Toxicology of metals—science confused by poor use of terminology

Toxicology of metals—science confused by poor use of terminology – Editorial

John H. Duffus

Development of the science of metal toxicology has been seriously hindered by the use of incorrect terminology. For example, I have argued previously that the continuing use of the term “heavy metals” as a synonym for “potentially toxic metals” has no scientific basis, has caused much confusion, (1) and should be abandoned. Even use of the term “metals” is misleading because it refers to a particular chemical species of the elements under consideration (see below). With the advantage of hindsight, it is remarkable that for so many years toxicologists have largely ignored chemical speciation of metals. Many toxicological studies have been based on the unwritten and unquestioned assumption that all that mattered was the elemental concentration. There is now an understanding that metallic elements can exist in many different chemical species, and that toxicity must be related to the precise nature of the toxic chemical species. This development has been aided by the formulation of authoritative definitions of the terms “chemical species” and “speciation” by the International Union of Pure and Applied Chemistry (IUPAC). (2) The approved IUPAC definitions are as follows:

Chemical species: Specific form of an element defined

as to isotopic composition, electronic or oxidation

state, and/or complex or molecular structure.

Speciation of an element: Distribution of an element

amongst defined chemical species in a system.

Although there is now a general acceptance of the importance of chemical speciation as defined above, much of the existing toxicological literature is not written in such a way as to make this clear, and is consequently misleading. To illustrate this and related terminology problems, we have only to look at the chapter “Toxic Effects of Metals,” by Goyer and Clarkson, in Casarett and Doull’s Toxicology. (3) Here we find various examples of the imprecise use of terminology which occurs throughout the literature concerned with toxic effects related to the metallic elements and their derivatives. Firstly, the title of the chapter is misleading. A metal is chemically defined as “a substance that con ducts electricity, has a metallic luster, is malleable and ductile, forms cations, and has basic oxides.” (4) In the pure elemental forms described by this definition, most metallic elements are nontoxic because they are not bioavailable. Thus, the chapter might better be titled “Toxic Effects of Metallic Elements, Their Ions and Their Compounds.”

The first sentence of the chapter cited shows how misleading it can be to ignore chemical speciation. “Metals differ from other toxic substances in that they are neither created nor destroyed by humans.” Put correctly, “Metallic elements, like all other elements, toxic or not, are normally neither created nor destroyed by humans (or by any other living organism).” Thus, for example, carbon, hydrogen, and oxygen are neither created nor destroyed by humans. Arguably, if you have a chapter entitled “Toxic Effects of Metals,” you should also have chapters entitled “Toxic Effects of Carbon,” “Toxic Effects of Hydrogen,” etc.

Moving to the second paragraph, we find “Metals are probably the oldest known toxins to humans.” Metals are not toxins. A toxin is defined as a “poisonous substance produced by a biological organism such as a microbe, animal or plant.” (5) Again, such loose use of terminology should have been avoided.

Further on in the second paragraph, we find “Arsenic and other metals are cited by Theophrastus of Erebus …” and arsenic is listed later as the first of the “Major Toxic Metals with Multiple Effects” (second in the list is the gas, arsine–clearly not a metal). Arsenic is not a metal. It is a metalloid or semimetal, which is defined as “an element that has the physical appearance and properties of a metal but behaves chemically like a nonmetal.” (4) Inasmuch as chemical properties determine toxicological properties, metalloids should not be considered as a subset of metals but as a subset of nonmetals. However, this confusion of arsenic with metals is very common in the literature.

In paragraph 5 of the cited chapter, we find the statement “In the case of chromium, DNA-protein complexes may serve as a biomarker of both exposure and carcinogenic potential.” Here we have the common misuse of the name of the element “chromium” in place of the name of the toxic chemical species “chromate.” This misusage implies that the oxygen in chromate does not contribute to its toxicity. At the very least, the presence of oxygen is essential for the bioavailability of chromate, making it chemically similar to phosphate and sulfate and thus permitting its uptake by the general anion channel protein. (6) It is difficult to believe that the powerful oxidizing capacity of chromate does not contribute to its toxic effects, but this tends to be ignored because of the emphasis on the metal atom–almost as though it were a free cation, which it never can be under normal biological conditions. Oddly, I have even found one instance in which chromate is referred to as a metal. (7)

In the first paragraph of the second section of the cited chapter, “Dose-Effect Relationships,” we find it stated that “The most precise definition of dose is the amount of metal within cells or organs manifesting a toxic effect.” This is equivalent to saying, for carbon compounds, that “The most precise definition of dose is the amount of carbon within cells or organs manifesting a toxic effect.” Rephrasing the first statement in this manner points up its absurdity Correctly, the general truth is that “The most precise definition of dose is the amount of toxic chemical species of any element within cells or organs which manifests a toxic effect.” It is unfortunate that, although the authors are well aware that different chemical species of metallic elements have different properties and hence different toxicities, they persist in making general statements about metallic elements which they would never make about carbon, hydrogen, or oxygen.

In the next paragraph of the chapter, we again find reference to arsenic as a metal in the phrase “… some metals such as arsenic and lithium…. “

Proceeding to paragraph 6 in Section 2, we find “Hexavalent chromium is highly toxic whereas the trivalent form functions as an essential trace element.” Hexavalent chromium does not exist in a free form under normal circumstances in living organisms, and thus is not in itself toxic. There is an implication in this usage that hexavalent chromium can exist as a free cation in normal exposures of living organisms, but, as already stated, the chromate cation is the form found, and the presence of oxygen is key to its toxicity. Further, trivalent chromium does not exist as a simple cation, but as a hydrated ion [[Cr[([H.sub.2O).sub.6]].sup.3+], from which are derived complicated, extensively hydrolyzed species (chromites) at pHs above 7, such as are found in the small intestine and in the blood. It is not surprising that such complex species are not readily bioavailable, and are therefore not toxic.

Further on in paragraph 6, we find “Lead attaches to thiol groups on enzymes involved in heme synthesis, trivalent arsenic to the thiol groups of alpha lipoic acid, and uranium to phosphate groups on glucose transporters, all metal ligand interactions underlying toxic action.” Ignoring the fact that arsenic is not a metal, to be correct this should say “Lead (II) ions attach to thiol groups”: arsenite, [(As[O.sub.2]).sup.-] or [(As[O.sub.3]).sup.3-], to the thiol groups of alpha lipoic acid; and uranyl ion [(U[O.sub.2]).sup.2+] to phosphate groups. It should be noted that the uranyl ion may exist in polymerized forms.

In Section 3, “Host Factors Influencing the Toxicity of Metals,” paragraph 5, we find the statement “Metals that provoke immune reactions include mercury, gold, platinum, beryllium, chromium, and nickel.” This is the last example I will quote, but it is typical of what can be found throughout the toxicological literature relating to metals. Herein, I will summarize the bases for a corrected version of the statement, paying proper attention to speciation.

According to Kazantzis, (8) hypersensitivity reactions are provoked by mercury vapor, thiomersal, and mercurochrome. Metallic mercury apparently does not cause hypersensitivity and is relevant only as a source of mercury vapor; it would appear that most inorganic mercuric and mercurous salts do not provoke immune reactions.

Metallic gold rarely causes hypersensitivity, but organic gold compounds used to treat rheumatoid arthritis do. (8)

Metallic platinum is nonallergenic; only a single alleged case of allergy to platinum has been recorded. (9) The results obtained by Cleare et al. (10) suggest that platinum allergy is confined to a small group of charged compounds that contain reactive ligand systems. Most effective are chloride ligands, and the allergic response generally increases with increasing number of chloro groups. Dipotassium tetranitroplatinate (II), [K.sub.2][Pt[(N[O.sub.2)].sub.4]], which contains no chloro group, was inactive. Ionic platinum compounds that contain bromide or iodide are also allergenic, but less effective. Neutral complexes are not allergenic. Beryllium ion is reported to act as a hapten, but it should be noted that the beryllium ion is hydrolyzed in the presence of water to give hydroxylated derivatives of varying and ill-defined complexity. (11) Complexes in which the beryllium ion is unavailable are immunologically inactive. (12)

Chromium metal is considered to be nonsensitizing. (13) Chromium as chromate enters cells, where it is reduced to Cr(III) ions. Antibodies to Cr(III) ions, but not to Cr(VI) (presumably as chromate), have been identified in sensitized animals. (14)

Nickel metal is involved in immune reactions only insofar as it may be dissolved gradually to yield the sensitizing species–nickel(II) ions–likely associated with chloride. (15)

Thus, the sentence “Metals that provoke immune reactions include mercury, gold, platinum, beryllium, chromium, and nickel” should read as follows: “Immune reactions may be caused by the following metal species: from mercury–mercury vapor, thiomersal, and mercurochrome; from gold–organic gold compounds used to treat rheumatoid arthritis and, rarely, the metal; from platinum–only certain halogenated coordination complexes, especially chlorinated salts; from beryllium–hydroxylated beryllium ions; from chromium–chromate, although the hapten is Cr(III) ion produced by biological reduction; and from nickel–Ni(II) ions, probably associated with chloride ions.”

It is hoped that this commentary will persuade scientists writing about metal toxicology to take more care in their use of terminology. Careless and imprecise use of terminology is not conducive to clarity of thought; it certainly cannot promote clear thought in the minds of students, and should not be found in our textbooks. Nor is it helpful in risk assessment and risk management of metals. In risk assessment, there is a continuing struggle to prevent metallic elements from being classified as persistent toxicants simply because no distinction is made between the element and its toxic chemical species. This distinction is always made for the element carbon, and there is no scientific justification for treating metallic elements differently.

All of us who have been active in the field of metal toxicology have probably been guilty of poor use of terminology in the past. Now it is time to apply stricter standards to our writing and teaching, to ensure that we provide a solid foundation for future progress.


(1.) Duffus JH. “Heavy Metals”–a meaningless term? Pure Appl Chem 2002; 74(5):793-807. Available at×0793.html

(2.) Templeton D, Aries F, Cornelis R, et al. Guidelines for terms related to chemical speciation. Pure Appl Chem 2000; 71(8):1453. Available at

(3.) Goyer RA, Clarkson TW. Toxic effects of metals. In: Klaasen CD (Ed). Casarett & Doull’s Toxicology, 6th ed. New York: McGraw-Hill, 2001; pp 811-67.

(4.) Atkins P, Jones L. Chemistry–Molecules, Matter and Change, 3rd ed. New York: W. H. Freeman, 1997.

(5.) Duffus JH. Glossary for chemists of terms used in toxicology. Pure Appl Chem 1993; 65(9):2003-122.

(6.) Costa M. Toxicity and carcinogenicity of Cr(VI) in animal models and humans. Crit Rev Toxicol 1997; 27:43142.

(7.) Costa M. Chromium and Nickel. In: Zalups RK, Koropatnick J (Eds). Molecular Biology and Toxicology of Metals. London and New York: Taylor & Francis, 2000; pp 115.

(8.) Kazantzis G. Hypersensitivity: clinical aspects. In: Dayan AD (Ed). Immunotoxicity of Metals and Immunotoxicology. New York: Plenum, 1990; pp 67-74.

(9.) Sheard C. Contact dermatitis from platinum and related metals. Arch Dermatol Syphilol 1955; 71:357-60.

(10.) Cleare MJ, Hughs EG, Jacoby B, et al. Immediate (Type I) allergic responses to platinum compounds. Clin Allergy 1976; 6:183-95.

(11.) Greenwood NN, Earnshaw A. Chemistry of the Elements, 2nd ed. Oxford, U.K.: Butterworth-Heinemann, 1997.

(12.) Krivanek M, Reeves AL. The effect of chemical forms of beryllium on the production of the immunologic response. Am Ind Hyg Assoc J 1972; 33:45-52.

(13.) Cronin E. Contact Dermatitis. Edinburgh: Churchill Livingston, 1980.

(14.) Novey HS, Habib M, Wells LD. Asthma and IgE antibodies induced by chromium and nickel salts. J Allergy Clin Immunol 1983; 72:407-12.

(15.) Avnstorp C, Menne T, Maibach H. Contact allergy to chromium and nickel. In: Dayan AD (Ed). Immunotoxicity of Metals and Immunotoxicology. New York: Plenum, 1990; pp 83-91.

John H. Duffus

The Edinburgh Centre for Toxicology

43 Mansionhouse Road

Edinburgh EH9 2JD

Scotland, U.K.

Response to Comments by Professor Duffus regarding Chapter 23, Toxic Effects of Metals, in Cassarett and Doull’s Toxicology, 6th edition, C. D. Klaassen, Editor. New York: McGraw Hill, 2001.

The editorial by Professor Duffus calls attention to the incorrect use of terminology in the field of metal toxicology. He focuses his criticisms on the chapter “Toxic Effects of Metals,” by Goyer and Clarkson, in Cassarett and Doull’s Toxicology. The concerns expressed in his critique on “the use of incorrect terminology” generally reflect differences between the approach that might be used by a physical scientist and the approach that might be used by medical scientists, who are primarily concerned with the evaluation of health risks. Many of these same concerns are addressed by the U.S. Environmental Protection Agency in an an action plan for metals risk assessment. (1,2) It may be more correct, as a chemist, to label a metal such as lead by its chemical species (e.g., lead acetate or oxide or chloride); however, the health risks from a metal such as lead are related to the presence of the metal per se in some body tissue, and that is the way the metal is identified in the body. Whether the exposure was to lead chloride or oxide is of concern only as a factor of bioavailability and route of exposure. Toxicity is related to a metal after it has been absorbed into a body tissue and bound to a cellular ligand. The measurement of the internal exposure of cellular targets to metals is generally performed using methodologies that only identify the metal (e.g., atomic absorption, or mass spectrography, or neutron activation), and not its chemical species. Therefore, it is the relationships between the measurement of the metal at its target site and the observed biological effect that constitute the basis for dose-response relationships and health risks, as discussed at length in the current (6th) edition of Toxicology. Chapter 23 does differentiate effects from exposure to different chemical species when differences in effects are known. For example, differences in effects between inorganic forms of arsenic and arsine gas, and between inorganic and organic forms of mercury and lead exposure, are considered in the current text. These comments are not meant to minimize the concerns expressed by Professor Duffus, but to provide the perspective of a health scientist. It is questionable whether expanding the title of the chapter would add clarity to its contents.


(1.) U.S. Environmental Protection Agency (EPA). Draft Action Plan for the Development of a Framework for Metals Assessment and Guidance for Characterizing and Ranking Metals. Federal Register (67 FR 39982, June 11, 2002). EPA/630/P-02/003A. Available at:

(2.) U.S. Environmental Protection Agency (EPA). Summary Report of the Meeting on Development of a Metals Assessment Framework. National Center for Environmental Assessment (NCEA-F-1292, Summary Report, June 10, 2002). Available at:

Robert A. Goyer, M.D.

Professor Emeritus

University of Western Ontario

London, Ontario, Canada

COPYRIGHT 2003 Heldref Publications

COPYRIGHT 2004 Gale Group