Sir

While we sympathize with some of the views expressed by Marcello Lotti and Pierluigi Nicotera in their Concepts essay “A risky business” (Nature 416, 481; 2002), the authors fail to acknowledge the wider role of toxicology in society. Toxicology is much more than the purely mechanistic studies they propose, and should not be incorporated into “the mainstream of fundamental biomedical research”. Rather, it should draw from and interact with such research.

Toxicology has several functions: to identify the origin and nature of toxic insults to human health; to identify their mechanism of action; and to guide regulatory and public-health bodies in measures to remove or reduce toxicological threats. Although mechanistic studies are essential, it is just as important to know the nature of the compounds or environmental insults responsible for adverse effects, and the dosimetry of such exposures. This enables effective risk assessments and development of strategies to reduce risks to human health. Lotti and Nicotera fail to cite any example where mechanistic knowledge alone has allowed reliable risk predictions for humans.

Contrary to the authors' claim that basic research has become irrelevant to many toxicologists (partly justified in the past), recent national and international meetings demonstrate that toxicology has embraced and contributed to the advances of basic molecular and cell biology, as well as incorporating pharmacology, pathology, chemistry and epidemiology. Toxicology is a multidisciplinary subject, and there will always be situations which demand urgent action, irrespective of our understanding of basic biological mechanisms.

If the strategy proposed by Lotti and Nicotera had been applied to liver cancer in Southeast Asia and Africa, it would certainly have produced some interesting mechanistic data on hepatocarcinogenesis. But it was the integration of molecular dosimetry and epidemiological studies that identified aflatoxin B1 as a major risk factor for this cancer. Subsequently, the knowledge of carcinogenic mechanisms provided the rationale for measures to reduce the risk, such as improved food storage, and the design of chemoprevention trials. Similar remarks could be made about organophosphorus pesticides and neurotoxicity; polycyclic aromatic hydrocarbons, aromatic amines and carcinogenicity; accidental or deliberate chemical poisoning; the development of antidotes; and the relevance of endocrine disruptors to human health. In all these it is essential to carry out 'chemical-driven' as well as 'mechanism-driven' toxicology, using cutting-edge methodologies, to generate meaningful risk assessments.

The rapidly expanding knowledge about genetic polymorphisms of enzymes involved in activation (such as cytochrome P450) and detoxification (for example glutathione S-transferase and N-acetyl transferase) of toxins shows that some population groups may be particularly susceptible to some chemicals. The mechanistic study of polymorphisms is valuable, but without considering the nature of the compounds to which individuals are exposed, such knowledge is of little practical value. The public is more concerned to know what compounds are toxic and to see them removed from the environment than to know in detail which biochemical mechanisms are occurring in their cells, important though that may be.

Society rightly demands assurances on the safety of the environment, and this can be achieved only by the application of toxicological principles together with mechanistic understanding. Nevertheless, we should strive to ensure that political agendas do not drive toxicology, but rather that strong science should drive regulatory decisions. It is the interaction between fundamental and applied science that is important in toxicology, not one or the other component.