Hazard Identification

Hazard identification is the process of determining whether exposure to an agent (e.g. a specific chemical) can cause a specific adverse health outcome. Hazards may be identified through epidemiological investigations, toxicological studies and other means.

Epidemiology
Epidemiology is the study of populations in order to determine the frequency and distribution of disease and measure risks. Epidemiological studies can range from simple clinical observations of adverse health outcomes in clusters of patients, to descriptive studies of mortality and morbidity rates, to more analytical studies designed to address specific hypotheses concerning cause and effect relationships. Epidemiological studies have the advantage of providing information on health hazards directly in humans. Some of the limitations of these studies include the limited sensitivity of some study designs and the difficulty in the detection of small effects. The study of environmental toxicants becomes increasingly difficult, as it is not appropriate to deliberately expose human test subjects to toxicants and observe any deleterious health effects. Rather, much of the empirical evidence linking endocrine disrupting chemicals and adverse health effects is based on studies of humans accidentally exposed to toxicants (Seveso, Italy, PCB contamination) or in small clusters of the population that are exposed because of occupation (pesticide applicators) or diet (contaminated fish).

Other limitations of epidemiological studies include problems of simultaneous exposure to a variety of hazards, the potential confounding effects of unsuspected hazards present in the environment, adequacy of exposure data, and the relative insensitivity of studies of even moderately large size to detect small effects. In contrast to controlled experiments, humans are continuously exposed to a plethora of toxicants, rather than a single toxicant. Identifying an individual toxicant as a causal factor in a health concern becomes difficult as exposure to other toxicants may play a role.

A major difficulty with environmental toxicants is the protracted time scale often required for disease development. Since some degree of human exposure must take place prior to the initiation of any epidemiological investigation, adverse health effects cannot be predicted in advance of introducing a new substance into the human environment. Consequently, epidemiological studies alone are insufficient for a purely preventive regulatory model, and are perhaps more often used to support public health decisions rather than to initiate them.

Toxicology
Toxicological experiments conducted under controlled conditions in a laboratory environment are often widely employed as a means of identifying potential human health hazards. Toxicology involves the study of toxic substances on human and laboratory animals used for testing on behalf of humans. While many highly sensitive tests are available to evaluate a wide range of deleterious effects including toxicity, metabolism of chemical substances and carcinogenic effects, it is difficult to ascertain whether a substance will interact with and disrupt the normal physiology of the endocrine system. There are several reasons for this. First, toxicological experiments use a laboratory animal to 'model' the human system. Animal models are helpful for the evaluation of the toxic and carcinogenic potential of chemical agents. To properly assess how a chemical will interact with the endocrine system requires an animal model with very similar physiology to humans. Second, many of the effects attributed to endocrine disruption may not be easily extrapolated or translated from animals to humans including infertility, cancer or neurocognitive deficits due to their multifactorial etiology. Third, although there are animal tests for the evaluation of behavioural abnormalities it is not possible to determine the effects of chemicals on cognitive development.

Although providing only indirect information on human health risks, toxicological studies conducted in the laboratory can provide valuable information on the toxic potential of environmental chemicals. Further development of appropriate animal models for the study of endocrine disease and dysfunction is still required.

Ecology
Many agents have been identified as potential endocrine disrupters based on ecological studies of wildlife. There are many examples of adverse effects present in wildlife populations inhabiting environments polluted by certain toxicants. As these adverse effects are not found in similar wildlife populations inhabiting unpolluted environments, the argument is made that exposure to the pollutant has caused the adverse effect. In many wildlife case studies, however, these polluted environments contain a mixture of chemicals with a few chemicals characterized as 'endocrine disrupters'. It is therefore difficult to link the observed adverse effects with exposure to a specific chemical agent.

Identification of endocrine disrupters
Classification of a group of environmental toxicants as endocrine disrupters as proved to be one of the most challenging tasks in this area. An endocrine disrupter is an exogenous substance or mixture that alters function(s) of the endocrine system and consequently causes adverse health effects in an intact organism, or its progeny, or (sub) populations. However, those environmental toxicants proposed to cause adverse effects through disruption of the endocrine system (ie. PCBs, dioxins, phthalates) are a diverse group of chemicals with dissimilar structures, biochemical properties and proposed mechanisms of action. While the term PCBs is used to refer to any of a group of toxic, chlorinated aromatic hydrocarbons used in a variety of commercial applications, not all PCB congeners may be endocrine disrupters. Thus, chemical class cannot be used to discriminate toxicants with the potential to act as endocrine disrupter. The identification of potential endocrine disrupters requires a step-wise, systematic approach best on criteria used commonly to identify chemicals with toxic or carcinogenic properties. These include mechanism of action, structure/activity analysis, metabolism and pharmacokinetics and weight of evidence.

Mechanism of action
Endocrine disrupters are believed to act at endogenous hormone receptors in several ways. First, toxicants may bind hormone receptors thereby competitively inhibiting the endogenous hormone. By binding to hormone receptors, toxicants may trigger the normal intracellular signaling cascade that results in gene expression and protein synthesis. These toxicants, or agonists, are loosely termed "estrogens" as they assume the same role as the endogenous hormones by binding to the estrogen receptor. Estrogenic toxicants include MXC, pesticides, bisphenol A and B, chlordecone, methoxychlor, octylphenol and nonylphenol. Estrogenic activity can be determined using in vitro assays such as estrogen receptor (ER) binding, breast cancer cell proliferation and transcriptional activation. Phytoestrogens present in a variety of plants such as soy (isoflavanoids) and berries, fruits, grains, vegetables and nuts (lignans) represent another source of exposure to estrogenic chemicals.

Some estrogenic toxicants appear to interact with the endocrine system in a very complex manner. Hormone receptors are actually families of receptors with each member possessing slightly different properties. There are two major estrogen receptor family members or 'isoforms'; ER- and ER- that are localized in different tissues throughout the body. Some toxicants may trigger different effects by interacting with different receptor isoforms. One methoxychlor metabolite, for example, exhibits estrogenic activity and interacts with ER- but is also a potent ER- antagonist, thereby inhibiting estrogen activity. Methoxychlor can also inhibit androgen receptors (ARs) and is therefore also an anti-androgen.

Mammals appear to possess a single androgen receptor (AR). Vinclozolin, a fungicide, is an anti-androgen. Vinclozolin metabolites, M1 and M2, competitively inhibit endogenous androgen binding to the receptor. Unlike methoxychlor, vinclozolin and its metabolites do not also act at the estrogen receptor, though it is possible that vinclozolin may exhibit a weak affinity for the progesterone receptor. Other toxicants also exhibit anti-androgen activity including the DDT metabolite p,p'-DDE, a methoxychlor metabolite, organophosphate fenitrothion and fungicide procymidone.

Some PCBs, PCDFs and dioxins such as TCDD interact with the aryl hydrocarbon receptor (AhR) to trigger signaling pathways, growth factor expression and enzyme activity. The AhR is a receptor protein located in the cytoplasm of cells. Through its ability to interact with multiple signal transduction pathways, to induce or inhibit a variety of gene products, AhR agonists can induce a wide spectrum of biological effects.

Not all toxicants disrupt the endocrine system by interacting directly with hormone receptors. Some toxicants inhibit hormone synthesis, transport or metabolism. An important enzyme in hormone synthesis is 'aromatase' which converts androgens to estrogens. Inhibition of aromatase would increase the ratio of androgens to estrogens. Certain fungicides have been shown to cause infertility through aromatase inhibition.

Other toxicants may trigger signaling cascades leading to changes in the biochemical structure of the hormone receptor. 'Phosphorylation' is the addition of a phosphate group by an enzyme called a 'protein kinase'. Phosphorylation of a compound, such as a hormone receptor, changes the biochemical properties of the compound including its interaction with other molecules, binding properties and its function. Some hormones require accessory compounds or complexes to function properly. Another potential mechanism of action for chemical toxicants is to disrupt the release of these cellular complexes necessary for hormone action.

Thus, classification of endocrine disrupters on the basis of mechanism of action is difficult. Toxicants may disrupt the endocrine system at many levels, thereby altering the normal hormonal function in the body. Classification of an environmental toxicant as an endocrine disrupter should be based on a mechanism of action, such as those described here, that would be expected to produce adverse effects through disruption of the endocrine system.

Structure/Activity Analysis
The basis for structure/activity analysis is that the chemical structure of a compound may serve as a predictor of its activity or function. In some cases, the chemical structure of a compound can be predictive of toxicity or carcinogenicity. The number of aromatic rings in polycyclic hydrocarbons or the number of chlorine atoms in chlorinated hydrocarbons has been used to establish relative potency of certain chemicals. Structure itself is not a sufficient predictor of potential endocrine disrupting activity. Several chemical classes (organochlorines, phthalates, dioxins etc) with very different structures have all been proposed to act as endocrine disrupters.

Similarly, toxicants that mimic or block hormone activity are not always structurally similar to the endogenous hormone. A characteristic of environmental toxicants is that they lack a consistent structural motif. There is generally the presence of an aromatic ring, many chemicals contain chlorine atoms but generally it is difficult to ascribe generic structural features to a 'typical' endocrine disrupter.

Structures of typical endocrine disrupters

Competitive binding assays are used to determine whether a compound can compete with the endogenous ligand for the hormone receptor. The binding constant (Kb) serves as a measure of the affinity of the ligand for the receptor and can be used to compare the potency of a test compound with other ligands including the endogenous hormone. Binding of the hormone receptor by the putative endocrine toxicant is suggestive of endocrine disrupting activity. Receptor binding assays can use rat, mouse, or human ER or AR. Limitations of binding assays include solubility in the culture medium, inability to distinguish agonists from antagonists, lack of metabolic capability, and risk of degradation of the receptor.

Functional assays may be more predictive of endocrine disrupting activity rather than chemical structure. Screening and testing is designed initially to identify and characterize effects that enhance, mimic, or inhibit estrogenic or androgenic hormone-related processes. Ideally, functional tests that detect multiple hormone interactions, address endpoints in multiple species, and predict long-term or delayed effects would be invaluable to the characterization of potential endocrine toxicants.

The reporter gene assay is used to detect transcriptional activity induced by chemical interaction and binding at the hormone receptor. Normally, the binding of the endogenous ligand to the hormone receptor triggers transcription, or synthesis of RNA, followed by translation of the mRNA to produce a protein. Endogenous estrogen-dependent gene transcription may be difficult to detect as multiple hormone pathways may regulate transcription and translation of the gene product. Thus, detection of the protein induced by receptor activation is achieved by transfecting (introducing) a 'reporter gene' into the genome of the cell. Transcription and translation of this reporter gene produces an identifiable protein that can be easily detected, thereby confirming functional activation of the hormone receptor. These reporter assays utilize the human ER of MCF-7 cells for transcriptional regulation of a reporter gene that codes for an exogenous enzyme that can be easily measured in a cell lysate. Examples of reporter genes products commonly used include luciferase and beta-galactosidase. Upon translation, the protein luciferase, a firefly enzyme, reacts with substrate and cofactors added to culture media by emitting a flash of light that can be easily detected. Beta-galactosidase, another enzyme, reacts with substrate and cofactors present in culture producing a product that can be detected by a change in colour.

The MCF-7 cell proliferation assay is another bioassay that can be used to detect estrogenic chemicals. MCF-7 cells contain various estrogen-regulated genes that enhance proliferation (increased cell replication) in response to estrogens. Putative toxicants are added to culture media with changes in cell number are assessed after 6 days. Estrogenic substances enhance cell growth thereby inducing proliferation of MCF-7 cells in culture. However, due to cross talk in signal transduction pathways it is possible to induce proliferation without the test compound possessing any estrogenic activity.

Other steps in risk assessment consist of four steps: dose-response assessment, exposure assessment and risk characterization.