The objective of the present study was to characterize the potential for 2-hydroxy-4-methoxybenzophenone (2H4MBP), a common component of sunscreen and personal care products, to adversely affect any phase of rat development, maturation, and ability to reproduce. Mechanistic screening studies have shown that 2H4MBP and its metabolites are capable of activating the estrogen receptor and antagonizing the androgen receptor to varying degrees.^102,103 In this study, Sprague Dawley (Hsd:Sprague Dawley^® SD^®) rats were exposed to 2H4MBP in 5K96 feed, using the National Toxicology Program (NTP) modified one-generation (MOG) study design. As disposition is similar following oral and dermal exposure, 2H4MBP exposure via the diet was selected for this study, rather than topical application, to sustain internal exposure; if applied topically, internal dose would be influenced by intra- and interanimal grooming behavior. To minimize the potential endocrine activity of phytoestrogens that are often present in rodent diets, 5K96 feed was used because it provides a diet low in phytoestrogens. This report complements ICH^c S5r2 guideline studies (fertility and early embryonic development, embryo-fetal development, and pre- and postnatal developmental studies in rats) on 2H4MBP⁵⁷ conducted by the U.S. Food and Drug Administration’s National Center for Toxicological Research (NCTR), an interagency NTP partner, and allows for the comparison of study designs and outcomes.

Exposure concentration selection was informed by a dose range-finding study that demonstrated that 25,000 ppm was well tolerated in pregnant rats and did not affect parturition, litter size, or pup viability. In that study, pup body weights of the 25,000 ppm group were significantly decreased compared to the control group, suggesting potential growth retardation; this response was severe at the 50,000 ppm exposure concentration and viable litter size was also affected. Therefore, 30,000 ppm was selected as the highest exposure concentration for the MOG study. The exposure concentrations of 3,000 and 10,000 ppm were selected to aid in identifying potential exposure concentration-response relationships. This spacing would ideally avoid significant overlap of the respective mg 2H4MBP/kg body weight (mg/kg) exposure concentrations, recognizing that the amount of feed consumed is dependent on pregnancy state (e.g., prior to mating versus lactation), sex, and age. Because 2H4MBP has been reported to induce estrogen-like activity, a low exposure concentration (0.05 ppm) of ethinyl estradiol (EE), a synthetic form of estrogen, was included as a positive control group. NTP studies have shown that comparing plasma concentrations of 2H4MBP in rats following feed exposure of 3,000–30,000 ppm to plasma concentrations in humans¹⁹ following repeated dermal application of 20 g/m² revealed rat-to-human dose multiples of 0.1 to 4. Collectively, these data demonstrate similar external (5- to 57-fold) and internal (0.1- to 4-fold) exposure of 2H4MBP in rats and humans.

Exposure of F₀ females to 2H4MBP or EE via the diet began on gestation day (GD) 6 (implantation). F₁ offspring were exposed to 2H4MBP or EE at the same exposure concentration as their respective dams. Upon weaning, F₁ offspring in each group were randomly assigned to one of three cohorts: (1) reproductive performance cohort (2/sex/litter), (2) prenatal cohort (1/sex/litter), and (3) biological sampling cohort (1/sex/litter). Upon sexual maturity, nonsibling F₁ rats allocated to the prenatal and reproductive performance cohorts were paired for mating to evaluate reproductive performance and F₂ prenatal and postnatal development. The likelihood of identifying potential 2H4MBP-induced adverse effects (similarity and magnitude thereof) at any phase of growth or development was increased by examining related endpoints in multiple pups within a litter during both preweaning and postweaning periods.

The concentrations of free (unconjugated compounds) and total (free and all conjugated forms) 2H4MBP, 2,4-dihydroxybenzophenone (DHB), 2,3,4-trihydroxybenzophenone (THB), and 2,5-dihydroxy-4-methoxybenzophenone (D2H4MBP) were quantified in plasma from the biological sampling cohort at postnatal days (PNDs) 28 and 56.⁶⁸ Free plasma 2H4MBP and DHB concentrations were similar to each other and increased with increasing exposure concentration, with no age or sex differences except in the 10,000 ppm group, as concentrations of both analytes were significantly increased in PND 56 animals relative to PND 28 animals. Free D2H4MBP and THB were not detected in these animals. The concentrations of total 2H4MBP and DHB were higher (approximately 100- to 300-fold) than the free 2H4MBP and DHB concentrations, demonstrating extensive conjugation of 2H4MBP and its metabolites. The rank order of the total concentrations was 2H4MBP ≈ DHB > D2H4MBP >> THB. Free and total analyte plasma concentrations were not sex-dependent in either PND 28 or PND 56 pup plasma.

In the current MOG study, 2H4MBP exposure was associated with lower F₁ and F₂ mean body weights (8%–24%). Lower preweaning F₁ pup mean body weights have also been observed in CD-1 mice exposed to 2H4MBP.⁴¹ The lower F₁ body weights observed postweaning to sexual maturity were not associated with lower feed consumption. Pregnant F₀ females and females in both F₁ cohorts exposed to 2H4MBP also did not display decreases in gestational or lactational feed consumption. Collectively, this suggests that 2H4MBP could have altered utilization of the consumed diet (and, thus, affected growth) and could have reduced or delayed preweaning growth. The observed lower mean body weights of the 2H4MBP groups, in the absence of effects on feed consumption, is consistent with findings reported in Fischer 344 (F344)/N rats administered 25,000 ppm 2H4MBP.⁴⁰

2H4MBP did not result in any significant effects on mating, pregnancy, or littering indices, nor did it result in adverse histopathological findings in the testis or changes in sperm parameters at concentrations up to 30,000 ppm. These observations contrast with those reported in the NTP Reproductive Assessment by Continuous Breeding study⁴¹ in CD-1 mice, in which 2H4MBP was associated with smaller litter sizes and decreases in pup viability. In a previous study, 50,000 ppm was associated with lower sperm density in both F344/N rats and mice.⁴⁰ No effects on sperm parameters were apparent at the next lower exposure concentration (12,500 ppm) in rats, however, percent of sperm cell abnormalities were significantly increased in mice at this exposure concentration.⁴⁰ These findings were collectively attributed to stress-induced toxicity, potentially by affecting metabolism or digestive processes, as evidenced by lower mean body weights. Chronic stress is known to affect rat spermatogenesis.^104,105 The absence of similarly robust effects on sperm parameters and reproductive performance in the current study might reflect strain and stock differences. Alternatively, it is possible that if higher 2H4MBP exposure concentrations could have been used in this study, a similar magnitude of response to that observed in the CD-1 mouse and F344/N rat in previous studies may have been observed, either as a stress-related response or as a more direct effect from 2H4MBP exposure.

Examining data across cohorts in the 30,000 ppm group, mean numbers of corpora lutea and F₂ implants on GD 21 were significantly decreased (3.7 and 2.7, respectively) relative to the control group, and the mean number of live fetuses was also lower (1.7). Mean live F₂ litter size on PND 0 was significantly decreased (1.5 pups) in the reproductive performance cohort, and F₁ live litter size on PND 0 was also slightly lower (<1 pup). These observations suggest that 2H4MBP exposure might have affected litter size, although the magnitude of this effect was small. The slightly smaller litter size might have been due to a direct effect (the decrease of the number of ova ovulated, as evidenced by the lower number of corpora lutea enumerated in the prenatal cohort) or an indirect effect of a stress-induced response (reflected in the lower mean body weights). 2H4MBP, administered at 50,000 ppm in the diet from GD 6 through PND 23, has been shown to delay follicular development, but this was not observed at 25,000 ppm.¹⁸ 2H4MBP has also been shown to affect early follicular assembly in rat ovary cultures.¹⁰⁶ Thus, the observed decrease in corpora lutea is consistent with alterations in follicular development.¹⁸ Nonetheless, no histopathological alterations of the ovaries were observed. No subsequent 2H4MBP-related effects on live litter size were observed. Collectively, given the minimal apparent response that may or may not be a direct effect of 2H4MBP, this was considered equivocal evidence of an adverse effect on reproductive performance.

EE exposure did not affect F₁ live litter size on PND 0; however, mean live F₂ litter size on PND 0 in the reproductive performance cohort and the mean number of live F₂ fetuses per litter on GD 21 in the prenatal cohort were both significantly decreased (approximately 2–3 pups per litter) relative to the control group. Fewer corpora lutea and total F₂ implants were observed in the EE prenatal cohort. Rat follicular development has been shown to be affected by EE (200 µg/kg) when exposed on PND 0 and examined on PND 21.¹⁰⁷ F₂ live litter size on PND 0 through PND 4 in the EE F₂ reproductive performance cohort was significantly decreased relative to the control group (approximately 2 pups per litter) in part because 3 of the 18 EE litters had 0 pups. After litter standardization on PND 4, survival in the EE group appeared higher than in the control group, but this was likely the result of several control litters that exhibited excessive pup loss. In a previously conducted multigenerational study, EE exposure at 0.05 ppm was not reported to significantly decrease (or increase) the number of live pups born.⁹⁷ Upon inspection of the NCTR study data, however, there is an apparent minimal nonsignificant decrease in mean live born (approximately 1 pup per litter) that is consistent with what was observed in the EE group in the current study.⁹⁷ A similar decrease in number of implants was observed in the NCTR Segment 1 study.¹⁰⁸

Progressively lower relative preweaning F₁ body weights were observed in males and females exposed to 30,000 ppm 2H4MBP. On PND 4, F₁ males and females displayed significantly decreased mean body weights of 15% and 12%, respectively, relative to the control group, and by PND 28, body weights of both males and females were significantly decreased by approximately 24%. In contrast, F₂ males and females did not exceed a 10% lower relative body weight until PND 25 and PND 19, respectively. The reason for this difference in F₁ versus F₂ generational response is unclear, but it could be related to increased 2H4MBP metabolism in the F₁ dams resulting from sustained 2H4MBP exposure. The no-observed-effect level (NOEL) for 2H4MBP-related effects on body weight is 3,000 ppm based on lower body weights in both sexes in both generations. The considerable effects on body weights associated with exposure to 2H4MBP were considered some evidence of developmental toxicity.

2H4MBP did not accelerate vaginal opening (VO), as would be expected if it displayed estrogenic activity, consistent with the expected robust acceleration of VO that was observed with EE. The day of VO attainment was delayed in the 30,000 ppm group, and body weights on day of acquisition were similar to those of the control group. When weaning weight was used as a covariate, addressing growth retardation, the apparent delay was mitigated. A similar VO delay, concomitant with lower mean body weight, has been reported for corticosterone administered in drinking water.¹⁰⁹ Intrauterine growth retardation—after ligation of the uterine artery on GD 17 and resulting in 16% lower body weight on PND 2 and lower postnatal body weights relative to the control group—has been shown to delay VO.¹¹⁰ Postnatal dietary restriction also has been shown to delay VO, with similar body weights relative to the control group at time of VO.¹¹¹ The lower PND 4 pup and postnatal mean body weights and the delay in VO observed in the current study are consistent with these findings.

2H4MBP exposure did not significantly alter any apical androgen-sensitive endpoints, demonstrating that it does not appear to affect androgen-mediated lengthening of anogenital distance or advancement of balanopreputial separation (BPS). 2H4MBP did not affect areola/nipple retention at the tested concentrations, indicating an absence of androgen-receptor antagonism. BPS was delayed in the 10,000 and 30,000 ppm 2H4MBP groups, as well as in the 0.05 ppm EE positive control group. Similar to VO, body weights on day of acquisition were comparable to those of the control group, and, when adjusted for weaning weight, there were also no differences relative to the control group. Intrauterine growth retardation and postnatal feed restriction, resulting in lower postnatal body weights, have been shown to delay BPS.¹¹⁰ It is plausible that, like VO, the similar weights on day of attainment observed in the current study suggest a weight or body mass requirement for the attainment of BPS.

Diaphragmatic hernias were observed at a low incidence in 2H4MBP-exposed animals in both the F₁ and F₂ generations but were not observed in any control animals. They were also not observed in control animals in two other MOGs (EHMC and BPAF).^112,113 This finding was also observed in the male F₁ and F₂ EE groups. Diaphragmatic hernias have been shown to be induced by 2,4-dichlorophenyl-p-nitrophenyl ether, which displays some similarity to 2H4MBP.^114,115 The presence of gross diaphragmatic hernias correlated with histologic hepatodiaphragmatic hernias in all but two animals. Although these incidences occurred only in exposed groups, there was no exposure response and no pairwise significance, and they have been observed in control groups in other developmental and reproductive toxicity studies. Therefore, it is unclear whether the occurrence of diaphragmatic and hepatodiaphragmatic hernias were related to 2H4MBP exposure.

No malformations observed at adult necropsy were consistent with perturbation of normal androgen-receptor-mediated development (grossly normal prostate, seminal vesicles, and epididymides). There was, however, a single incidence of hypospadias, a lesion commonly seen when androgen action is attenuated.^116,117 Given the singular incidence and the absence of corresponding changes in androgen-dependent processes, the hypospadias was likely not related to 2H4MBP exposure. In F₁ adult males in the reproductive performance cohort, the weights of androgen-dependent reproductive tissues (testes, epididymides, ventral prostate gland) and levator ani/bulbocavernosus muscle complex were all slightly lower in the 30,000 ppm group compared to the control group. All of those organ weight changes occurred concurrently with lower body weights, however, and are likely secondary to the apparent growth retardation. Moreover, there were no apparent 2H4MBP-related histopathological findings in the reproductive tissues, nor was the ability of males to successfully mate and induce pregnancy adversely affected in either the prenatal or reproductive performance cohorts. Sperm and spermatid counts, which are androgen-sensitive endpoints, were also not affected. In totality, unlike what has been reported in cell models, 2H4MBP exposure had no apparent effect on androgen-receptor-dependent processes, nor did it affect mating or pregnancy indices.

2H4MBP exposure was associated with greater kidney weights and histologic lesions consistent with obstructive nephropathy, including renal tubule epithelial regeneration, renal tubule degeneration (females only), interstitial chronic active inflammation, renal tubule and pelvic concretions, renal tubule dilation, papillary necrosis, urothelial hyperplasia, and urothelial ulcers. In addition, increased chronic progressive nephropathy, pelvic dilation, and renal mineralization were present in females. These findings are consistent with renal effects previously reported following subchronic exposure⁴⁰ and those observed with chronic exposure.³⁴ F₁ males and females exposed to 2H4MBP also displayed greater liver weights. This finding is consistent with the fetal malformation finding of enlarged liver. The absolute weights of the adrenal glands were significantly decreased in the female 30,000 ppm reproductive performance cohort. Chronic stress would be expected to increase corticosterone levels and result in lower adrenal gland weights due to negative feedback; however, sustained elevated adrenocorticotropic hormone (or equivalent) would be expected to increase both adrenal gland weight and the levels of corticosterone.¹¹⁸ The NOEL for adult general toxicity necropsy findings is 3,000 ppm based on increases in kidney weights and histopathological findings in the urinary system consistent with chronic obstructive nephropathy.

There was no effect of 2H4MBP exposure on the incidence of fetal skeletal abnormalities. Fetal findings were limited to an increase in the incidences of hydronephrosis of the kidney and enlarged liver in the 30,000 ppm group. A relatively high background incidence was found in this strain of rat for hydronephrosis (fetal incidence and range: 4/1,385 and 0.00%–0.81%), along with dilated renal pelvis (fetal incidence and range: 6/1,385 and 0.00%–1.06%), distended ureter (fetal incidence and range: 151/1,385 and 4.83%–15.36%), and hydroureter (fetal incidence and range: 11/1,385 and 0.17%–2.83%). Moreover, the background incidence of some findings (e.g., dilated renal pelvis and/or ureter) could be greater in fetuses than in pups, suggesting that these changes might be transient.^18,119,120 At necropsy of the F₂ offspring on PND 28, dilation of the renal pelvis was observed grossly in six rats in the 30,000 ppm group and in one F₂ rat in the control group. No incidences of hydronephrosis were observed in F₂ pups at necropsy; nevertheless, the observed fetal findings are consistent with the finding that the kidney and liver are target tissues for 2H4MBP-mediated toxicity.

In the current study, 2H4MBP exposure was associated with minimal apparent responses on litter size (fetal or PND 0) and fewer corpora lutea. A similar decrease in the numbers of corpora lutea and implants has also been observed at 30,000 ppm in the NCTR fertility and early embryonic development study, in which female dosing started two weeks prior to cohabitation through GD 6. No apparent responses were observed in the NCTR embryo-fetal toxicity study in which dosing is for a shorter duration (GD 6–15).⁹⁷ If 2H4MBP-related, this difference in response may be the result of the longer duration of exposure. The observed EE exposure-related decreases on PND 0 live F₂ litter size in the reproductive performance cohort, and GD 0 in the prenatal cohort (as well as total number of implants) is consistent with what has been observed in the 0.05 ppm EE group in the NCTR fertility and early embryonic development study.⁹⁷ These similarities demonstrate the consistency of responses observed with conducting a single study versus conducting three independent studies that would necessitate the use of more animals.