Johnson, MD, N Kenney, A Stoica, L Hilakiva-Clarke, B Singh, G Chepkko, R Clarke, PF Sholler, AA Lirio, C Foss, R Reiter, B Trock, S Paik and MB Marin. 2003. Cadmium mimics the in vivo effects of estrogen in the uterus and mammary gland. Nature Medicine, online 13 July 2003.

With these findings, Johnson et al. confirm earlier indications that cadmium acts an estrogen mimic report that the action occurs with detectable effects at astoundingly low levels of cadmium in the affected tissues in the rats they studied. Direct measurements of cadmium confirmed that effects were caused at concentrations in tissue of one hundred-thousandth of a part per billion, or 1/100th of a part per picogram, or less. The authors speculate that cadmium effects on mammary gland development may be important to causation of breast cancer.

What did they do? Johnson et al. conducted a series of relatively classic experiments to test the estrogenicity of compounds in live animals. Importantly, however, they conducted their experiments using very low levels of cadmium, and complemented their measurements of effects with direct measurements of cadmium in the tissues affected by exposures.

They asked 3 questions:

Does exposure to cadmium affect organs targeted by estrogen? To answer this they first surgically removed the ovaries of young female rats, allowed them to recover, and then injected the rats with a single dose of 5 µg/kg cadmium. One group of rats serving as a positive control instead received a pellet implant releasing 60µg/kg/day estradiol. A third group was injected with both cadmium and the anti-estrogen, ICI-182-780. Responses in each of these groups was compared to untreated controls, whose ovaries had also been removed.

The uterine responses in these animals were examined 4 days later, and the mammary gland responses 4 and 14 days later. At the levels of exposure used in the experiments, there were no signs of overt toxicity, nor alterations in body weight.

Uterine weight increased 1.9-fold in cadmium-treated animals compared to controls. The estradiol positive control also increased relative to untreated control, by 3.8 times. Exposure to ICI-182-780 eliminated the cadmium effect, indicating it is mediated by an estrogen receptor. Histological examination of the uterine tissues was also consistent with an estrogenic response, and differed substantially from the control animals.

Responses in mammary tissue were also consistent with estrogen exposure. Epithelial density in cadmium-exposed animals increased significantly compared to controls, and similarly to animals exposed to estradiol. As with uterine tissue, the response was eliminated by adding the anti-estrogen ICI-182-780.

Does cadmium mimic the effects of estradiol on gene expression? Johnson et al. measured the effect of cadmium on gene activation of genes well established to be under estrogenic control, PGR and C3. To do this they measured changes in the amount of messenger RNA produced by exposure to estradiol and cadmium. Changes were caused by activation of the genes that produce the messenger RNA.

Using estradiol exposure as a positive control, they first found that estradiol treatment produced a 3-fold increase in PGR messenger RNA and a 124-fold increase in C3 mRNA in the uterus compared to controls (untreated females with ovaries removed). In mammary glands, estradiol increased PGR mRNA and C3 mRNA by 42-fold and 416-fold, respectively.

In animals exposed to cadmium the increases compared to controls were also large: uterine expression of PGR and C3 mRNA increased 2-fold and 12-fold respectively. In mammary glands the increases were 9-fold and 16-fold compared to controls. These changes were highly significant statistically, and were blocked by adding ICI, the anti-estrogen, indicating the responses are mediated by the estrogen receptor.

Does exposure to cadmium in the womb affect development? Johnson et al. here took advantage of the well-established impacts of in utero exposure to estrogen on the rate of sexual development, mammary gland development and body weight. Estradiol exposed females reach sexual maturity faster than unexposed (or less exposed). In their experiments, cadmium mimicked the effect of estradiol on each of these endpoints. At low exposure levels, cadmium decreased the time needed to reach sexual maturity, increased adult body weight, and altered the structure of the mammary glands.

What does it mean? Most toxicological studies heretofore with cadmium have worked in the range of 1-5 milligrams per kilogram (parts per million) and higher. Johnson et al. chose instead to use cadmium levels in these experiments relevant to human environmental exposures. They delivered doses of 0.5 - 60 micrograms per kilogram (parts per billion), less than 1% of the usual level. At this low range, their experiments revealed dramatic responses to cadmium mediated by the estrogen receptor. The prior work at high doses was uninformative about these low dose responses, a pattern explored from a theoretical and regulatory perspective by Welshons et al. earlier in 2003.

Johnson et al.'s measurements went one step farther than simply knowing how much cadmium was delivered by injection. Using a highly sensitive technique, they measured cadmium levels in the affected tissues directly. This technique is capable of measure down well below 1 part per billion. They found the cadmium levels in mammary gland tissue and in the uterus to often be below the limits of detection using their instruments, even in treated animals with responses caused by cadmium. "When detectable in the uterus or mammary gland, the amount of cadmium was approximately 10^-2 pg per g of tissue (10^-5 parts per billion)." This staggering result reinforces the conclusions by Welshons et al. that traditional studies of contaminants at toxicological levels can be expected to miss responses at lower, "physiological" levels mediated by hormone receptors.

Johnson et al. note that the range of exposures they use in their experiments are similar to dietary guidelines from the World Health Organization and to estimates of daily dietary exposures in the US, Germany, the UK and Sweden. The highest exposures occur in children between the ages of 1 and 6 years. They also note that cadmium levels found typically in human breast tissue are far above the levels measured in animals exposed in their experiments, in which mammary gland development had been altered significantly, in ways consistent with heightened risk of mammary gland tumors. They suggest it would be appropriate for more research into links between cadmium exposure and breast cancer risk.