The inhibition of prostaglandin production during ovulation by exposure to a phthalate mixture is circumvented by cAMP analogue supplementation in a human granulosa cell model.

Active PGE 2 and PGF 2α are increased by LH/hCG late in the ovulatory period to exert their functional roles in oocyte release and CL formation [ 16 , 21 ]. Congruently, PG metabolism is decreased to maintain elevated levels of active PGs [ 16 , 21 ]. In this study, active PGs and inactive metabolites were measured late in the ovulatory period (24 and 36 hr) following exposure to hCG+PHTmix. Treatment with hCG increased the levels of PGE 2 and PGF 2α at the 24 and 36 hr time-points when compared to the DMSO group that did not receive hCG ( Fig. 2A and B ). However, these increases were inhibited by exposure to hCG+PHTmix at several doses. Specifically, hCG+PHTmix decreased PGE 2 levels at the 100 and 500 μg/ml doses at 24 hr and at the 10, 100, and 500 μg/ml doses at 36 hr when compared to the hCG alone ovulatory control group ( Fig. 2A ). Interestingly, hCG+PHTmix initially increased PGE 2 levels at the 1 and 10 μg/ml doses at 24 hr compared to hCG controls; however, levels either returned to hCG (1 μg/ml) or were decreased compared to hCG (10 μg/ml) at 36 hr ( Fig. 2A ). The hCG-induced increases in PGF 2α were inhibited by hCG+PHTmix at the 1, 100, and 500 μg/ml doses at 24 hr and at the 10, 100, and 500 μg/ml doses at 36 hr compared to hCG ( Fig. 2B ). The observed decreases in PGE 2 and PGF 2α levels by hCG+PHTmix were either comparable to the DMSO group (1, 10, and 100 μg/ml) or even further decreased compared to the DMSO group (500 μg/ml) that did not receive hCG treatment. Treatment with hCG did not significantly alter the levels of inactive PGE 2 and PGF 2α metabolites at either time-point when compared to DMSO ( Fig. 2C and D ). Similarly, hCG+PHTmix did not significantly alter the levels of either metabolite when compared to hCG. Although exposure to hCG+PHTmix at the 1 μg/ml dose increased PGE 2 metabolite levels at 24 hr when compared to DMSO, the levels in this hCG+PHTmix group were comparable to the hCG alone group ( Fig. 2C ). Certain PG synthases are increased by LH/hCG in granulosa cells to increase ovulatory PG production [ 16 , 21 ]. Since exposure to hCG+PHTmix decreased active PG levels, the mRNA levels of several PG synthases were measured in the human granulosa cells. Treatment with hCG did not alter the mRNA levels of PLA2G4A at 6 and 12 hr, but the levels were decreased at 24 and 36 hr when compared to DMSO ( Fig. 3A ). Treatment with hCG+PHTmix decreased PLA2G4A levels at the 12 hr time-point at the 500 μg/ml dose, increased levels at the 24 hr time-point at the 1 μg/ml dose, and decreased levels at the 36 hr time-point at the 1 μg/ml dose when compared to hCG alone ( Fig. 3A ). The mRNA levels of PTGS1 were increased by hCG at 12 hr and decreased by hCG at 24 and 36 hr when compared to DMSO ( Fig. 3B ). Exposure to hCG+PHTmix further increased PTGS1 levels at 24 hr at the 1 μg/ml dose and decreased levels at 36 hr at the 500 μg/ml dose when compared to hCG ( Fig. 3B ). The mRNA levels of PTGS2 were increased by hCG at each time-point when compared to DMSO; however, this increase was inhibited by exposure to hCG+PHTmix at 6 hr (100 and 500 μg/ml doses), 12 hr (500 μg/ml dose), 24 hr (100 and 500 μg/ml doses), and 36 hr (all doses tested) when compared to hCG ( Fig. 3C ). However, hCG+PHTmix exposure at the 1 μg/ml dose further increased PTGS2 levels at 24 hr when compared to hCG ( Fig. 3C ). Similar to PTGS2 , the mRNA levels of PTGES were increased by hCG at each time-point when compared to DMSO ( Fig. 3D ). This ovulatory induction of PTGES was inhibited by exposure to hCG+PHTmix at the 500 μg/ml dose at 6, 12, and 24 hr and at the 10 μg/ml dose at 36 hr when compared to hCG ( Fig. 3D ). The mRNA levels of PTGES were also increased by exposure to hCG+PHTmix at the 100 μg/ml dose at 24 hr when compared to hCG ( Fig. 3D ). The mRNA levels of PTGES2 were decreased by hCG at each time-point when compared to DMSO ( Fig. 3E ). Exposure to hCG+PHTmix increased PTGES2 levels at the 10 μg/ml dose at 6 hr and decreased levels at the 1 μg/ml dose at 12 hr when compared to hCG ( Fig. 3E ). The mRNA levels of PTGES3 were not significantly altered by hCG treatment at 6 and 12 hr compared to DMSO, but they were decreased at 24 and 36 hr ( Fig. 3F ). Exposure to hCG+PHTmix decreased the mRNA levels of PTGES3 at the 10 and 100 μg/ml doses at 12 hr, and increased levels at the 100 and 500 μg/ml doses at 36 hr when compared to hCG ( Fig. 3F ). Aldo-keto reductases (AKRs) favor production of PGF 2α over the more potent ovulatory mediator, PGE 2 , whereas HPGD metabolizes PGs to an inactive state [ 21 ]. The mRNA levels of these reductases/metabolic enzymes were measured to explore if increased PG metabolism may serve as a mechanism by which PHTmix exposure decreases ovulatory PGs. Treatment with hCG increased the mRNA levels of AKR1C1 at 6, 12, and 24 hr when compared to DMSO, but decreased levels at 36 hr ( Fig. 4A ). Exposure to hCG+PHTmix increased the levels of AKR1C1 at 6 hr (1, 100, and 500 μg/ml doses), 12 hr (10, 100, and 500 μg/ml), 24 hr (10, 100, and 500 μg/ml), and 36 hr (500 μg/ml) when compared to hCG alone ( Fig. 4A ). The mRNA levels of AKR1C3 were decreased by treatment with hCG at 6, 12, and 36 hr when compared to DMSO; however, hCG+PHTmix exposure increased the levels of AKR1C3 at 6 hr (100 and 500 μg/ml doses), 12 hr (500 μg/ml dose), and 36 hr (100 and 500 μg/ml doses) when compared to hCG ( Fig. 4B ). The mRNA levels of the metabolic enzyme HPGD were decreased by hCG treatment at all time-points tested when compared to DMSO ( Fig. 4C ). Exposure to hCG+PHTmix increased HPGD levels at 6 hr (1, 100, and 500 μg/ml doses) and at 12 hr (500 μg/ml dose) when compared to hCG ( Fig. 4C ). PG transporters modulate the influx/efflux of PGs to/from the granulosa cells, and these transporters are also regulated by hCG to control ovulatory PG function [ 20 , 21 ]. The mRNA levels of two key PG transporters were measured to investigate if phthalate mixture exposure potentially alters granulosa cell PG influx and efflux. Treatment with hCG increased the mRNA levels of ABCC4 at 6 and 12 hr when compared to DMSO; however, this increase was inhibited by hCG+PHTmix exposure at 12 hr at the 1 and 10 μg/ml doses when compared to hCG alone ( Fig. 5A ). Even though hCG did not significantly alter the levels of ABCC4 at 36 hr, exposure to hCG+PHTmix decreased levels at the 10 μg/ml dose when compared to hCG ( Fig. 5A ). The mRNA levels of SLCO2A1 were increased by hCG treatment at 12, 24, and 36 hr when compared to DMSO ( Fig. 5B ). This ovulatory induction was decreased by exposure to hCG+PHTmix at the 500 μg/ml dose at 12 and 24 hr, but SLCO2A1 levels were further increased by hCG+PHTmix at 24 hr at the 10 μg/ml dose when compared to hCG ( Fig. 5B ). PGs exert their ovulatory functions by binding to their receptors, which are present on numerous cell types within the ovarian follicle, including granulosa cells during ovulation [ 21 , 29 ]. To further elucidate the effects of phthalate mixture exposure on PG function, the mRNA levels of the 4 PGE 2 receptors ( PTGER1–4 ) and 1 PGF 2α receptor ( PTGFR ) were measured. The mRNA levels of PTGER1 were decreased by hCG at 12, 24, and 36 hr when compared to DMSO ( Fig. 6A ). Exposure to hCG+PHTmix at the 500 μg/ml dose increased PTGER1 levels at 12 and 36 hr when compared to hCG controls ( Fig. 6A ). Treatment with hCG increased the mRNA levels of PTGER2 at 6, 12, and 24 hr when compared to DMSO; however, this increase was inhibited by hCG+PHTmix at the 100 (6 hr) and 500 (6–24 hr) μg/ml doses when compared to hCG ( Fig. 6B ). Even though hCG decreased PTGER2 levels at 36 hr when compared to DMSO, the levels in the hCG+PHTmix groups were not significantly altered compared to hCG ( Fig. 6B ). The mRNA levels of PTGER3 were also increased by hCG at 6, 12, and 24 hr when compared to DMSO, and this ovulatory induction was decreased by hCG+PHTmix at the 1, 10, and 500 μg/ml doses at 6 hr, the 500 μg/ml dose at 12 hr, and the 10 μg/ml dose at 24 hr when compared to hCG ( Fig. 6C ). Exposure to hCG+PHTmix at the 100 and 500 μg/ml doses also decreased PTGER3 levels at 36 hr compared to hCG controls ( Fig. 6C ). Treatment with hCG decreased the mRNA levels of PTGER4 at each time-point tested when compared to DMSO ( Fig. 6D ). Exposure to hCG+PHTmix increased PTGER4 levels at 6 hr at the 1 μg/ml dose when compared to hCG alone ( Fig. 6D ). The mRNA levels of PTGFR were increased by hCG at 12 and 24 hr compared to DMSO, and while this hCG-stimulation was not statistically significant at 36 hr, exposure to hCG+PHTmix decreased PTGFR levels at the 10, 100, and 500 μg/ml doses when compared to hCG alone at this time-point ( Fig. 6E ). The rise in PG levels during the ovulatory period involves a complex signaling cascade. Activation of LHCGR first increases cAMP/PKA signaling, resulting in increased PGR levels [ 5 , 16 , 17 ], and we have shown that PGR directly modulates the increase in PG synthases, PG transporters, and PG levels in human granulosa cells [ 20 ]. We have also shown that hCG+PHTmix exposure in human granulosa cells decreases PGR levels via decreased cAMP/PKA signaling, and that supplementation with cAMP at the hCG+PHTmix 500 μg/ml dose can restore PGR signaling [ 19 ]. Thus, cAMP supplementation at the hCG+PHTmix 500 μg/ml dose was used in the present study to further elucidate the mechanism of PHTmix-induced ovulatory toxicity, and supplementation was also tested as an approach to rescue the PHTmix-induced decreases in PG levels. As seen in previous experiments ( Fig. 2 ), treatment with hCG increased PGE 2 and PGF 2α levels at 24 and 36 hr when compared to DMSO, and exposure to hCG+PHTmix decreased these active PGs when compared to hCG alone controls ( Fig. 7A and B , solid bars). Supplementation with cAMP increased the levels of PGE 2 and PGF 2α in each treatment group at each time-point when compared to the same treatment groups that did not receive cAMP supplementation ( Fig. 7A and B , dotted bars). Supplementation with hCG+PHTmix 500 μg/ml+cAMP did not increase levels of PGE 2 to the levels comparable to the DMSO+cAMP and hCG+cAMP groups; however, hCG+PHTmix+cAMP had fully restored (24 hr) or further increased (36 hr) levels of PGE 2 when compared to the hCG alone ovulatory control group ( Fig. 7A ). Similarly, supplementation with hCG+PHTmix+cAMP increased PGF 2α levels compared to hCG alone, albeit the levels were decreased compared to the hCG+cAMP group ( Fig. 7B ). Similar to the PG metabolite results in Fig. 2 , treatment with hCG did not significantly alter the levels of either PG metabolite at both time-points tested compared to DMSO, and exposure to hCG+PHTmix did not significantly alter the levels of the metabolites compared to hCG ( Fig. 7C and D ). However, cAMP supplementation decreased PGE 2 metabolite levels in the groups that received hCG treatment. Specifically, supplementation with hCG+cAMP and hCG+PHTmix 500 μg/ml+cAMP decreased PGE 2 metabolite levels at 24 hr compared to hCG alone and hCG+PHTmix, and at 36 hr compared to hCG alone and DMSO+cAMP ( Fig. 7C ). A phthalate treatment effect was not observed in PGF 2α metabolite levels. However, hCG+cAMP decreased PGF 2α metabolite levels at 24 hr compared to hCG alone, and DMSO+cAMP increased PGF 2α metabolite levels at 36 hr compared to hCG alone ( Fig. 7D ). Exposure to hCG+PHTmix decreased PG levels ( Figs. 2 and 7 ) and the mRNA levels of certain PG synthases ( Fig. 3 ) and PG transporters ( Fig. 5 ). Since PGR signaling directly modulates the hCG-induced increases in PTGS2 , PTGES , and SLCO2A1 [ 20 ], supplementation with cAMP, thereby enhancing PGR signaling, may restore phthalate toxicity. In the present study, these were the synthases and transporters that were most abundantly increased by hCG and downregulated by hCG+PHTmix 500 μg/ml. As shown previously, hCG increased PTGS2 and PTGES mRNA levels at each time-point tested compared to DMSO, and hCG+PHTmix 500 μg/ml decreased levels at each time-point compared to hCG alone ( Fig. 8A and B ). Similar results were obtained with SLCO2A1 at 12 and 24 hr; hCG increased levels relative to DMSO and hCG+PHTmix decreased levels relative to hCG ( Fig. 8C ). Supplementation with cAMP increased PTGS2 mRNA levels in each treatment group at each time-point when compared to the same treatment groups that did not receive cAMP supplementation ( Fig. 8A ). Supplementation with hCG+PHTmix+cAMP restored PTGS2 levels beyond hCG alone control levels at each ovulatory time-point, albeit the levels were decreased relative to DMSO+cAMP and hCG+cAMP at 6, 24, and 36 hr ( Fig. 8A ). Supplementation with hCG+PHTmix+cAMP increased the mRNA levels of PTGES beyond hCG+PHTmix and hCG control levels at 6 hr ( Fig. 8B ). At 12 hr, supplementation with hCG+PHTmix+cAMP increased PTGES levels relative to hCG+PHTmix, but the levels were decreased compared to hCG alone ( Fig. 8B ). At the 24 and 36 hr time-points, treatment with DMSO+cAMP and hCG+cAMP increased the mRNA levels of PTGES compared to hCG; however, supplementation with hCG+PHTmix+cAMP had no significant increases compared to hCG alone and hCG+PHTmix ( Fig. 8B ). The mRNA levels of SLCO2A1 were restored to hCG alone control levels when cells were supplemented with hCG+PHTmix+cAMP at 12 hr and were further increased from hCG alone and hCG+PHTmix at 24 hr ( Fig. 8C ). Although hCG+PHTmix did not decrease SLCO2A1 levels at 36 hr compared to hCG, supplementation with hCG+PHTmix+cAMP increased SLCO2A1 levels beyond hCG and hCG+PHTmix ( Fig. 8C ). The additional synthases and transporters that were characterized in the full dose response studies ( Figs. 3 and 5 ) were also measured in the cAMP supplementation experiments. The results for PLA2G4A , PTGS1 , PTGES2 , PTGES3 , and ABCC4 are included in Supplemental Figure 1 . These genes were not abundantly induced by hCG during the ovulatory period and/or their levels were not consistently altered by hCG+PHTmix 500 μg/ml exposure compared to hCG. Further, and unlike the findings in Fig. 8 , supplementation with DMSO+cAMP, hCG+cAMP, and hCG+PHTmix+cAMP did not alter the mRNA levels of these genes compared to hCG alone (when the gene is known to increase with hCG treatment) or DMSO alone (when the gene is known to decrease with hCG treatment) ( Supplemental Fig. 1 ). However, some significant findings specific to phthalate treatment were noted. Supplementation with hCG+PHTmix+cAMP increased the mRNA levels of PLA2G4A at 12 hr when compared to hCG+PHTmix ( Supplemental Fig. 1A ). The levels of AKRs, favoring production of PGF 2α over PGE 2 , and HPGD , which metabolizes PGs, were measured to further elucidate the impact of cAMP supplementation on the rescue of phthalate toxicity. As previously seen ( Fig. 4 ), treatment with hCG increased the mRNA levels of AKR1C1 at 6 and 12 hr compared to DMSO, and hCG+PHTmix further increased AKR1C1 levels at each time-point measured when compared to hCG alone ( Fig. 9A ). Supplementation with cAMP in each treatment group increased the levels of AKR1C1 at 12, 24, and 36 hr when compared to the respective treatment groups that did not receive cAMP, with hCG+PHTmix+cAMP increasing levels beyond hCG alone and hCG+PHTmix ( Fig. 9A ). Treatment with hCG decreased the mRNA levels of AKR1C3 at 6 and 12 hr when compared to DMSO, and hCG+PHTmix exposure increased AKR1C3 levels at 6 and 12 hr when compared to hCG ( Fig. 9B ). Supplementation with hCG+PHTmix+cAMP tended to decrease AKR1C3 levels at 6 hr (decreased levels statistically equivalent to hCG controls), and significantly decreased AKR1C3 levels at 12 hr when compared to both hCG alone and hCG+PHTmix ( Fig. 9B ). The mRNA levels of the metabolic enzyme, HPGD , were once again decreased by treatment with hCG compared to DMSO at each time-point tested, and levels were increased by exposure to hCG+PHTmix compared to hCG at 6, 12, and 24 hr ( Fig. 9C ). Supplementation with hCG+PHTmix+cAMP decreased HPGD levels compared to hCG+PHTmix at 24 hr, where levels were restored to hCG controls ( Fig. 9C ). Levels in the hCG+PHTmix+cAMP group tended to decrease at 6, 12, and 36 hr (decreased levels statistically equivalent to hCG controls); however, the only time-point with statistical significance was 24 hr, as noted above. The mRNA levels of PTGER2 , PTGER3 , and PTGFR were decreased by exposure to hCG+PHTmix ( Fig. 6 ), so their levels were measured following supplementation with cAMP. As seen previously, treatment with hCG increased PTGER2 levels at 6 and 12 hr compared to DMSO, and exposure to hCG+PHTmix decreased PTGER2 levels at 6, 12, and 24 hr compared to hCG alone ( Fig. 10A ). Even though supplementation with DMSO+cAMP increased PTGER2 levels at 6 and 12 hr compared to DMSO, there was not a rescue effect with hCG+PHTmix+cAMP supplementation compared to hCG+PHTmix nor a further increase with hCG+cAMP supplementation compared to hCG ( Fig. 10A ). At 24 hr, each treatment group receiving cAMP supplementation had increased levels of PTGER2 compared to the respective groups that did not receive cAMP, and hCG+PHTmix+cAMP supplementation increased levels when compared to hCG alone and hCG+PHTmix ( Fig. 10A ). Treatment with hCG increased the mRNA levels of PTGER3 at 6, 12, and 24 hr compared to DMSO, and exposure to hCG+PHTmix inhibited this increase at 6 and 12 hr ( Fig. 10B ). Supplementation with hCG+PHTmix+cAMP did not restore PTGER3 levels at 6 hr, but the levels were increased at 12 hr when compared to hCG+PHTmix and were comparable to hCG alone, DMSO+cAMP, and hCG+cAMP ( Fig. 10B ). PTGER3 was not induced by hCG at 36 hr, but hCG+PHTmix decreased levels compared to hCG. Supplementation with hCG+PHTmix+cAMP had comparable levels of PTGER3 with hCG+PHTmix at the 36 hr time-point ( Fig. 10B ). Treatment with hCG increased the mRNA levels of PTGFR at 12 and 24 hr when compared to DMSO, and exposure to hCG+PHTmix decreased PTGFR levels at 36 hr when compared to hCG, even though hCG had comparable levels to DMSO at 36 hr ( Fig. 10C ). Supplementation with DMSO+cAMP and hCG+cAMP increased PTGFR levels when compared to DMSO and hCG at 12 hr, but the hCG+PHTmix and hCG+PHTmix+cAMP groups had statistically similar levels ( Fig. 10C ). The PHTmix-induced decrease in PTGFR levels at 36 hr was rescued by supplementation with hCG+PHTmix+cAMP, as each group that received cAMP supplementation had increased PTGFR levels compared to the respective treatment groups that did not receive cAMP ( Fig. 10C ). Measurements of the mRNA levels of PTGER1 are included in Supplemental Figure 1 , as this was another gene that was not abundantly induced by hCG during the ovulatory period. Exposure to hCG+PHTmix increased the levels of PTGER1 at 6, 12, and 36 hr compared to hCG alone ( Supplemental Fig. 1F ). Supplementation with hCG+PHTmix+cAMP decreased the levels of PTGER1 to hCG controls at the 12 hr time-point ( Supplemental Fig. 1F ). Further, the DMSO+cAMP group had comparable levels of PTGER1 with hCG at 24 hr that were decreased compared to DMSO without cAMP ( Supplemental Fig. 1F ). The mRNA levels of PTGER4 were not measured in the cAMP experiments because this gene was decreased by treatment with hCG, and exposure to hCG+PHTmix did not significantly alter the levels of PTGER4 when compared to hCG ( Fig. 6D ).

The PHTmix included the following composition: 35% diethyl phthalate (DEP), 21% di(2-ethylhexyl) phthalate (DEHP), 15% dibutyl phthalate (DBP), 15% diisononyl phthalate (DiNP), 8% diisobutyl phthalate (DiBP), and 5% butyl benzyl phthalate (BBzP). These percentages were derived from the respective phthalate metabolite levels in the urine of pregnant women enrolled in the Children’s Environmental Health Research Center at the University of Illinois [ 19 , 33 , 34 ]. All phthalates were purchased from Sigma-Aldrich (St. Louis, MO) and were >98% purity. Dimethyl sulfoxide (DMSO) (Sigma-Aldrich, St. Louis, MO) was used as the vehicle to create stock solutions of the mixture at various concentrations (1.33, 13.3, 133, and 655 mg/ml). This allowed for an equal volume of each stock to be added to the culture to control for vehicle concentration (0.75 μg/ml). The final concentrations of the PHTmix in culture were 1, 10, 100, and 500 μg/ml. These doses were selected because they have previously been shown to disrupt ovarian function. Our previous studies identified that these doses inhibit ovulatory progesterone/progesterone receptor signaling in primary human granulosa cells in vitro [ 19 ], and they decrease ovulation rates via decreased PG levels in mouse antral follicles in vitro [ 18 ]. In other studies, these same doses inhibit antral follicle growth, decrease steroid hormone levels, and induce oocyte fragmentation in mouse antral follicles in vitro [ 33 ]. Thus, the selected doses appear to target crucial ovarian processes that are required for ovulation and fertility in human and mouse cells/follicles. When considering the doses of each individual phthalate as a percentage of the total mixture, the lowest dose of the PHTmix (1 μg/ml) contains 150 ng/ml DBP and 210 ng/ml DEHP, which are below the highest measurements of each of the respective metabolites (MBP and MEHP) in human follicular fluid (415 ng/ml for MBP and 239 ng/ml for MEHP) [ 35 ]. Due to the high levels of DEHP, DBP, and DEP in certain medications and medical equipment, humans undergoing certain medical procedures are exposed to much higher levels of phthalates than the general population, especially if interventions are performed intravenously. The higher doses of the PHTmix used in this study may resemble exposure levels in this population, given the doses of the individual phthalates in relation to their percentage in the mixture [ 36 – 39 ]. For the supplementation experiments, 8-Bromoadenosine 3’, 5’-cyclic monophosphate (8-Br-cAMP; hereafter abbreviated to cAMP) was selected as a stable cAMP analogue (Sigma-Aldrich, St. Louis, MO). The final concentration of cAMP in culture was 1.5 mM because this dose was shown in our previous study to rescue the PHTmix-induced inhibition of progesterone/progesterone receptor signaling in primary human granulosa cells [ 19 ]. This dose was also used in previous studies to stimulate steroidogenesis in a similar primary human granulosa cell model [ 40 , 41 ]. The in vitro culture model employed in the present study utilized primary human granulosa cells from patients undergoing in vitro fertilization (IVF). This model has previously been shown to recapitulate periovulatory outcomes as observed in women in vivo , including increased PG production [ 19 , 20 , 42 – 44 ]. Women undergoing IVF at the Bluegrass Fertility Center (Lexington, KY) underwent a standardized ovarian stimulation protocol, which included administration of recombinant human follicle-stimulating hormone (FSH) for 9–11 days. Following FSH treatment, the women were given 10,000 U of hCG to induce the ovulatory cascade, and follicular aspiration was conducted to collect the contents of the follicles at 36 hr post-hCG treatment, which is the time-point just prior to oocyte release. Cumulus oocyte complexes were isolated in the fertility clinic, and the remaining cells from the aspirates were subjected to a Percoll gradient to separate the granulosa cells from red blood cells, which provided a purified population of granulosa cells for culture. The isolated granulosa cells were platted at 2–2.5 × 10 5 cells/ml and were cultured in OptiMEM media containing 10% fetal bovine serum and 1% antibiotic-antimycotic. The granulosa cells were cultured for 6 days and the media were changed every 24 hr. Following this 6-day acclimation period, the granulosa cells possess functional LH/hCG receptors and regain responsiveness to hCG to recapitulate periovulatory outcomes [ 19 , 20 , 42 , 43 ]. Specifically for the scope of this manuscript, these revitalized granulosa cells produce PGs in response to hCG and have increased levels of PG synthases/transporters in response to hCG [ 20 ]. Thus, the acclimation period allows the granulosa cells to revert to a preovulatory phenotype, and re-treatment with hCG is capable of inducing ovulatory outcomes as observed in women in vivo [ 19 , 20 , 42 , 43 ]. The cells were first pre-treated with DMSO (vehicle control) or the PHTmix (1, 10, 100, and 500 μg/ml) for 48 hr prior to hCG treatment. Following 48 hr, the cells were cultured in media without serum for 1 hr and were then treated with or without hCG (1 IU/ml), with or without PHTmix, and with or without cAMP. Cell and media samples were collected at 0, 6, 12, 24, and 36 hr post-hCG treatment for the assays described below. Following each time-point post-hCG, media were collected for prostaglandin measurements. Active prostaglandins (PGE 2 and PGF 2α ) and inactive prostaglandin metabolites (PGE 2 metabolite and 13,14-dihydro-15-keto PGF 2α (hereafter abbreviated to PGF 2α metabolite)) concentrations were measured via enzyme linked immunosorbent assays (ELISAs) (Cayman Chemical, Ann Arbor, MI). The PGE 2 metabolite assay quantifies a single, stable metabolite derivative after the conversion of both 13,14-dihydro-15-keto PGA 2 and 13,14-dihydro-15-keto PGE 2 . The PGE 2 assay sensitivity was 15 pg/ml and intraassay and interassay coefficients of variation were 5.7% and 12.8%, respectively. The PGF 2α assay sensitivity was 10 pg/ml and intraassay and interassay coefficients of variation were 13.8% and 11%, respectively. The PGE 2 metabolite assay sensitivity was 2 pg/ml and intraassay and interassay coefficients of variation were 12.5% and 15.1%, respectively. The PGF 2α metabolite assay sensitivity was 15 pg/ml and intraassay and interassay coefficients of variation were 16.9% and 16%, respectively. Due to patient to patient variability, data were expressed as percent change relative to the hCG alone control group for each patient at each time-point. Following each time-point post-hCG, cells were lysed for quantitative real-time polymerase chain reaction (qPCR) analysis. Total RNA (200ng) was extracted from the cells using the RNeasy Mini Kit (Qiagen, Inc., Valencia, CA) according to the manufacturer’s protocol, and was then reverse transcribed to cDNA using the iScript RT kit (Bio-Rad Laboratories, Inc., Hercules, CA) according to the manufacturer’s protocol. The cDNA samples were further diluted using nuclease-free water. The QuantStudio 3 Real-Time PCR System (Thermo Fisher Scientific – Applied Biosystems, Waltham, MA) and accompanying QuantStudio 3 qPCR Data Analysis Software were used for analysis according to the manufacturer’s protocol. The QuantStudio 3 Real-Time PCR System quantifies the amount of PCR product generated by measuring the fluorescence from the SsoAdvanced Universal SYBR Green Supermix (Bio-Rad Laboratories, Inc.) or TaqMan Gene Expression Master Mix (Invitrogen Life Technologies, Inc., Waltham, MA) for reactions performed with laboratory designed primers or TaqMan primers, respectively. Specific qPCR primer information/sequences (Integrated DNA Technologies, Inc., Coralville, IA; Invitrogen Life Technologies, Inc., Waltham, MA) can be found in Table 1 . GAPDH (TaqMan) and GNPDA2 (SYBR) were selected as the reference genes, as their expression did not significantly differ across time-point or treatment group, as determined statistically. The SsoAdvanced program consisted of an enzyme activation step (95°C for 30 seconds), an amplification and quantification program (45 cycles of 95°C for 10 seconds, 60°C for 10 seconds, single fluorescence reading), and a melt curve (65°C−95°C heating 0.5°C per second with continuous fluorescence readings), as per the manufacturer’s protocol. The TaqMan program included the following steps: 2 minutes at 50°C to permit AmpErase uracil- N -glycosylase optimal activity, denaturation step for 10 minutes at 95°C, 15 seconds at 95°C, and 1 minute at 60°C for 50 cycles, followed by 1 minute at 95°C, 30 seconds at 58°C, and 30 seconds at 95°C for ramp dissociation. Expression data were generated using the mathematical standard comparative ( ΔΔ Ct) method. The Δ Ct was calculated by subtracting the reference gene Ct value from the gene of interest Ct value. The ΔΔ Ct was calculated as the difference between the ΔCt between the treatment groups and the 0 hr DMSO groups (immediately prior to hCG treatment). The relative fold-change of expression was then equaled to 2 −ΔΔCT for each sample. Due to patient to patient variability, data were expressed as percent fold change relative to the hCG alone control group for each patient at each time-point. Data analysis was conducted using SPSS statistical software (SPSS, Inc., Chicago, IL). Data were expressed as means ± standard error of the mean (SEM). To account for normal patient to patient biological variability, data were presented as percent change/percent fold change relative to the hCG alone control group for each patient at each time-point, with hCG set at 100%. Therefore, each patients’ cells serve as their own controls. Multiple comparisons between experimental groups were made using Kruskal-Wallis tests with Dunn-Bonferroni pairwise comparisons. Statistical significance was assigned at p≤0.05.

The main findings from the present study are the first, to our knowledge, to show that exposure to a mixture of phthalates decreases the levels and impairs the function of vital ovulatory PGs in primary granulosa cells obtained from women. Depending on the time-point and PG measured, each dose of the PHTmix tested decreased the levels of PGE 2 and PGF 2α , with both PGs being decreased by the 10–500 μg/ml doses at 36 hr, which is immediately preceding ovulation. Most strikingly, these PHTmix-induced decreases are comparable to or even further decreased than the DMSO non-ovulatory control group that did not receive hCG treatment. Given the requisite physiological roles of PGs in oocyte release and CL formation [ 16 , 21 – 29 ], these findings suggest that phthalate exposure may impair ovulation and fertility in women. A major limitation of the present study is the use of the diester parent compounds and not the phthalate metabolites. To enhance the environmental relevance of the findings, future cellular/tissue in vitro studies should utilize phthalate metabolite mixtures, as parent phthalate compounds are rapidly metabolized to monoester metabolites. It is these metabolites that predominantly reach organ systems and are considered bioactive [ 30 , 45 ]. In vitro studies using a phthalate metabolite mixture have been done by our colleagues [ 46 – 48 ], and this more environmentally relevant exposure paradigm is currently being used by our group in our primary human granulosa cell model. However, it is possible that the parent phthalates included in the PHTmix can reach the ovary during increased exposure conditions. Specifically, certain women undergoing medical procedures/treatments with equipment/medicines containing elevated phthalate levels are likely to have parent phthalate concentrations within the ovary [ 49 – 53 ]. Additionally, mature antral follicles have the capacity to metabolize the parent phthalates that may reach the ovary to their respective bioactive, monoester metabolites [ 49 – 52 ]. Thus, the present findings enhance our understanding of the impact of phthalate exposure on human reproductive health by providing mechanistic evidence of phthalate mixture toxicity in human ovarian granulosa cells. In the present study, the primary mechanism by which PHTmix exposure decreased PG levels is due to inhibited PG production via decreased levels of PTGS2 and PTGES across several ovulatory time-points and PHTmix doses tested. These are the two PG synthetic enzymes that are most abundantly induced during ovulation in human granulosa cells as depicted by the present data and our previously published study [ 20 ]. Further, PHTmix exposure appeared to promote an imbalanced shift towards metabolism of the more functional ovulatory PGE 2 to PGF 2α , via increased levels of AKR1C1 and AKR1C3 , and of active PGs to their inactive metabolites, via increased levels of HPGD . However, these molecular changes related to metabolism did not result in a PHTmix-induced increase in inactive PG metabolite levels, and this is perhaps attributed to the decreased availability of PGE 2 and PGF 2α to be metabolized due to decreased synthesis. In addition to decreased levels, PG function may be impaired by phthalate exposure due to the PHTmix-induced decreases in the levels of transporters ( SLCO2A1 and ABCC4 ) and receptors ( PTGER2 , PTGER3 , and PTGFR ). These decreases can potentially impair influx/efflux of PGs across granulosa cells and decrease receptor availability, both of which would hinder PGs from exerting their functions. To further elucidate the molecular mechanism by which phthalate exposure decreased PG levels, and to investigate preliminary means to circumvent phthalate toxicity, cAMP supplementation was used to rescue the toxic effects on PG levels. Specifically, cAMP supplementation to hCG+PHTmix treated granulosa cells either fully restored or further increased the levels of PGE 2 and PGF 2α to hCG alone ovulatory control levels. This was achieved via restored or further increased levels of PTGS2 and PTGES in the hCG+PHTmix+cAMP supplementation group and via restored decreases in AKR1C3 and HPGD to hCG levels. Additionally, PG function was potentially rescued with cAMP supplementation because the hCG+PHTmix+cAMP group had levels of SLCO2A1 , PTGER2 , PTGER3 , and PTGFR that were comparable to or further increased than hCG controls. However, an additional limitation to this study is that cAMP supplementation was only performed in the hCG+PHTmix 500 μg/ml group, and future studies will measure cAMP levels and PKA activity at the lower doses and will establish if cAMP supplementation at the lower doses rescues toxicity. We have previously shown that PHTmix exposure inhibited PGR signaling via decreased cAMP levels and PKA activity in human granulosa cells, and that cAMP supplementation was able to rescue this toxic effect on PGR at the 500 μg/ml dose [ 19 ]. We have also shown that PG synthases, PG transporters, and the ovulatory induction of PG levels are dependent on PGR signaling in human granulosa cells [ 20 ]. Thus, the underlying molecular mechanism of PHTmix toxicity on ovulatory PG function appears to include the following temporal defects: 1) decreased cAMP/PKA, 2) inhibited PGR signaling, 3) decreased expression of PG synthases, transporters, and receptors, and 4) decreased active PG levels. A working model of these findings is provided in Fig. 11 . The cAMP rescue findings further establish this mechanism and provide potential targets and signaling pathways in which to intervene on phthalate toxicity at the cellular level. Alongside the current findings of decreased synthesis ( PTGS2 and PTGES ) and increased catabolism/metabolism ( AKR1C1 , AKR1C3 , and HPGD ), additional patterns of PHTmix-induced changes were noted that conflict with the typical temporal changes in gene expression during the ovulatory period. In this study, treatment with hCG alone generally increased the levels of PTGS2 , PTGES , AKR1C1 , ABCC4 , SLCO2A1 , PTGER2 , PTGER3 , and PTGFR , and generally decreased the levels of PLA2G4A , PTGS1 , PTGES2 , PTGES3 , AKR1C3 , HPGD , PTGER1 , and PTGER4 . These findings are consistent with the temporal regulation of PG synthetic/metabolic enzymes, transporters, and receptors by hCG during the ovulatory period that has been documented in several species [ 20 , 21 , 29 ]. However, in this study, when hCG induced the genes mentioned above, exposure to the PHTmix generally decreased the mRNA levels of those genes. The major exception was with AKR1C1 , where hCG is known to increase its levels, but exposure to the PHTmix further increased AKR1C1 levels, again aligning with a mechanism involving favored PG metabolism following PHTmix exposure. For the genes mentioned above that are known to be downregulated by hCG, exposure to the PHTmix generally increased the mRNA levels, but this pattern was less consistent than with the pattern of genes that are typically induced by hCG. Specifically, there were both increases and decreases in PLA2G4A , PTGS1 , and PTGES3 following PHTmix exposure, which were impacted by the dose of PHTmix treatment and time-point of collection. However, the overall PHTmix patterns described above contrast with the typical hCG ovulatory response, and these general patterns arguably align with that of the non-ovulatory, DMSO phenotype. This is especially clear at 36 hr with PTGS2 , where each tested dose of hCG+PHTmix had levels that were comparable to the DMSO group, which were well-below that of hCG. Future studies will measure protein and enzymatic activity levels of the synthases/reductases, metabolic enzymes, transporters, and receptors to clarify any discrepancies in mRNA levels with changes in PG levels and function. The hCG+PHTmix 1 μg/ml treatment group increased the mRNA levels of multiple synthases ( PLA2G4A , PTGS1 , and PTGS2 ) at 24 hr ( Fig. 3A – C ), which corresponded to the only observation of a significant increase in PG levels by phthalate treatment (PGE 2 at 24 hr; Fig. 2A ). This may explain the increase in PGE 2 levels at this particular dose and time-point, but it is important to note that the increases in PLA2G4A and PTGS1 were at time-points where hCG alone downregulated these genes when compared to DMSO. Therefore, the molecular effects in the hCG+PHTmix 1 μg/ml group were more similar to those in the DMSO group than those in the hCG controls. In other words, perhaps hCG temporally decreases the levels of certain synthases ( PLA2G4A at 24 and 36 hr, PTGS1 at 24 and 36 hr, PTGES2 at all time-points, and PTGES3 at 24 and 36 hr) in order to facilitate the increases in more functional/potent synthases ( PTGS2 and PTGES are both increased by hCG at all time-points). Regardless, hCG+PHTmix 1 μg/ml exposure decreased the mRNA levels of PTGS2 (36 hr) and increased the levels of AKR1C1 (6 hr) and HPGD (6 hr) that likely contributed to the functional defect of decreased PGF 2α levels at this dose. Importantly, the 1 μg/ml treatment group contains concentrations of individual phthalates whose metabolites’ concentrations are within the levels found in women’s follicular fluid, suggesting potential environmental relevance [ 35 ]. Findings from the present study provide several important strengths to the field of EDC/reproductive toxicology by circumventing previous limitations in the literature and enhancing the mechanistic knowledge pertaining to ubiquitous exposure to toxicants, such as phthalates. Historically, phthalate ovarian toxicology studies utilize single phthalate exposures [ 1 , 5 ], even though women are exposed to a mixture of phthalates on a daily basis [ 30 – 32 ]. The use of this PHTmix, which was derived from urinary phthalate metabolite levels in pregnant women, is a more physiologically relevant approach for toxicology studies. Concurrently, the direct effects of phthalates on human samples are markedly absent in the literature, as the majority of studies use rodent models [ 1 , 5 ]. Our well-characterized ovulation model utilizes primary granulosa cells obtained from women undergoing IVF, and this in vitro model replicates known signaling cascades (including cAMP/PKA and PGR), hormonal changes (including P4 and PGs), and molecular signatures (including PG synthases, transporters, and receptors) that have been documented during the ovulatory period in normally cycling women in vivo [ 19 , 20 , 42 – 44 ]. Further, investigating the impact on the ovulatory process is a novel component of this study, as direct effects of exposure to phthalates on this crucial period for fertility are understudied [ 1 , 5 ]. Meanwhile, defects in ovulation are the top contributors for female infertility [ 9 ], and toxicant exposures are increasingly being recognized as contributing to female infertility [ 6 , 10 – 12 ]. The source of the primary samples from women introduces several complexities and may be considered another limitation to the current study. PGs are implicated in various pathological conditions, including hypertension, cancer, and inflammatory disorders [ 54 ]. As such, women with these pathologies may have pre-existing alterations in PG levels and function and/or a potential altered ovarian response to hCG to produce PGs in our study design. In the present study, we were unable to collect patient information related to these PG-implicated conditions, so it is unknown if these pathologies impact the patients’ ovarian response to hCG. However, PGs, including those outside of PGE 2 and PGF 2α , exert effects that are organ and tissue specific [ 54 ]. The increase in PGE 2 and PGF 2α levels in the human ovary during ovulation is well-timed timed following LHCGR activation and is localized specifically to the granulosa cells [ 16 ]. These increases were consistently replicated in the granulosa cells treated with hCG in the present experiments, regardless of any underlying pathology, suggesting that our in vitro granulosa cell model is suitable for investigating the impact of phthalate exposure on ovulatory PG production. The fertility status of the women undergoing IVF adds an additional complexity in interpretation of this study. Samples were collected from women that are fertile (oocyte donors, oocyte preservation, and male factor infertility), women with non-ovarian infertility diagnoses (tubal factor infertility), women with ovarian-related disorders (endometriosis, polycystic ovary syndrome, and advanced maternal age), and couples with an unknown infertility diagnosis. Samples from women with a known infertility diagnosis may complicate interpretation of the results because it is presently unknown how these reproductive disorders may impact the function of the granulosa cells in this in vitro model. However, we were unable to stratify samples based on fertility status due to limited patient availability and limited cell number collected from some patients. In this study, we collected granulosa cells from 42 patients in which 57% had no known ovarian diagnosis and 45% had no known female infertility diagnosis ( Supplemental Table 1 ). We are currently increasing our patient population to investigate if any non-reproductive pathology or infertility diagnosis correlates with negative measurable outcomes in this model. Importantly, we did not observe major differences in the hCG response between the patients’ samples in the present study, regardless of fertility issues and potential pathologies associated with altered PG function. Further, we are actively measuring phthalate levels in these follicular fluid aspirates at the time of oocyte collection to also investigate associations of increased phthalate exposure with infertility diagnoses and/or poor IVF outcomes and to determine if a woman’s baseline level of phthalate exposure during the IVF process impairs the granulosa cells’ response to hCG and/or predisposes her granulosa cells to increased phthalate toxicities in this model. Regardless of the potential limitations associated with using samples obtained from IVF, the use of primary human granulosa cells is a major strength of the current study due to the lack of information available regarding the direct effects of phthalates on the human ovary. In summary, findings from the present study are the first, to our knowledge, to demonstrate that phthalate mixture exposure impairs the production, metabolism, and function of the vital ovulatory PGs in human granulosa cells. The cAMP supplementation findings provide a potential target for toxicity rescue at the cellular level and support a mechanism by which phthalate exposure impairs ovulatory PKA signaling, P4/PGR signaling, and disrupted PG function during the process of ovulation [ 19 ]. These findings in women support our observations in the in vitro mouse follicle culture model where the same doses of the PHTmix decreased PGE 2 and PGF 2α levels ultimately resulting in decreased ovulation rates [ 18 ]. The current observations contribute to the growing literature that suggests that phthalate exposure may impair women’s reproductive health and fertility.

Phthalates are a group of chemicals used as plasticizers and solvents in the manufacturing of numerous common consumer products including food and beverage packaging, personal care products, and building materials. Leaching of phthalates from these products results in daily human exposure via oral ingestion, inhalation, and dermal contact [ 1 – 5 ]. This continuous exposure is considered a public health and women’s health concern because phthalates are endocrine-disrupting chemicals (EDCs) and female reproductive toxicants, in which the ovary is a target organ of phthalate toxicity [ 1 , 5 , 6 ]. Specifically, phthalates can directly act on ovarian cells in women because measurable levels of phthalates are detected in human ovarian follicular fluid [ 7 , 8 ]. Further, phthalate exposure in rodent models has been shown to disrupt essential ovarian processes, including steroidogenesis and ovarian follicle development (folliculogenesis) [ 1 , 5 , 6 ]. Impaired steroidogenesis and folliculogenesis suggest that phthalate exposure can inhibit the release of the egg for fertilization (ovulation). However, much less is known regarding the direct effects of phthalate exposure on the ovulatory process. Defects in ovulation are the leading cause of infertility in women [ 9 ], and increasing evidence suggests that exposure to EDCs, such as phthalates, contributes to the prevalence of infertility [ 6 , 10 – 12 ]. In addition, adult women have an increased exposure profile to certain phthalates compared to men, due to the abundance of phthalates in personal care products [ 1 , 13 – 15 ]. Thus, there is an urgent demand to investigate the effects of phthalate exposure on ovulation [ 5 ]. Ovulation is initiated by the luteinizing hormone (LH) surge or clinical treatment with human chorionic gonadotropin (hCG; a potent LH analogue). LH/hCG bind to the LH/hCG receptor (LHCGR; a G-protein coupled receptor) on granulosa cells within the ovulatory follicle, which results in the immediate activation of numerous intracellular signaling molecules/pathways, most notably the cyclic adenosine monophosphate (cAMP)/protein kinase A (PKA) pathway [ 5 , 16 , 17 ]. This signal transduction pathway increases the levels of multiple genes/proteins and hormones that serve as mediators of the ovulatory process [ 5 , 16 , 17 ]. Among the most well-characterized LH/hCG-induced mediators of ovulation across species are progesterone (P4), progesterone receptor (PGR), and prostaglandins (PGs) [ 16 ]. Our group has utilized a mixture of six phthalates (PHTmix) that was derived from urinary phthalate levels in pregnant women, and we reported that PHTmix exposure decreased ovulation rates in mouse follicles in vitro via impaired P4/PGR signaling and decreased PG levels [ 18 ]. We have also shown that the same PHTmix inhibited ovulatory P4/PGR signaling in primary human granulosa cells obtained from women undergoing in vitro fertilization (IVF), and that this inhibition was due to decreased cAMP levels and PKA activity [ 19 ]. However, the effect of PHTmix exposure on ovulatory PG production in the human ovary is unknown. We have previously identified that ovulatory PG production is downstream and dependent on P4/PGR signaling in human granulosa cells. Specifically, PGR in human granulosa cells directly regulates the transcription of prostaglandin synthases and transporters, resulting in the increased levels of ovulatory PGE 2 and PGF 2α (two active PGs in the ovary) [ 20 ]. Since we have previously shown that the PHTmix inhibited P4/PGR signaling [ 19 ], ovulatory PG production and function in women may also be impaired by phthalate exposure. Functionally, PGs are involved in the regulation of angiogenesis, increased blood flow in the ovulatory follicle, expansion of the cumulus-oocyte complex, and tissue remodeling, which are processes required for oocyte release and formation of the corpus luteum (CL; remnant ovarian structure following ovulation that supports pregnancy) [ 16 , 21 ]. Blocking the synthesis and action of PGs result in ovulatory failure, which underscores the importance of PGs for fertility. Specifically, treating with PG synthase inhibitors, nonsteroidal anti-inflammatory drugs, and PG receptor antagonists, or knocking out PG synthases results in anovulation and infertility in women and animal models [ 16 , 21 – 28 ]. The abundant increase in PGE 2 levels is requisite for ovulation and fertility, whereas PGF 2α , though still increased during the ovulatory period, has a less understood role and may be a catabolic product of increased PGE 2 [ 16 , 21 ]. Several PG synthases/reductases (PTGS1, PTGS2, PTGES, and AKR1C1), transporters (SLCO2A1 and ABCC4), and receptors (PTGER2, PTGER3, and PTGFR) are temporally increased in granulosa cells by LH/hCG; whereas PG metabolic enzymes (HPGD) are decreased during the ovulatory period to maintain elevated levels of active PGs [ 16 , 21 , 29 ]. Additional PG synthases (PLA2G4A, PTGES2, and PTGES3), conversion enzymes (AKR1C3), and receptors (PTGER1 and PTGER4) are constitutively expressed and not induced in granulosa cells by LH/hCG, but phthalate exposure may still alter their levels leading to impaired ovulatory PG function. To initiate PG production, arachidonic acid is cleaved from membrane phospholipids by PLA2G4A. PTGS2, and to lesser extent in the human ovary PTGS1, converts arachidonic acid to PGH2, which is a short-lived PG intermediate. PTGES, and to lesser extent PTGES2 and PTGES3, covert PGH2 to active PGE 2 , which is considered to be the key ovulatory PG in comparison to PGF 2α [ 21 ]. AKR1C1 converts PGE 2 to PGF 2α , whereas AKR1C3 converts the PGH2 intermediate directly to PGF 2α . Finally, PGs are rendered inactive by the metabolic enzyme, HPGD. A schematic of the ovulatory PG production/metabolic pathway, including transporters, receptors, and upstream PGR-mediated events, is depicted in Fig. 1 . Given the importance of PGs in ovulation and fertility, the present study exposed primary human granulosa cells to the PHTmix to test the hypothesis that phthalate exposure decreases ovulatory PG levels and function by decreasing PG synthases, transporters, and receptors, and increases PG metabolism. Additionally, the present study utilized cAMP supplementation (8-Br-cAMP; stable cAMP analogue) to PHTmix treated cells to further elucidate the mechanism of phthalate-induced ovulatory dysfunction and to provide a potential cellular target for intervention of phthalate toxicity. Considering that the vast majority of studies in the literature focus on single phthalate exposures in rodent models, the use of a phthalate mixture in human granulosa cells in the present study seeks to better mimic exposure in women and to circumvent these gaps in knowledge [ 5 , 30 – 32 ]. As such, this study contributes to the field of EDC/reproductive toxicology by providing mechanistic evidence by which phthalate mixture exposure impairs a crucial ovulatory pathway in the human ovary.

Supplemental Figure 1: Effect of cAMP supplementation on the levels of PG synthases, transporters, and receptors that were not abundantly induced by hCG. Human granulosa cells from in vitro fertilization patients (2–2.5 × 10 5 cells/ml) were treated with DMSO (vehicle control) or phthalate mixture (PHTmix; 1–500 μg/ml) for 48 hr prior to hCG and 8-Br-cAMP (cAMP) treatment. Following 48 hr, the cells were treated with or without hCG, with or without PHTmix, and with or without cAMP. Cells were collected at multiple time-points following hCG treatment to measure the mRNA levels of PLA2G4A (A), PTGS1 (B), PTGES2 (C), PTGES3 (D), ABCC4 (E), and PTGER1 (F) via qPCR. Values were normalized to the reference gene and are presented as a percent fold change relative to the hCG alone control group at each time-point. Graphs represent mean ± SEM. Bars that do not share a letter designation are significantly different (n=3–16 patients/group, p≤0.05). Supplemental Table 1 : Characteristics of IVF patients from whom granulosa cells were collected for experiments in the present study. Percent of patients without ovarian diagnosis was calculated as the number of patients with only male factor infertility, were oocyte donors without a female fertility diagnosis, were same sex couples without a female fertility diagnosis, were women undergoing fertility preservation without a female fertility diagnosis, and were women with only tubal factor issues divided by the total number of patients in the present study and multiplied by 100. Anovulation, endometriosis, PCOS, low AMH levels, advanced maternal age, and unexplained diagnoses were considered potential ovarian defects, and these patients were not included in the calculation. Percent of patients without any female infertility diagnosis was calculated as the number of patients with only male factor infertility, were oocyte donors without a female fertility diagnosis, were same sex couples without a female fertility diagnosis, and were women undergoing fertility preservation without a female fertility diagnosis divided by the total number of patients in the present study and multiplied by 100. Not applicable (n/a); Polycystic ovary syndrome (PCOS); Anti-Müllerian Hormone (AMH).