valuable
to have included additional measures of cytotoxicity and perhaps to have gone past simple description to the use of specific agonists or antagonists (see Lovekamp et al. (160), above).
Utility (Adequacy) for CERHR Evaluation Process: The data are adequate for the evaluation process and are presumed relevant for humans. They show a direct effect on the ovary, although not a unique effect.
They add little to mechanistic understanding. They do, however, expand the list of affected species.
Figure 6. Percent of Bovine Oocytes Reaching Metaphase II after 24-hour Culture with MEHP
0 20 40 60 80
0 20 40 60 80 100
MEHP Concentration (µM)
J J J
Cumulus-Oocyte Complexes Denuded Oocytes
*Significant difference from 0 ppm group. Drawn from Anas et al. (161).
Sekiguchi.et.al..(162), supported by the Cooperative System for Supporting Priority Research in Japan Science and Technology Corporation, evaluated number of ova ovulated in response to 15 or 30 international units (IU) equine chorionic gonadotropin as an assay of female reproductive toxicity in F344 rats. DEHP was used as an illustrative chemical in their system. The equine chorionic gonadotropin was injected intramuscularly (IM) at 25 days of age. The animals were killed 72 hours later, and uteri and ovaries were removed and weighed. Ova were flushed from the oviducts with saline, denuded with hyaluronidase, and counted under a microscope. Because the maximum number of ova recovered in spontaneously ovulating rats was 11, superovulation was defined as the presence of more than 11 ova in the oviducts. To evaluate the effects of the test chemical, DEHP [purity.not.specified] 0 or 500 mg/kg bw/day in olive oil was given by sc injection daily from 24 to 27 days of age. There were 12 rats treated with 15 IU equine chorionic gonadotropin, 6 of which received DEHP, and 7 rats treated with 30 IU equine chorionic gonadotropin, 3 of which received DEHP. Ovulation was induced in 4 of the 6 rats that received 15 IU equine chorionic gonadotropin in the presence or absence of DEHP treatment. In rats given 30 IU equine chorionic gonadotropin, ovulation occurred in 1 of 3 rats given DEHP and 4 of 4 rats given the vehicle. This difference was described by the authors as not statistically significant, possibly due to the small number of animals [statistical.methods.not.described;.P.=.0.14,.Fisher.exact.test.by.
CERHR]. None of the rats treated with DEHP demonstrated superovulation compared to 2 of 4 rats given 15 IU equine chorionic gonadotropin and 1 of 4 rats given 30 IU equine chorionic gonadotropin.
The mean number of recovered ova was described by the authors as having been reduced by DEHP treatment (mean ± SEM 12.7 ± 7.75 [control] compared to 2.50 ± 0.85 [DEHP] in animals given 15 IU equine chorionic gonadotropin and 8.00 ± 4.69 [control] compared to 0.33 ± 0.33 [DEHP] in animals given 30 IU equine chorionic gonadotropin [P.=.0.22.and.0.23,.t.test.by.CERHR]). The authors concluded that DEHP may have suppressed ovulation due to disrupted ovarian steroidogenesis and/or follicle growth.
Strengths/Weaknesses: The use of intact animals allows the assessment of an integrated physiologic response. This is overshadowed by the small sample sizes, the lack of appropriate statistics, and the fact that the studies did not move past simple description, and thus, add little that is new or of value to our understanding.
Utility (Adequacy) for CERHR Evaluation Process: These data are inadequate for the evaluation process, based on small sample size and inadequate statistics.
A 65-week feeding study in marmosets performed at Mitsubishi Chemical Safety Institute, Ltd. (92) contained information on ovarian and uterine weight and histology. Because the focus of this study was on testicular effects, the study is discussed in Section 188.8.131.52.