The Causative Chemicals

Probably the most difficult problem in this difficult area of research is to link exposure to specific chemicals to the effects observed in wild populations of organisms (such as those described above); that is, to identify the chemicals causing endocrine disruption in wildlife. The reason for this is that we know relatively little about what chemicals (either natural or man-made) are present in the aquatic environment, or about their concentrations (which will be variable, and depend on the specific situation). The problem is easily understood when it is realised that around 70,000 man-made chemicals are in everyday use. Many, if not most, of these will enter the aquatic environment, where they will degrade at varying rates, and varying extents, to produce many more chemicals. In addition there will be many, probably very many, natural chemicals present. This very large number of chemicals makes it likely that we will never have a complete picture of what chemicals are present, and at what concentrations. Even if all the chemicals in a particular aquatic environment could be identified and their concentrations determined, this may not help a great deal. In cases where direct exposure (e.g. water to organism) occurs, then knowledge of what is present in the water will be very helpful, but in cases where indirect exposure occurs (as might well occur in the case of top predators, which probably ingest contaminants in their food, and subsequently pass them to their offspring via the yolk in their eggs), knowledge of contaminant concentrations in the water and/or sediment might not be very informative, and may even mislead.

One of the most useful approaches adopted to date has been that of Toxicity Identification and Evaluation (TIE), which incorporates a chemical fractiona-tion procedure with a method of detecting endocrine activity in the fractions. Essentially, a very complex effluent is split into fractions of decreasing complexity, with each fraction being analysed for oestrogenic activity (or whatever endocrine activity is of interest). Fractions identified as active are separated further, until they are simple enough to be analysed by a technique such as gas chromatography-mass spectrometry, leading to the identification of the active chemicals.

Such a TIE approach was used to identify the main oestrogenic chemicals in the effluents of seven UK STWs receiving primarily domestic influent. These were found to be the natural hormones oestradiol and oestrone, and the syn thetic hormone ethinyl oestradiol [39], which presumably were excreted by (primarily) women. These steroids were detected as 'free' hormones, yet they are excreted as conjugates (primarily sulfates and glucuronides) which are biologically inactive. This suggested that de-conjugation must occur in the sewage system, a suggestion supported by the study of Panter et al. [40], who showed that conjugated steroids are very readily de-conjugated (to the active steroid) by minimal microbial activity.

To verify that the right chemicals had been identified, laboratory experiments were conducted in which fish were exposed to varying concentrations of these oestrogens, centred around the concentrations present in effluents. It was found that vitellogenin concentrations were elevated in a dose-dependent manner [41], thus demonstrating that environmentally-relevant concentrations of natural and synthetic oestrogens are alone sufficient to account for the elevated vitellogenin concentrations observed in both wild fish [2] and caged fish [24] living downstream of STW discharges in British rivers. Subsequent research from a number of other European countries has verified that natural and synthetic oestrogens are present in STW effluent at concentrations similar to those reported to be present in UK effluents [42-44].

Although this particular example demonstrated the power of the TIE approach, it was aided both by the fact that the chemicals of interest were known to be oestrogenic (because they led to elevated vitellogenin concentrations [22]), and that relatively straightforward assays for oestrogens are widely available. However, if it is unclear what type of endocrine activity is responsible for the effects observed, then the TIE approach becomes problematical, because it is not clear what bioassay should be used to guide the chemical fractionation. Ideally, the bioassay should be based upon the effects observed in the wild animal (perhaps, for example, vitellogenin induction), but this is rarely feasible; instead, a more simple (usually in vitro) assay is used. Such an approach can, however, mislead, because most relatively simple in vitro assays for endocrine activity do not represent factors such as bioaccumulation and metabolism, which occur in vivo and can be critical factors in determining the effects of chemicals. Furthermore, although many different in vitro assays for oestrogens (and mimics) are widely available, this is not true for other endocrine activities - for example, androgenic activity. Thus, further advances in the techniques used to measure different types of endocrine activity are needed before the TIE approach can be used widely. When the effects observed in wild animals are caused by many chemicals having different mechanisms of action (as may often be the case), then again the TIE approach is made much more difficult.

The only other realistic approach (besides TIE) to identifying the causative chemicals is to screen chemicals in different, usually in vitro, assays to identify those possessing appropriate endocrine activity. Rather than conduct a random screen (which would anyway be impractical, considering the very large number of chemicals present in the aquatic environment), the chemicals tested should be those likely to be present in the particular environment in which the affected organisms were found. For example, if pesticides are suspected of causing the adverse effects (as is the case with alligators on Lake Apopka), then it would obviously be appropriate to screen a selection of such chemicals. This approach has been quite widely used, sometimes because it is readily "doable", rather than because it is the most appropriate to the situation under investigation. Thus, for example, Jobling et al. [45] used such an approach to demonstrate that many common aquatic pollutants, including some phthalates, are weakly oestrogenic, findings that have since been corroborated by many other studies. However, despite phthalates being ubiquitous aquatic pollutants, there is presently no evidence showing that they cause endocrine disruption in wild fish (or any wildlife). This example demonstrates the difficulties associated with this approach; the ease of conducting many in vitro assays has led to many chemicals being tested in such assays, and to quite a few being shown to possess one, or more, types of endocrine activity. However, demonstrating that a particular chemical is present in the environment at concentrations which cause endocrine disruption to wildlife has proved infinitely more difficult. To date, this "shotgun" approach to identifying chemicals with endocrine activity has not led, to my knowledge, to a single example of the identity of a chemical causing endocrine disruption in wildlife actually being revealed.

These problems serve to highlight just some of the many difficulties that are encountered in trying to identify the chemicals responsible for endocrine disruption in aquatic organisms. They help explain why, with few exceptions (e.g. TBT and imposex in molluscs), the causative chemicals remain elusive. Yet, unless the chemical, or mixture of chemicals, is identified, appropriate laboratory studies (which can best address many aspects of endocrine disruption) cannot be undertaken, and remedial action, to reduce or eliminate the problem, cannot be considered.

Most of the issues associated with identifying the chemical, or chemicals, causing endocrine disruption in wildlife are summarised in Table 1. This considers just three chemicals (17^-oestradiol, ethinyl oestradiol and nonylphenol) which probably all play a role in causing the "feminisation" reported in freshwater fish in the UK. The factors that need to be considered are the potencies (efficacies) of these chemicals, their concentrations in the aquatic environment, their

Table 1. Factors to consider when attempting to assess the relative contributions of different oestrogenic chemicals to the "feminising" effects of effluent from sewage treatment works. Some of the information is taken from publications, but some is my own "best guess", based sometimes on relatively little information

Table 1. Factors to consider when attempting to assess the relative contributions of different oestrogenic chemicals to the "feminising" effects of effluent from sewage treatment works. Some of the information is taken from publications, but some is my own "best guess", based sometimes on relatively little information


Concentration in effluent

Bioconcentration factor


Minimum effective concentration in vivo



Low |g/l



10 |g/l


(days ^ months)





10 ng/l

Short (hours?)


Low ng/l


Very high

1 ng/l




fate and behaviour in the aquatic environment, and their fate and behaviour in aquatic organisms (that is, their persistence, accumulation and metabolism within the organisms of interest). Note that, even with these three chemicals, which have been extensively studied in the last few years, quite a lot of the data required are not available, or supported by very little information. Nevertheless, Table 1 illustrates the difficulty in determining whether an extremely potent oestrogen, ethinyl oestradiol, but one which is present in the aquatic environment at extremely low concentrations (parts per trillion, or lower), is more, or less, important than nonylphenol, a rather weak oestrogen, but one which is widespread and often present at appreciable concentrations (parts per billion in solution and higher in sediments). Even this type of approach is very simplistic (and probably unrealistic), because aquatic wildlife are very rarely ever exposed to an individual chemical, but instead are nearly always exposed to highly complex, undefined, mixtures of chemicals simultaneously. Such an exposure scenario makes it extremely difficult to identify the chemical(s) causing the adverse effects; in many cases I suspect that it may never be possible to pin-down the causative chemicals (this has been achieved surprisingly rarely, despite more than two decades of research on endocrine disruption in wildlife). Further, the exposure regime is usually not static, but instead is constantly changing as the use of chemicals waxes and wanes, their rates of release into the environment change, and waste treatment processes (prior to release of effluent into the aquatic environment) change and improve. Thus, when effects are observed in wildlife, especially in adult animals, the exposure responsible for the effects may no longer persist by the time the (long-lived) effects are noticed, probably making it impossible ever to link cause with effect with certainty.

0 0

Post a comment