Rajapakse, N, D Ong and A Kortenkamp. 2001. Defining the Impact of Weakly Estrogenic Chemicals on the Action of Steroidal Estrogens. Toxicological Sciences 60: 296-304.
Rajapakse et al. address one of the central questions of endocrine disruption: can "weakly" estrogenic substances present in mixtures at low levels exert an impact on hormonal function in the presence of estrogen? This is important because most of the contaminants with estrogen-like activity (xeno-estrogens) are much less potent than estrogen itself--often thousands of times less powerful--and estrogen is virtually always present, at least at some level, in a developing fetus. Skeptics of the potential impacts of endocrine disruption point to the comparatively weak potency of xeno-estrogens as a fatal flaw in proposals that contaminants like bisphenol A could have health consequences.
In an elegant set of experiments, Rajapkse et al. demonstrate that the additive interactions of weak xeno-estrogens significantly change cellular responses to 17ß-estradiol. Their results effectively eliminate the argument that xeno-estrogens can't be important because they are only "weak estrogens."
What
did they do?
Rajapaske et al. used bisphenol A and o,p'-DDT in a series
of cell-culture experiments that first established the dose-response relationship
for each of the two contaminants separately across a broad range of doses.
The assay they used to quantify these relationships was the yeast estrogen
screen, a now-standard method of determining estrogenic activity (Routledge
and Sumpter 1996). This assay employs yeast cells into which human
estrogen receptors have been integrated; when an estrogenic substance
binds with the receptor, the yeast secretes an enzyme that can be measured
easily and reliably. While the assay has many limitations with respect
to what it can reveal about how estrogenic responses are integrated into
whole-body mammalian systems, it is an excellent and highly reliable way
exploring hormonal interactions at the molecular level. And it is at this
molecular level that the "weak estrogen" criticism described
above is relevant.
After establishing single contaminant dose-response curves, Rajapaske et al. introduced combinations of the contaminants into the assay along with 17ß-estradiol. They predicted what responses should occur based on the single-contaminant experiments, assuming additivity of effects of the contaminants, and compared the predictions with what they actually observed when the contaminants were combined in mixtures.
What
did they find?
First, as expected, 17ß-estradiol was far more potent at stimulating
enzyme release, i.e., the estrogenic response, than either BPA
or o,p'-DDT. In fact, 17ß-estradiol was "approximately
30,000 times more potent than BPA and about 17,000 times more potent than
o,p'-DDT." Dose response curves for each of the three compounds
were similar to prior studies, although DDT's effect was apparent at somewhat
lower levels.
In their experiments with mixtures, Rajapakse et al. found that their additive model of effects was largely accurate in its predictions of estrogenicity of the mixtures. For example, when BPA was added to to the assay in combination with 17ß-estradiol, BPA enhanced the assay response, if its concentration was sufficiently high relative to 17ß-estradiol. They were able to calculate what "sufficiently high" was on the basis of the relative potency of the compounds. Because 17ß-estradiol was 30,000 times more potent than BPA, an effect of BPA became apparent in the experiments when the concentration of BPA was thousands of times higher than estradiol. At equivalent concentrations, no effect was apparent.
What
does this mean?
According to the authors:
| "There can be no doubt that weak xenoestrogens such as BPA or o,p'-DDT, when combined with 17ß-estradiol, are able to contribute to estrogenic mixture effects. We show that the impact of xenoestrogens on the actions of the steroid hormone depends on the mixture ratio and on its potency relative to 17ß-estradiol. When combined at approximately equieffective concentrations, substantial modulations of the effects of 17ß-estradiol by the xenoestrogen become discernible." |
As Rajapaske et al. discuss, xenoestrogens by themselves may approach concentrations "equieffective" with 17ß-estradiol only under unusual circumstances, but they regularly occur in mixtures with other xenoestrogens. They go on to state:
| However, human tissues contain many compounds with estrogenic activity. On the basis of our studies, it appears conceivable that a multitude of xenoestrogens, when present in sufficient number and/or concentration, might in principle act together to impact on the actions of steroidal estrogens. Whether such impacts will be physiologically relevant remains to be seen. Definitive answers to this question are currently hampered by our lack of knowledge about the full spectrum of estrogenic agents in human tissues. |
These guarded words mean that it scientifically indefensible to use the "weak estrogen" argument to dismiss concerns about endocrine disruption. Considerable research is required before it will be possible to know when mixtures of xenoestrogens are sufficiently potent, and when they are not.
In the meantime, as the Royal Society of London has concluded, it is prudent to limit exposures of people to hormonally-active chemicals.