Environmental estrogens

There is probably no bigger controversy in the field of toxicology than whether chemical pollutants are responsible for congenital malformations in wild animals, the decline of sperm counts in men, and breast cancer in women. One of the sources of these pollutants is pesticide use. Americans use some 2 billion pounds of pesticides each year, and some pesticide residues stay in the food chain for decades. Although banned in the United States in 1972, DDT has an environmental half-life of about 100 years (Nature Genetics 1995). Recent evidence has shown that DDT (dichloro-diphenyl-trichloroethane) and its chief metabolic by-product, DDE (which lacks one of the chlorine atoms), can act as estrogenic compounds, either by mimicking estrogen or by inhibiting androgen effectiveness (Davis et al. 1993; Kelce et al. 1995). DDE is a more potent estrogen than DDT, and it is able to inhibit androgen-responsive transcription at doses comparable to those found in contaminated soil in the United States and other countries. DDT and DDE have been linked to such environmental problems as the decrease in the alligator populations in Florida, the feminization of fish in Lake Superior, the rise in breast cancers, and the worldwide decline of human sperm counts (Carlsen et al. 1992; Keiding and Skakkebaek 1993; Stone 1994; Swan et al. 1997). Guillette and co-workers (1994; Matter et al. 1998) have linked a pollutant spill in Florida's Lake Apopka (a discharge including DDT, DDE, and numerous other polychlorinated biphenyls) to a 90% decline in the birthrate of alligators and to the reduced penis size in the young males.

Dioxin, a by-product of the chemical processes used to make pesticides and paper products, has been linked to reproductive anomalies in male rats. The male offspring of rats exposed to this planar, lipophilic molecule when pregnant have reduced sperm counts, smaller testes, and fewer male-specific sexual behaviors. Fish embryos seem particularly susceptible to dioxin and related compounds, and it has been speculated that the amount of these compounds in the Great Lakes during the 1940s was so high that none of the lake trout hatched there during that time survived (Figure 21.23; Hornung et al. 1996; Zabel and Peterson 1996; Johnson et al. 1998).

Some estrogenic compounds may be in the food we eat and in the wrapping that surrounds them, for some of the chemicals used to set plastics have been found to be estrogenic. The discovery of the estrogenic effect of plastic stabilizers was made in a frightening way. Investigators at Tufts University Medical School had been studying estrogen-responsive tumor cells. These cells require estrogen in order to proliferate. Their studies were going well until 1987, when the experiments suddenly went awry. Then the control cells began to show the high growth rates suggesting stimulation comparable to that of the estrogen-treated cells. Thus, it as if someone had contaminated the medium by adding estrogen to it. What was the source of contamination? After spending four months testing all the components of their experimental system, the researchers discovered that the source of estrogen was the plastic tubes that held their water and serum. The company that made the tubes refused to tell the investigators about its new process for stabilizing the polystyrene plastic, so the scientists had to discover it themselves. The culprit turned out to be p-nonylphenol, a chemical that is also used to harden the plastic of the plumbing tubes that bring us water and to stabilize the polystyrene plastics that hold water, milk, orange juice, and other common liquid food products (Soto et al. 1991; Colburn et al. 1996). This compound is also the degradation product of detergents, household cleaners, and contraceptive creams. A related compound, 4-tert-pentylphenol, has a potent estrogenic effect on human cultured cells and can cause male carp (Cyprinus carpis) to develop oviducts, ovarian tissue, and oocytes (Gimeno et al. 1996).

Some other environmental estrogens are polychlorinated biphenyls (mentioned earlier). These PCBs can react with a number of different steroid receptors. PCBs were widely used as refrigerants before they were banned in the 1970s when they were shown to cause cancer in rats. They remain in the food chain, however (in both water and sediments), and they have been blamed for the widespread decline in the reproductive capacities of otters, seals, mink, and fish. Some PCBs resemble diethylstilbesterol in shape, and they may affect the estrogen receptor as DES does, perhaps by binding to another site on the estrogen receptor. Another organochlorine compound (and an ingredient in many pest- icides) is methoxychlor. Pickford and colleagues (1999) found that methoxychlor blocked progesterone-induced oocyte maturation in Xenopus at concentrations that are environmentally relevant. This would severely inhibit the fertility of the frogs, and it may be a component of the worldwide decline in amphibian populations.

Some scientists, however, say that these claims are exaggerated. Tests on mice had shown that litter size, sperm concentration, and development were not affected by environmental concentrations of environmental estrogens. However, recent work by Spearow and colleagues (1999) has shown a remarkable genetic difference in the sensitivity to estrogen among different strains of mice. The strain that had been used for testing environmental estrogens, the CD-1 strain, is at least 16 times more resistant to endocrine disruption than the most sensitive strains such as B6. When estrogen-containing pellets were implanted beneath the skin of young male CD-1 mice, very little happened. However, when the same pellets were placed beneath the skin of B6 mice, their testes shrunk, and the number of sperm seen in the seminiferous tubules dropped dramatically (Figure 21.24). This widespread range of sensitivities has important consequences for determining safety limits for humans.

Environmental thyroid hormone disruptors

The structure of some PCBs resembles that of thyroid hormones (Figure 21.25), and exposure to them alters serum thyroid hormone levels in humans. Hydroxylated PCB were found to have high affinities for the thyroid hormone serum transport protein transthyretin, and can block thyroxine from binding to this protein. This leads to the elevated excretion of the thyroid hormones. Thyroid hormones are critical for the growth of the cochlea of the inner ear, and rats whose mothers were exposed to PCBs had poorly developed cochleas and hearing defects (Goldey and Crofton in Stone 1995; Cheek et al. 1999).

Deformed frogs: pesticides mimicking retinoic acid?

Throughout the United States and southern Canada there is a dramatic increase in the number of deformed frogs and salamanders in what seem to be pristine woodland ponds (Figure 21.26A; Ouellet et al. 1997). These deformities include extra or missing limbs, missing or misplaced eyes, deformed jaws, and malformed hearts and guts. Some of these malformations (especially the limb anomalies, see Figure 3.23) may be due to trematode infestation, but other malformations do not seem to be explainable by that route. Some lakes containing high proportions of malformed frogs do not appear to be infested with trematodes, and water from these lakes is able to disrupt development in frog eggs that are placed into it. It is not known what is causing these disruptions, but there is speculation (see Hilleman 1996, Ouellet et al. 1997) that pesticides (sprayed for mosquito and tick control) might be activating or interfering with the retinoic acid pathway. The spectrum of abnormalities seen in these frogs resembles those malformations caused by exposing tadpoles to retinoic acid (Crawford and Vincenti 1998; Gardiner and Hoppe 1999).

New research has focused on compounds such as methoprene, a juvenile hormone mimic that inhibits mosquito pupae from metamorphosing into adults. Since vertebrates do not have juvenile hormone, it was assumed that this pesticide would not harm fish, amphibians, or humans. This has been found to be the case: methoprene, itself, does not have teratogenic properties. However, upon exposure to sunlight, methoprene breaks down into products that have significant teratogenic activity in frogs (Figures 21.26B,C). These compounds have a structure similar to that of retinoic acid and will bind to the retinoid receptor (Harmon et al. 1995; La Claire et al 1998). When Xenopus eggs are incubated in water containing these compounds, the tadpoles are often malformed, and show a spectrum of deformities similar to those seen in the wild (La Claire et al. 1998).

Chains of causation

Whether in law or science, establishing chains of causation is a demanding and necessary task. In developmental toxicology, numerous endpoints must be checked, and many different levels of causation have to be established (Crain and Guillette 1998; McNabb et al. 1999). For instance, one could ask if the pollutant spill in Lake Apopka was responsible for the feminization of male alligators. To establish this, one has to ask how might the chemicals in the spill contributed to reproductive anomalies in males alligators and what would be the consequences of that happening. Table 21.3 shows the postulated chain of causation. After observing that the population level of the alligators has declined, at the organism level one discovers the unusually high levels of estrogens in the female alligators, the unusually low levels of testosterone in the males, and the decrease in the number of births among the alligators. On the tissue and organ level, the decline in birth rate can be explained by the elevated production of estrogens from the juvenile testes, the malformation of the testes and penis, and the changes in enzyme activity in the female gonads. On the cellular level, one sees ovarian abnormalities that correlate with unusually elevated estrogen levels. These cellular changes, in turn, can be explained at the molecular level by the finding that many of the components of the pollutant spill bind to the alligator estrogen and progesterone receptors and that they are able to circumvent the cell's usual defenses against overproduction of steroid hormones (Crain et al. 1998).

While there is little dispute about the damage to wildlife being wrought by endocrine disrupting chemicals, it is difficult to document the effects of environmental compounds on humans. There is enormous genetic variation in the human species, and one cannot perform controlled experiments to determine the effect of any particular compound on a human population. Rather, we are exposed to “cocktails” consisting of different compounds ingested at different times. There is a great deal more research that needs to be done on the biochemistry of these compounds, their effects on development, and the epidemiology of developmental abnormalities. At the moment, evidence coming from animal studies suggests that humans and natural animal populations are at risk from these hormonal modulators, but not all the needed data are in.