Drinking water with uranium below the U.S. EPA water standard causes estrogen receptor-dependent responses in female mice

Abstract

Background: The deleterious impact of uranium on human health has been linked to its radioactive and heavy metal-chemical properties. Decades of research has defined the causal relationship between uranium mining/milling and onset of kidney and respiratory diseases 25 years later.

Objective: We investigated the hypothesis that uranium, similar to other heavy metals such as cadmium, acts like estrogen.

Methods: In several experiments, we exposed intact, ovariectomized, or pregnant mice to depleted uranium in drinking water [ranging from 0.5 microg/L (0.001 microM) to 28 mg/L (120 microM).

Results: Mice that drank uranium-containing water exhibited estrogenic responses including selective reduction of primary follicles, increased uterine weight, greater uterine luminal epithelial cell height, accelerated vaginal opening, and persistent presence of cornified vaginal cells. Coincident treatment with the antiestrogen ICI 182,780 blocked these responses to uranium or the synthetic estrogen diethylstilbestrol. In addition, mouse dams that drank uranium-containing water delivered grossly normal pups, but they had significantly fewer primordial follicles than pups whose dams drank control tap water.

Conclusions: Because of the decades of uranium mining/milling in the Colorado plateau in the Four Corners region of the American Southwest, the uranium concentration and the route of exposure used in these studies are environmentally relevant. Our data support the conclusion that uranium is an endocrine-disrupting chemical and populations exposed to environmental uranium should be followed for increased risk of fertility problems and reproductive cancers.

Keywords: Navajo reservation; depleted uranium; endocrine disruption; estrogen; estrogen receptor; female reproduction; heavy metal.

Figure 1

Effects of UN at 0.5, 2.5, 12.5, or 60 μg/L (0.001, 0.005, 0.025, or 0.120 μM U, respectively) on dam follicle populations and in utero exposed pup ovary primordial follicles. B6C3F₁ dams were exposed to control tap water or U in drinking water for 30 days before mating and through gestation. Ovaries from dams (A) and pups (B) were removed on the day of birth. Values shown are mean ± SE (n = 7–11). *Significantly different compared with controls (p < 0.05, ANOVA).

Figure 2

Effect of UN or DES alone and in combination with ICI 182,780 on uterine weight in ovariectomized C57Bl/6J mice. (A) Uteri were removed after 30 days of exposure, and wet weights were recorded and normalized to body weight; values shown are mean ± SE (n = 5–6). (B) Uteri were removed after 10 days of exposure, and wet weights were recorded and normalized to body weight; values shown are mean ± SE (n = 6–7). Different letters (a, b, c) indicate significant differences among exposure groups (p < 0.005). *Significantly different compared with other exposure groups (p < 0.001).

Figure 3

Uterine luminal epithelial cell growth in ovariectomized C57Bl/6J mice stimulated by UN or DES in drinking water for 30 days. (A) Cell height in uteri collected and prepared for scanning electron microscopy; values shown are mean ± SE (n = 5 uteri at 40 measurements from each tissue). Representative scanning electron microscopy images at the same magnification of uterine epithelial cell layers from tap water control (B), 0.19 μM DES (C), or 0.12 μM U (D). Arrows highlight epithelial cell height in DES-exposed (C) and U-exposed (D) ovariectomized mice. *Significantly different compared with control (p < 0.0001).

Figure 4

Effect of UN in drinking water on VO and presence of cornified vaginal cells. Ovariectomized C57Bl/6J mice (50 days of age) were exposed to control tap water, 0.19 μM DES, or 0.06 or 0.12 μM U for 10 days, or one of these doses plus vehicle or 500 μg/kg ICI 182,780 in vehicle. (A) Mice were examined daily for VO from 50 days of age to the day of vaginal opening; values shown are mean day of VO ± SE (n = 6–7). (B) Vaginal cell cornification determined from vaginal smears collected daily; the presence and absence of vaginal cornified cells were analyzed by chi-square test (p < 0.05). *Statistically significant compared with control (p < 0.05 by Fisher’s exact test). **Significantly different from control (p < 0.001).