Howdeshell, KA, PH Peterman, BM Judy, JA Taylor, CE Orazio, RL Ruhlen, FS vom Saal, and WV Welshons 2003. Bisphenol A is released from used polycarbonate animal cages into water at room temperature. Environmental Health Perspectives doi:10.1289/ehp.5993.

BPA leaching out of water bottles causes aneuploidy

Howdeshell et al. report that significant amounts of bisphenol A leaches into water that comes in contact with polycarbonate animal cages. Their results raise the specter of widespread inadvertent and hitherto unsuspected contamination of generations of animal testing by an estrogenic compound, bisphenol A. The result may be that many experiments have provided false assurances of safety about hormonal impacts of contaminants.

These results may also help explain inconsistencies between different experiments, given that the type of animal housing differs among labs and the rate of leaching is affected by the age of the polycarbonate cages.

How could this be? Most hormonal systems have maximal response levels to exposure, demonstrating what scientists call an asymptote.

As dose increases, there comes a point at which further increases in dose no longer increase the effect. In hormone systems this occurs by the time that all available hormone receptors have bound to hormone molecules.

.

Consider the following two circumstances (graphs below):

In this hypothetical experiment, there is an asymptotic relationship between the dose and the response. In the first version, to the left, there is no inadvertent estrogenic contamination. Thus controls have 0 response. Addition of an experimental treatment provokes an increase in response. The difference between E and C is the observed effect.

In the graph to the right, the "control" has significant estrogenic contamination. Hence "C", the control response, is significantly greater than true 0 (although the experimenter is unaware that they are not measuring "true 0") and close to the maximum response, "E", that can be achieved via experimental treatment.

The difference between "E" and "C" is much smaller than that between E and C (in the first graph) and more likely to be insignificant, particularly with small sample sizes or in experiments with statistical variability.

If bisphenol A contamination has been shifting control animals to the right along the dose curve, then experimenters are less likely to have found the estrogenic effects they would have observed without the bisphenol A contamination.

This effect would be seen in experiments with compounds that share molecular mechanisms with bisphenol A, e.g., an ability to bind with the estrogen receptor.

What did they do? Howdeshell et al. put purified water in different types of cages at room temperature and neutral pH, let the water sit for one week, and then measured both the amount of BPA in the water using high pressure liquid chromatography and the estrogenicity of the water using a standard assay (the MCF-7 human breast cancer cell proliferation assay).

Cage types were new vs. old polycarbonate, new polysulfone, old polypropylene, and glass. Polycarbonate and polysulfone cages are made with bisphenol A. They also compared the uterine weight of mice reared shortly after weaning for one week in old polycarbonate vs. old polypropylene cages. Animals in polycarbonate cages drank water from old polycarbonate water bottles, while those in polypropylene drank water from glass bottles.

The cages were rinsed prior to the experiments to remove prior leaching, and were not scrubbed for the duration. Thus any accumulation was a result of passive leaching during the week of the experiment.

What did they find? Old polycarbonate leached significantly more bisphenol A than any other type of cage. New polycarbonate and new polysulfone leached detectable amounts, approximately one-tenth or less that of the old polycarbonate. No bisphenol A was detectable in the water from the glass or polypropylene cages, although trace amounts of nonylphenol, another estrogenic substance, were measured in the latter.

In the MCF-7 assay, all water samples from old polycarbonate cages induced proliferation. By adding an anti-estrogen to these experiments, they were able to block the response, indicating it was caused by an estrogenic substance in the water. Additional chemical analysis strongly indicated that the estrogenicity was due to BPA.

Water from new polycarbonate and new polysulfone cages did not stimulate proliferation. Water from one of the old polypropylene cages, however, did produce a slight elevation in response, most likely because of the trace amounts of nonylphenol (noted above).

In the mouse experiments, "prepubertal uterine wet weight was approximately 16% heavier in females housed in the used polycarbonate caging with used polycarbonate water bottles relative to mice housed in used polypropylene cages, although the results were not statistically significant (P =0.31). The authors note, however, that uterine weight is not a very sensitive whole-animal assay for BPA exposure. Patterns of mammary duct development are approximately 4000 times more sensitive.

What does it mean? According to the authors: "Our findings here suggest that aquatic laboratory animals may be exposed to sufficient BPA due to leaching from worn polycarbonate caging to result in a significant impact on reproductive parameters. This prediction is based on a number of recent reports of significant effects at very low concentrations of BPA in frogs, fish and mollusks."

As to mammals like rats and mice, the cornerstones of regulatory toxicology, while the uterine results were not significant, it is likely other assays such as in utero effects on prostate weight or time to puberty would have yielded significant results, as these, like mammary duct development, are far more sensitive to BPA.

Howdeshell et al. also observe that polycarbonate is not the only source of bisphenol A in the laboratory: "BPA migration into human serum has been reported with the use of polycarbonate and polysulfone plastic hemodialysis equipment. Another potential route of BPA exposure in the laboratory is polyvinyl chloride (PVC) pipes used in the supply of tap water; BPA is added as a stabilizer in the production of PVC products.

They recommend that researchers be aware of the potential confounding effects of BPA contamination and take extra measures to minimize this possibility, including filtering water.

The key question, which they do not even begin to address, is how many negative results now pervade the scientific and regulatory literature because of BPA leaching from laboratory equipment. Their findings make this a strong possibility.