EPA-600/1-78-016
February 1978
TOXICOLOGY OF METALS - Volume III
by
Subcommittee on the Toxicology of Metals
Permanent Commission and International Association of
Occupational Health
Prof. Lars Friberg, Chairman
in cooperation with
The Swedish Environmental Protection Board, and
The Karolinska Institute
Contract No. 68-02-1287
Project Officer
Robert J. M. Horton
Criteria and Special Studies Office
Health Effects Research Laboratory
Research Triangle Park, N.C. 27711
U.S. ENVIRONMENTAL PROTECTION AGENCY
OFFICE OF RESEARCH AND DEVELOPMENT
HEALTH EFFECTS RESEARCH LABORATORY
RESEARCH TRIANGLE PARK, N.C. 27711
-------
DISCLAIMER
This report has been reviewed by the Health Effects
Research Laboratory, U.S. Environmental Protection Agency,
and approved for publication. Approval does not signify
that the contents necessarily reflect the views and policies
of the U.S. Environmental Protection Agency, nor does
mention of trade names or commercial products constitute
endorsement or recommendation for use.
-------
TABLE OF CONTENTS
Page
Introduction 1
List of Contributors 3
GENERAL CHEMISTRY OF METALS
Velimir Vouk 5
SAMPLING AND ANALYTICAL METHODS
Magnus Piscator and Velimir Vouk 29
SOURCES, TRANSPORT, AND TRANSFORMATION OF
METALS IN THE ENVIRONMENT
Karin Beijer and Arne Jernelov 51
EFFECTS - GENERAL PRINCIPLES
UNDERLYING THE TOXIC ACTION OF METALS
Thomas W. Clarkson 83
FACTORS INFLUENCING EFFECTS AND
DOSE-RESPONSE RELATIONSHIPS OF METALS
Gunnar Nordberg, Jiri Parizek and 117
Magnus Piscator
EPIDEMIOLOGICAL ASPECTS OF ASSESSMENT OF
DOSE-RESPONSE AND DOSE-EFFECT RELATIONSHIPS
Tord Kjellstrom 141
GENERAL ASPECTS OF THE PREVENTION OF
METAL POISONING
Sven Hernberg 171
GENERAL ASPECTS AND SPECIFIC DATA ON
ECOLOGICAL EFFECTS OF METALS
Karin Beijer and Arne Jernelov 201
STANDARDS AND CRITERIA
Lars Friberg and Velimir Vouk 223
CLINICAL EFFECTS, DIAGNOSIS AND TREATMENT
OF METAL POISONING - GENERAL CONSIDERATIONS
George Kazantzis 237
MUTAGENIC AND CARCINOGENIC EFFECTS OF METALS
George Kazantzis and Lorna Lilly 265
111
-------
INTRODUCTION
Introduction to Volume I
At the time of the XVIth International Congress on
Occupational Health in Tokyo in 1969 the Permanent Commission
and International Association of Occupational Health decided
to form a subcommittee on the toxicology of metals. The need
for such a group had been recognized during an international
meeting on mercury toxicology in 1968. The subcommittee was
formed with Prof. Lars Friberg of Stockholm as chairman. The
program developed by this group consisted of exploration and
documentation of the general principles of raetal toxicology
by means of symposia and workshops of experts in the field. A
broad approach was selected including general environmental
exposures as well as those related to occupation. Emphasis was
placed on greater depth and precision of understanding in terms
of metabolic processes and dose-effect relationships. The
Permanent Commission, the Swedish Environmental Protection
Board, and the Karolinska Institute have supported the committee
in carrying out this program. Two workshops were conducted, one
in conjunction with the meeting of the Permanent Commission in
1971, and another in conjunction with the XVIIth International
Congress on Occupational Health in 1972.
Recently the subcommittee has decided to broaden its work
by preparing a handbook on metal toxicology which will contain
the general material being developed in its workshops and also
specific information on a large number of metals for use by
toxicologists and occupational and environmental health workers.
-------
The Environmental Protection Agency is cooperating with the other
sponsors in supporting this phase of the subcommittee's work. As
information is developed on the subject it will be collected in
three annual progress reports. The first of these is contained
in this volume. It consists of the conclusions of the sub-
committee's third workshop which was held in Tokyo in 1974.
Since the three workshops which have been held constitute a
cumulative series, with frequent reference back to previous
definitions and statements, it was thought best on this occasion
to reproduce all of them together as a unit.
Introduction to Volume III
This third volume on Toxicology of Metals contains material
prepared for the Handbook of the Toxicology of Metals - Environ-
mental and Occupational Aspects which is being prepared by the
Scientific Committee on the Toxicology of Metals of the Permanent
Commission and International Association of Occupational Health.
Volume II contained chapters on the toxicology of specific metals.
This third volume contains chapters on the general aspects of
metal toxicity. Although these are finished chapters, it has
occurred to the authors and editors that publication at this time
offers the opportunity to solicit comments and suggestions from
the scientific community which might be helpful in improving
these texts before final publication in the handbook. Readers
who know of additional pertinent material which should be
included, or have other suggestions for improvements are urged
to communicate directly as soon as convenient with the responsible
senior authors whose addresses are given on page 3.
-------
List of Contributors to this Volume
Karin Beijer, M.S.
The Swedish Water and Air Pollution Research Laboratory
Halsingegatan 43
S-100 31 Stockholm, Sweden
Thomas W. Clarkson, Ph.D.
Department of Radiation Biology and Biophysics
School of Medicine and Dentistry
University of Rochester
Rochester, N.Y. 14627, U.S.A.
Lars Friberg, M.D.
Department of Environmental Hygiene
of The Karolinska Institute
and of The National Swedish Environment Protection Board
S-104 01 Stockholm 60, Sweden
Sven Hernberg, M.D.
Institute of Occupational Health
Haartmaninkatu 1
SF-00290 Helsinki 29, Finland
Arne Jernelov, Ph.D.
The Swedish Water and Air Pollution Research Laboratory
Halsingegatan 43
S-100 31 Stockholm, Sweden
George Kazantzis, M.D.
Department of Community Medicine
The Middlesex Hospital Medical School
Central Middlesex Hospital
London NW10 7NS, England
Tord Kjellstrom, M.D.
Department of Community Medicine
School of Medicine
University of Auckland
Private Bag
Auckland, New Zealand
Lorna Lilly, M.D.
Department of Biology as Applied to Medicine
Middlesex Hospital Medical School
Cleveland Street
London W 1, England
Gunnar Nordberg, M.D.
Institute of Community Health and Environmental Medicine
Odense University
Winslovsvej 19
DK-5000 Odense, Denmark
-------
Jiri Parizek, M.D.
Institute of Physiology
Czechoslovak Academy of Sciences
Budejovicka 1083
CS-142 20 Prague 4
CSSR
Magnus Piscator, M.D.
Department of Environmental Hygiene
The Karolinska Institute
S-104 01 Stockholm 60, Sweden
Velimir Vouk, Ph.D.
Control of Environmental Pollution and Hazards
Division of Environmental Health
World Health Organization
1211 Geneva 27, Switzerland
Technical editor:
Pamela Boston, B.A.
Department of Environmental Hygiene
The National Swedish Environment Protection Board
S-104 01 Stockholm 60, Sweden
-------
GENERAL CHEMISTRY OF METALS
Velimir Vouk
Out of 104 identified elements about 80 are usually considered
as metals, and the chemistry of metals represents a major
part of inorganic chemistry. The chemical basis of metal
toxicology is still inadequately understood although signi-
ficant advances have been made with the emergence of bio-
inorganic chemistry.
For details on the chemistry of metals, the reader is referred
to the recent texts on inorganic chemistry by Cotton and
Wilkinson (1972, 1976). Those less familiar with modern
chemistry are advised to consult the excellent introductory
text by Pauling (1970). A good source of information on the
properties of ions in solution, a field of great importance
for the understanding of metal toxicology, is the introductory
monograph by Parr (1973) . The principles of bio-inorganic
chemistry are discussed by Hughes (1972) and Phipps (1976) and
some theoretical and chemical aspects of the toxicology of
metals by Brakhanova (1975).
Most of the information contained in this chapter is based
on the sources referred to above and, in view of its introductory
character, original literature is not cited.
1. Definition of metals
Metals are usually defined on the basis of their physical
properties in the solid state, which are of great technological
significance. These are 1) high reflectivity which is responsible
for the characteristic metallic luster, 2) high electrical
conductance which decreases with increasing temperature, 3)
high thermal conductance, and 4) mechanical properties such
as strength and ductility. Metals are also characterized by
their crystal structure, by their specific type of chemical
bond in which the electrons are delocalized and mobile, and
-------
by their magnetic properties. Of particular technological
usefulness are alloys, metallic materials consisting of two
or more elements. These properties of metals have a limited
interest for the understanding of systemic toxic effects,
although some of them may be of importance as regards the
local effects of metal aerosols.
From the toxicological viewpoint, a more useful definition
of metals is that based on the properties of their ions in aqueous
solutions: a metal is an element which under biologically
significant conditions may react by losing one or more
electrons to form a cation. Much of the discussion that
follows will be concerned with the behavior of metal ions in
solution.
The distinction between metals and non-metals, whether based
on their physical or chemical properties, is not a sharp
one. Some elements such as selenium or germanium have certain
properties which are typical of metals; other properties
make them similar to non-metals. This also applies to such
elements as boron and silicon which have to be classified
separately as metalloids or metal-like elements. In general,
there is in some groups of the periodic system a gradual
transition of properties from non-metals to metals as we
descend from the lighter to the heavier elements (e.g. C,
Si, Ge, Sn, Pb in the IVb group of the periodic system).
2. Classification of metals
The properties of chemical elements depend on the electronic
configuration of the atom and vary with the atomic number
in a systematic way. The periodicity of properties may be
shown by arranging the elements in a periodic table (Table 1).
The horizontal rows, called periods, consist of a very short
period (containing hydrogen and helium), two short periods of
8 elements each, two periods of 18 elements each, a very long
period of 32 elements, and an incomplete period. The vertical
columns of the periodic table are the groups of elements. There
are 9 groups of elements, groups I to VII being further sub-
divided into subgroups a and b.
-------
The electronic configuration of elements is usually described
in terms of shells, subshells and "orbitals" which indicate
the spatial distribution of electrons. The so-called s-orbitals
are spherically symmetrical around the nucleus of the atom.
Other orbitals (p, d, etc.) are not spherically symmetrical
(Figure 1). The first shell consists of one orbital (Is) which
can accommodate at most two electrons, as indicated by the symbol
2
Is . This is the electronic structure of the element helium.
The second shell consists of two subshells (2s and 2p). The 2p
subshell is further subdivided into 3 orbitals (2p , 2p and
2p ). Since each orbital can accomodate only 2 electrons
z
(Pauli's exclusion principle), the second shell contains a
maximum of 8 electrons. This is indicated by the symbol
fj /-
2s 2p . The element neon has both shells completely filled, and
9 9 ft
its electron structure is (Is ) (2s 2p ). The next shell con-
sists of 3 subshells: 3s (one orbital), 3p (three orbitals)
and 3d (five orbitals). When completed this shell will
accommodate a total of 18 electrons arranged as follows:
3s 3p 3d . The fourth shell contains these subshells, and in
addition a 6f subshell which consists of seven 4f orbitals.
The filling of the successive shells, subshells and orbitals
proceeds in the sequence of increasing energies which is, for
the lighter elements, ls<2s<2p<3s<3p<4s<3d<4p<4d<4f.
According to their electronic configuration, metals are
classified into s-, p-, d- and f-blocks (see Table 1). The
s-block contains the alkali metals (Li, Na, K, Rb, Cs, Fr),
beryllium (Be), magnesium (Mg) and the alkaline earths (Sr,
Ba, Ra). Their electronic configuration is characterized by
a noble gas core and one s (alkali metals) or two s electrons
(Be, Mg and alkaline earths).
The d-block contains at least 32 elements which can be
further subdivided into 3d and 4d and 5d metals. With the
exception of Zn, Cd and Hg, the d-block metals belong to the
so-called transition elements which have an incompletely
filled d-orbital. The first series of transition elements
(3d - sub-block) contains several biologically important
metals such as vanadium, chromium, manganese, iron, cobalt
and copper.
-------
The f-block metals also belong to the transition elements.
Their electronic configuration is characterized by partially
filled 4f and 5f orbitals. The 4f elements are called
lanthanides and have a very simple chemistry. In this respect
they are similar to the s-block metals. The 5f metals(actinides)
have a much more complex chemical behavior. They are the
heaviest elements in the periodic table and are radioactive.
The p-block consists of 25 elements out of which only 9 are
usually classified as metals (Al, Ga, In, Tl, Ge, Sn, Pb,
Sb, Bi) although As, Se, Te and Po also exhibit some metallic
properties.
3. Compounds of metallic elements
3.1 Covalent and ionic bonds
There are two extreme types of chemical bonds, covalent and
ionic, and examples of each type of bond can be found among
metal compounds. Since electrons are in continuous motion
around the atomic nucleus, the positive charge of the nucleus,
if averaged over all directions in space, is balanced by the
negative charge of the electrons. However, if we select a
specific direction in space, the positive charge of the nucleus
will not be completely shielded by the electrons. Thus, if
two atoms are brought together, the effective positive charge
of each atom will displace the outer electrons (the valence
electrons) of the other atom, so that the electron density
becomes greater in the region between the two nuclei. In other
words the two nuclei will share a pair of electrons. The
resulting chemical bond is called covalent bond. Although
quantum mechanical concepts such as pairing of electron spins
are needed to describe a covalent bond, the basic binding force
is the electrostatic interaction of the bonding electrons with
the nuclei between which they are located. When the two atoms
of a diatomic molecule are the same (e.g. two hydrogen atoms)
the electron density is distributed symmetrically between the
two nuclei and the covalent bond is homopolar. If the two atoms
are not identical, the distribution of electrons between the
-------
two nuclei will be asymmetrical, and the electron density
will be displaced towards the atomic nucleus which is more
electronegative, i.e. which has a higher capacity to attract
electrons. This is a heteropolar covalent bond. The greater
the difference in electronegativities of the two atoms
forming a bond, the more uneven will be the distribution of
electrons between the two atoms, and an extreme case can be
envisaged in which an electron has been entirely transferred
from one atom to the other, forming a completely electro-
static or ionic bond. Metallic elements have a low electronega-
tivity, and are relatively easily ionized by non-metallic
elements with high electronegativity (e.g. chlorine), thus
forming ionic bonds. The ionic bond is a useful model but it
should be understood that a chemical bond is rarely either
entirely covalent or entirely ionic. The electron distri-
bution in homopolar, heteropolar and ionic bonds is shown
in Figure 2.
Predominantly ionic bonds are found in metal salts such as
NaCl or Ca(NO,)9. Predominantly covalent bonds are formed
between metals and carbon atoms, for example in organometallic
compounds such as dimethylmercury CH^-Hg-CH^, but there are
also other metal compounds with covalent bonds.
3.2 Inorganic compounds
Metallic elements form a variety of inorganic compounds. They
may be classified into the binary compounds and multi-
element compounds.
The most important binary compounds, both from the technological
and toxicological viewpoint, are oxides and sulfides. This is
the chemical form in which most metals appear in nature as
minerals, and metal-containing aerosols produced in the course
of metallurgical processes are usually metal oxides or sulfides.
Closely related to sulfides are binary compounds of metals
with the elements tellurium and selenium (tellurides and
selenides).
-------
Another class of important binary metal compounds consists of
halides, i.e. compounds of metallic elements with fluorine,
chlorine, bromine and iodine. Most metal halides are substances
of predominantly ionic character although there is a
gradation from purely ionic to essentially covalent compounds.
For example, KC1 is completely ionic but TiCl. is a covalent
molecular compound.
Metallic elements can also form binary compounds with hydrogen
(hydrides), carbon (carbides), and nitrogen (nitrides). All
these compounds may be either predominantly ionic (e.g. LiH,
CaC-, Ba..Np) or essentially covalent (e.g. PbH. , SnH. and BeC) .
Some metal oxides, for example, alkali metal oxides (Na~0, K2O),
react with water to give hydroxides (NaOH, KOH) which are
very soluble in water, giving strong alkaline solutions.
Other examples of metal hydroxides are Zn(OH)_ and Cd(OH)2.
Some transition metals form a variety of oxyhalides, ternary
compounds containing a halogen element, oxygen and a metal.
Examples of such compounds are VOC1,, VO-Cl, CrOF., MnCUF and
FeOCl.
Salts are ionic compounds formed by a reaction between metal
hydroxides, such as sodium hydroxide (NaOH) and acids which
may be either binary acids such as HC1, oxoacids such as
H2SO4 (sulfuric acid), HNO^ (nitric acid) and H-CO^, or
organic acids such as acetic acid (CH_,COOH) or oxalic acid
(HOOCCOOH). "Acid" salts are compounds such as sodium
hydrogen carbonate NaHCO., or sodium hydrogen phosphate
NaH(PHO3). "Basic" salts may be exemplified by magnesium
hydroxide chloride Mg(OH)Cl; or BiO(NO^) and BiO(OH)-BiO(NO,)
•s ~*
which are the products of hydrolysis of bismuth nitrite
Bi(NO3)3-5H2O.
Transition metals such as manganese and chromium form, in
their highest oxidation states, oxoanions such as manganate
2— - 2 —
ion (MnO . ), permanganate ion (MnO. ) , chromate ion
10
-------
and dichromate ion (Cr-O. ) . Typical compounds containing such
anions are potassium permanganate (KMnO.) or lead chromate
(PbCrO .) . Selenium and tellurium also form oxoacids, for
example, HgTeO, (telluric acid) and H^SeO. (selenic acid) .
As regards the nomenclature of such polyatomic anions, the
termination -ate indicates a higher oxidation state, e.g.
arsenate AsO. , compared to an oxoanion with the central
4 3_
atom in a lower oxidation state, e.g. arsenite AsO., . The
termination -ide is applied to binary compounds such as
barium arsenide, BaAs-.
3 . 3 Metal complexes
A metal complex is formed by the association of a metal atom or
ion and another chemical species, called ligand, which may
be either an anion or a polar molecule. Many complexes are
rather stable structures and may remain unchanged in a
sequence of chemical or physical reactions. All metals form
complexes, but the formation of complexes is so widespread a
phenomenon for transition metals that it provides a key for
the understanding of their chemistry. Each metal can be
assigned a principal valency which is the oxidation state.
In addition, there is an auxiliary valency which represents
the number of ligands associated with the central metal ion;
this is equivalent to the coordination number of the metal.
Most common are the coordination numbers 4 and 6 , although
others are known as well, ranging from 2 to 10.
Ligands can be unidentate, i.e. they provide only one electron
pair. Examples are simple anions such as chloride or ammonia.
Multidentate ligands contain more than one donor atom and
it may be sterically possible to coordinate one metal atom
at two or more positions in its coordination shell. A well-
known example is ethylenediamine or 1, 2-diaminoethane
NH? -CHp -CH2 -NHp which contains two donor nitrogen atoms,
and acts as a bidentate ligand. In some cases, such bi-
or polydentate ligands form coordination compounds in the form
of a heteroatomic ring, shown below. This process of ring forma-
tion is called chelation. Ligands such as ethylenediamine are
often referred to as chelating agents.
11
-------
H 2 H£
/N N\
\ / CH2
H2 H2
bis (ethylenediamine) platinum (II)
Chelates are usually very stable complexes and have therefore
a great importance in analytical chemistry of metals and in
the treatment of metal poisoning. Other multidentate ligands
of great interest are nitrilotriacetic acid (NTA) which is
tetradentate and ethylenediaminotetraacetic acid (EDTA) which
is sexadentate.
A great variety of ligands exists. For example, they may have
two basic groups as 1,10-phenanthroline, two acidic groups as
carbonate ion, and one acidic and one basic group as is the
case with ot-amino acids, such as glycine. Corrin and porphin
molecules are other examples of multidentate ligands which
are of great biological importance, the corrin structure being
the basic chelating arrangement in a cobalt complex, vitamin
B12'
The most useful chelating agents in the treatment of lead
poisoning are EDTA and other aminopolycarboxylic acids.
Arsenic is a "soft" Lewis acid (see section 5.3) which pref-
erably binds ligands with sulfur donor atoms . This property
has been used in developing a chelating ligand 2,3 dimercapto-
propanol as an antidote to Lewisite, Cl-CH=CH-AsCl_ . BAL acts
by forming a five-membered chelate ring with arsenic. BAL is
effective in the treatment of mercury poisoning because mer-
cury, like arsenic, shows a preference for ligands with sulfur
donor atoms. D-isomer of penicillamine is the most effective
chelating agent for binding copper. D-penicillamine can
coordinate with a suitable metal through carbon, sulfur and
nitrogen donors.
CH.SH CH3
CH-SH SH — C — CH
3
CH2'SH H2N—C— CH2K
BAL H
D-penicillamine
12
-------
3-
Further details on the chemistry and application of chel-
ating agents can be found in a recent monograph by Bell
(1977) .
Another class of d-transition metal complexes of considerable toxi-
cological interest are the complexes of ^-acceptor (T acid)
ligands. These are the complexes of a metal with various
neutral molecules such as carbon monoxide, substituted phos-
phines, stibines, nitric oxide and pyridine, and include,
for example, binary molecular compounds such as Cr(COfc but
also such complex ions as [Fe(CN)^COJ and [MO (CO) ^IJ . in
many of these complexes the metal ion is in a low positive,
zero or even negative formal oxidation state. The most im-
portant of these types of compounds are metal carbonyls which
can be mononuclear, i.e. containing a single metal atom like
Fe(CO)_, or polynuclear like Fe (CO) 0, or heteronuclear like
r* "1 A f "i **
MnRe(CO),... Cyanide complexes such as Fe(CN),] er[Mo(CN)R|
belong to the same group of metal compounds and are restricted
to the d-block transition metals.
3.4 Organometallic compounds
Organometallic compounds contain a carbon atom of an organic
group bound to metal atoms. There are three main types of
Organometallic compounds:
1. Ionic compounds of electropositive metals. Examples of such
- + 2+
compounds are (C-Hj-i-C Na or (Cc^c ) ^ Ca
2.6-bonded compounds in which the organic residue is bound
to a metal atom by a normal 2-electron covalent bond.
Examples of such compounds include (CH.,) _ SnCl and CH_.HgCl.
3. Non-classically bonded compounds. One class of these com-
pounds includes the alkyls of Li and Be that have bridging
alkyl-groups. A second much larger class includes compounds
of transition metals with alkenes, alkynes, benzene and
other ring systems.
Of considerable toxicological interest are alkyl- or aryl-
mercury compounds RHgX or R_Hg. The RHgX type belongs to
either ionic or 6-bonded compounds depending on X. When X
13
-------
can form a covalent bond with Hg, as for example Cl, Br, or
OH, the compound is a covalent non-polar substance. When X
2- -
is SO or NO- , the substance is ionic in character. Di-
alkyls and diaryls are non-polar, volatile, toxic, colorless
liquids. Biological methylation of mercury or its compounds
gives highly toxic (CH^)0Hg or CH.,Hg . Methylcobalt oximes
2 +
containing a Co-CH bond can transfer CH_ to Hg
Practically all the organometallic compounds of germanium,
tin and lead are formed in the oxidation state +4. Of toxico-
logical interest are only lead and tin compounds, particularly
(CH7).Pb and (C H ) .Pb and trialkyl and dialkyltins which have
j ~i £ O ^
a variety of industrial and other uses.
Examples of the third category of organometallic compounds are
5 5
n -cyclopentadienyls such as (C H5) Fe. { r\ indicates that
5 carbon atoms are bound to the metal, and is pronounced
"pentahapto".) This type of compound is shown below.
ferrocene ^ -cyclopentadienyl-Mn-tricarbonyl
Of potential toxicological importance are the cyclopentadienyl
manganese tricarbonyl (n - C2H5^ Mn (c°) •> and its methylated
analog, MMT, which have been considered as a replacement for
lead tetraethyl as an antiknock agent.
4. Solubility
The solubility of metal compounds in water and in lipids is
of great toxicological importance because it is one of the
major factors influencing the biological availability and
absorption of metals. The solubility of metal compounds in
water depends on the presence of other chemical species,
14
-------
particularly H -ions (pH), and so it may be quite different
depending on whether the solvent is "pure" water or a bio-
logical fluid. Biological fluids are slightly alkaline in
mammals (pH 7.4) with the exception of the fluids in the
gastrointestinal tract which are acid (pH 2-6) in the stomach
and almost neutral (pH 6.8) in the intestines. In addition,
biological fluids may also contain a variety of organic ligands
which exert a further influence on the solubility. Experimental
data on the solubility of metal compounds in biological fluids
are very limited.
In spite of the mentioned differences, these are some simple
rules governing the solubility of metal compounds in water
which may be a useful indicator for the solubility of these
compounds in biological fluids. A simple rule, used in chemistry,
divides various substances into "soluble" and "insoluble".
"Soluble" substances are those which have a solubility in
water >1 g/100 ml; "insoluble" are those which have a solubility
< 0.1 g/100 ml. This distinction may not be meaningful if a
substance is highly toxic. Within each group of the periodic
system the solubility of metal compounds generally decreases
with increasing atomic numbers.
Mainly soluble metal compounds are nitrates, acetates and
all chlorides, bromides and iodides except those of silver,
mercury (I) and lead. All sulfates, except those of barium,
strontium and lead are also soluble. All salts of sodium,
potassium and ammonium are soluble except sodium antimonate,
potassium hexachloroplatinate and potassium cobaltonitrite.
Mainly insoluble are all hydroxides (except those of alkali
metals, ammonium and barium), all normal carbonates and phos-
phates (except those of alkali metals and ammonium), and sul-
fides (except those of alkali metals, ammonium and the alkaline
earth metals).
In addition to the factors mentioned above, such as pH, the
presence of other ions, etc., solubility may depend on the
15
-------
oxidation state of a metal, and on the rate of oxidation/
reduction conversions.
The solubility of sparingly soluble substances will also
depend on their particle size. Finely divided material is
usually more soluble.
5. Properties of metal ions
5.1 Formation of metal ions
Metals ions are formed by the removal of one or more outer
electrons from the neutral atom. The energy required for the
ion formation depends on the environment in which this
process takes place. The formation of ions in the gas phase
requires a considerable amount of energy. The same process
requires much less energy if it takes place in water because
a part of the ionization energy is provided by the energy of
hydration, i.e. the energy that is gained when a positively charged
metal ion binds dipolar water molecules. The number of water mole-
cules that are bound directly to the metal ion (first hydration
sphere) depends on the size and the charge of the metal ion,
+ 2+
and varies from 4 for Li to about 10 for Ra . Because there
is further polarization of water molecules contained in the
first hydration sphere, additional water molecules will be
attracted to form a second hydration sphere. This association
can, of course, continue but its extent decreases rapidly
with distance from the ion. The size of the hydration sphere
will obviously depend on the polarizing power of the ion
which in turn depends on the charge/radius ratio. The hydrated
ion is a dynamic system in which water molecules in the
hydration sphere rapidly exchange with those in the bulk
phase of the solution.
5.2 Redox potentials
The removal of electrons from solid metal is an oxidation
process. The reverse process is called reduction. Oxidation
and reduction processes are always coupled, i.e. when one
substance is oxidized (reducing agent), another must be
16
-------
reduced (oxidizing agent). The oxidizing or reducing power of
an oxidation-reduction system is measured in terms of standard
electrode potentials. If the standard electrode potential of
a metal is large and negative, the metal is a powerful
reducing agent because it loses electrons easily. Standard
electrode potentials, or redox potentials, depend on the
concentration of metal ions, on the temperature, and on the
presence of other ligands that can displace water from a
hydrated ion. Oxidation-reduction reactions are of fundamental
importance in biochemistry, and transition metals which can
easily change their oxidation state play a very important
role.
5.3 Metal ions as Lewis acids
A useful definition of acids and bases is that of G.N.
Lewis who defined an acid as an electron pair acceptor and a
base as an electron pair donor. Thus all the positively
charged metal ions can be classified as Lewis acids or
electron acceptors. In the same way water molecules which
are formally electron donors can be described as Lewis
bases. In this terminology the formation of a hydrated metal
ion is in fact the formation of a complex. This can be
indicated by the following equations:
Metal ion + water > hydrated metal ion
(Lewis acid + Lewis base * complex)
Hydrated ion + ligand > complex and water
The last equation indicates that water in the hydration shell
can be replaced by another ligand (another Lewis base) to
give a different complex. The chemistry of metal complexes
is very important not only in analytical chemistry but also
for the understanding of metal biochemistry and toxicology.
Lewis acids and bases can be classified as "hard" and "soft".
A "hard" metal ion or Lewis acid is one which retains its
valence electrons very strongly and is not readily polarized.
Ions of small size and high charge are "hard". A "soft" ion
is relatively large, does not retain its valence electrons
17
-------
firmly and is easily polarized. Examples of hard ions are
+ 2+
L and Mg as well as bases with donor atoms such as
oxygen, nitrogen (e.g. amines, ammonia, water) or fluorine.
"Soft" metals include copper (I), silver (I), gold (I),
thallium (I), palladium (II) and platinum (II). Easily
polarized ligands containing phosphorus, arsenic or sulfur
donor atoms are "soft" bases.
As a rule the formation of stable complexes results from
interaction of hard acids with hard bases, or soft acids
with soft bases. The bonding in hard-hard complexes is
largely electrostatic whereas in soft-soft complexes it is
more covalent.
5.4 Hydrolysis
Hydrolysis is a reaction that may occur between the metal ion and
a water molecule in the coordination (solvation) sphere, in which
a proton (hydrogen ion) is released and the solution becomes acidic.
This can be described by the equation
[M (OH2)x]n+i=^ [M(OH2)x-lp~1) + + H* (a)
Hydrolysis may proceed in several stages until the last coordinated
water molecule is removed. The process of hydrolysis may be inter-
rupted if at one stage an insoluble compound is produced. Hydrolysis
occurs most readily with metal ions which strongly polarize the co-
ordinated water molecules. The polarizing power increases with the
increasing charge and decreasing size of the ions. For example, in
2 +
the case of alkaline earths, the hydrolysis is confined to Be
which has a small radius compared to other members of the group.
5.5 Main group metals
The main group metals are those belonging to the periodic system
groups la (alkali metals), Ila (beryllium and alkaline earths),
lib (zinc, cadmium and mercury), Illb (aluminum, gallium, indium
and thallium), IVb (germanium, tin and lead), Vb (antimony and
bismuth). Arsenic (group Vb), selenium and tellurium (group VIb)
are metal-like elements that also belong to the main group elements.
18
-------
The salts of the alkali metals are generally soluble and produce
unipositive aquo ions with a probable coordination number 4 for Li,
Na and K and a probable coordination number 6 for Cs. Lithium salts
are frequently less soluble in water than the corresponding salts of
other alkali metals. The hydrated cations of the alkali metals show
little tendency to hydrolyze, or to form complex ions. Potassium
and sodium have important biological functions (e.g. osmo-regulation,
acid-base balance in the organism, transmission of nerve impulses)
and lithium is of toxicological interest. By losing its outer
s-electron, lithium +1 ion becomes spherically symmetrical and
non-polarizable. The similarity in chemistry of Li and Mg^+ can
be explained by their similarity in charge/radius ratios.
The salts of strong acids of alkaline earths are soluble, with the
exception of the sulfates of Ca, Sr and Ba. Salts of weak acids are
considerably less soluble. All the hydroxides with the ex-
ception of Be-hydroxide are strong bases but much less soluble
than the corresponding alkali-metal compounds. Be-hydroxide
is amphoteric. All cations of this group are hard Lewis acids
and bind best anionic oxygen donor ligands such as polyphos-
phates and carboxylates. Of the six elements only magnesium
and calcium have a definite biological role (e.g. enzyme
activation, muscular contraction, skeleton formation); all
others are toxic to varying degrees. Beryllium toxicity might
be due to its capacity for forming strong covalent bonds, thus
displacing other divalent cations. For example, even in con-
~6 — 3
centrations as low as 10 mol dm , it inhibits in vitro
alkaline phosphatase which is magnesium-activated. Barium
toxicity could be due to its interference with the calcium
function.
There is a great similarity in properties between zinc and
2+ 2 +
cadmium, and the Zn and Cd ions are similar in many
respects to Mg , particularly as regards the solubility of
2 +
halides and oxo-acid salts. In contrast to Mg , zinc and
2 +
cadmium ions are rather strong Lewis acids. The salts of Hg
with strong oxo-acids (H»SO , HNO.) are soluble but at the
same time hydrolysis will occur because mercury (II) hydroxide
19
-------
is a weak base. Hg cation forms coordination compounds with
a coordination number 4 but has a pronounced tendency towards
complex formation with a coordination number 2 as well, if
elemental mercury is present, mercury (II) ion will dispropor-
tionate to form mi
metal-metal bond.
tionate to form mercury (I) ion, Hg - Hg , with a strong
Because metals of group lib have a filled set of d-orbitals
interposed between the core and the valence electrons, and
because these d-orbitals provide only a weak shield for the
increased nuclear charge, the ionization. energies of Zn, Cd
and Hg are higher than those of Ca, Sr and Ba. The result is
2+ 2 + 2 +
that the complexes of Zn , Cd and Hg have a considerable
covalent character due to the strong attraction these cations
exercise on electrons. For the same reason, the sulfides of
Zn, Cd and Hg are insoluble. The increase in covalency
with increasing nuclear charge is also demonstrated by the
stability of metal alkyls of mercury, a fact of considerable
toxicological and ecological importance. Zinc has an important
biochemical function as potentiator of many enzymes. In
contrast, cadmium and mercury have no known biological role.
Cadmium toxicity may be related to its similarity to zinc.
In the mammalian kidney, cadmium has been found in a ternary
Cd-Zn protein complex, metallothionein, which has up to 10%
cadmium and a very high content of cysteine residues. The
toxic effects of mercury appear to be related to its ability
to bind thiol groups, which is a consequence of the affinity
of both cadmium and mercury for sulfur ligands.
Boron, the first element of group Illb, is a non-metal.
Aluminum, the next member of the group, has a larger atomic
radius and consequently a lower ionization potential. There-
fore, it behaves like a metal and so do other members of the
group, i.e. Ga, In and Tl. The salts of strong oxo-acids are
soluble, but those of weak acids undergo extensive hydrolysis.
All the members of this group form compounds in the oxidation
state +3, but because of the very low first ionization poten-
tial, in some cases univalent ions are also known in aqueous
20
-------
solutions (e.g. In+ and Tl+). In fact Tl is usually more
stable than Tl . Chemically, Tl is similar to K and Ag .
Complex formation is confined almost entirely to ligands
having oxygen donor atoms and to the halide ions.
In group IVb, only lead has strongly pronounced metallic
properties. When the cations are in the oxidation state +4,
the high charge and the small size will provide a strong
polarizing power leading to compounds of covalent character.
The stability of compounds in +4 oxidation state decreases
down the group. When in lower oxidation state +2, tin
and lead form typical ionic compounds. The covalent char-
acter of the +4 oxidation state is reflected in the wide range
of organometallic compounds formed by germanium, tin and lead.
The elements of group Vb and group VIb show an increasingly
non-metallic character. In group Vb, antimony and bismuth
are considered to be metals, but they show many non-metallic
properties. In group VIb only polonium is a metal.
None of these metals has a defined biochemical role. Gallium
and indium are moderately toxic, but thallium is a very toxic
element which seems to interfere with sodium-potassium metab-
olism. Germanium has no known biological role; the role of
tin is uncertain and lead is very toxic.
5 .6 Transition metals
The usual definition of the transition elements is that they
have, as elements, partly filled d or f shells. It is, however,
more appropriate to include also those elements which have
partly filled d or f shells in any of the oxidation states
in which they form compounds. This broader definition includes
among transition elements also Cu, Ag and Au. There are about
56 transition elements of the d- and f-block. All of them
have some common properties:
21
-------
1) they are all metals;
2) with a few exceptions, they exhibit variable oxidation
rates;
3) because of partly filled shells, they form at least some
paramagnetic compounds;
4) their ions and compounds are colored in one or all
oxidation states.
Properties 2) and 3) are of great biological importance and
make some of them particularly suitable for biological
catalysis and for electron transport function.
The transition elements are further subdivided into three
main groups:
a) the main transition elements or d-block elements
b) the lanthanide elements, and
c) the actinide elements.
The chemistry of transition elements is extremely complex
and the present discussion will be limited to the d-block
elements. Among them, of particular biological and toxicol-
ogical interest are some members of the first row, i.e. Ti,
V, Cr, Mn, Fe, Co, Ni and Cu. In the second row Mo is
certainly of greatest toxicological interest, and to some
extent Ru, Rh, and Pd. In the third row attention should
be paid to the triad Os, Ir and Pt.
The chemistry of the d-block cations is dominated by their
ability to act as Lewis acids forming complexes with Lewis
bases. With ligands of biological interest having nitrogen,
oxygen and sulfur donor atoms, the most stable complexes
are formed in the +2 and +3 oxidation states. An important
feature of such compounds is that under very similar conditions
one can obtain complexes that differ only with respect to
the oxidation state of the metal. Examples are Co (NH_)J
T III "13+ L 3 6J
and Co (NH-), which can both be prepared at room
L 3 6J XI
temperature. The Co complex can be easily oxidized by oxygen
to the Co complex. The process can be easily reversed by
a suitable reducing agent. This ability to move easily between
22
-------
the two adjacent oxidation states is particularly valuable
in the biological transport of electrons.
The simplest theory that has been developed to explain the
behavior of transition metal ions and complexes is the
crystal field theory. This theory makes the assumption that
all ligands, including water, can be treated as point negative
charges, subject only to the laws of electrostatics.
The reactions of the transition-metal complexes include
ligand substitution, redox processes and reactions of co-
ordinated ligands. These and other topics of interest for
the understanding of the biochemistry and toxicology of
some transition metals are discussed in detail in a monograph
by Hughes (1972).
7. Biological molecules containing metals
Many metals have very important biological functions which
they achieve in most cases in the form of complexes with
other biological molecules. An excellent treatment of this
aspect of biochemistry may be found in Hughes' monograph
(1972).
7.1 Metalloporphyrins
The metalloporphyrins include two extremely important
ones: the chlorophyll molecule and the heme group.
The ability of the chlorophyll molecule to absorb light
is related to the conjugated polyene structure of the
porphyrin ring. Magnesium ions which are coordinated to
the nitrogen atoms of the four pyrrole rings have at least
two functions: to provide the necessary structural rigidity
and to increase the rate of conversion of the singlet ex-
cited state resulting from photon absorption into the
triplet state which enables the transfer of the excitation
energy into the redox chain.
The two main functions of iron-containing biological complexes
are the transport of oxygen and mediation in electron trans-
23
-------
fer chains. The heme group is in all cases associated with
a protein molecule as in hemoglobin, myoglobin, cytochromes
and enzymes such as catalase and peroxidase. Cytochromes serve
as electron carriers and the heme-containing enzymes, cata-
lase and peroxidase, catalyze the decomposition of hydrogen
peroxide.
7.2 Non-heme iron proteins
Non-heme iron proteins, such as rubredoxins, ferredoxins
and high potential iron proteins (Hi PIPs) contain strongly
bound functional iron atoms attached to sulfur, but they do
not contain porphyrins. All of them participate in electron
transfer processes.
The third group of iron-containing biological structures
are ferritin and hemosiderin. Both serve to store iron in
a protein structure. Transferrin binds ferric iron and
transports it from ferritin to red cells. In microorganisms,
iron is transported by ferrichromes and ferrioxyamines,
structures containing cyclic or acyclic polypeptide chains.
7.3 Cobalt-containing biological molecules
The best known biological molecules containing cobalt are
the vitamin B, ~ coenzymes (cobalamines). Cobalamines consist
of a cobalt atom, a microcyclic ligand corrin, and a complex
organic part containing a phosphate group, a sugar and an
organic base also coordinated to the cobalt atom. Methyl-
cobalamine is involved in the methane-producing bacteria
and transfers the CH. group to Hg11, Tl111, Pt11 and Au1.
7.4 Metalloenzymes and metal-activated enzymes
Some enzymes incorporate one or more metal atoms in their
normal structure. They are called metalloenzymes of which
at least 50 have been identified. About 20 of them incorporate
zinc. The best known members are carbonic anhydrase and
carboxypeptidase. The structure of carboxypeptidase is
partly known. The zinc ion is bound in a distorted tetrahy-
dral configuration , with two histidirie-nitrogen atoms,
one glutamate carboxyl oxygen atom and a water molecule as
ligands.
24
-------
There are about 15 copper-containing metalloenzymes. Their
structure is not well understood. Examples are ascorbic acid
oxidase and various tyrosinases. In many lower animals (crabs,
snails) the oxygen carrying molecule is a copper-containing
protein hemocyanin which, however, does not contain any heme
groups.
Metalloenzymes containing molybdenum and iron (nitrogenases)
play an important role in nitrogen fixation.
Metal-ions can be bound to proteins in a reversible way.
This is the case with metal-activated enzymes. Such systems
are much less amenable to study than metalloenzymes,
because they cannot be isolated with metal ion in place.
Most enzymes associated with phosphate group transfer
or hydrolysis appear to be activated by Mg
25
-------
REFERENCES (General)
Bell, C.F. (1977). "Principles and Applications of Metal
Chelation." Clarendon Press, Oxford.
Brakhanova, I.T. (1975). "Environmental Hazards of Metals."
Consultants Bureau, New York.
Cotton, F.A. and Wilkinson, G. (1972). "Advanced Inorganic
Chemistry." 3rd edition. Interscience Publishers, New
York.
Cotton, F.A. and Wilkinson, G. (1976). "Basic Inorganic
Chemistry." John Wiley & Sons, New York.
Hughes, M.N. (1972). "The Inorganic Chemistry of Biological
Processes." John Wiley & Sons, New York.
Parr, G. (1973). "Ions in Solution (3), Inorganic Proper-
ties." Clarendon Press, Oxford.
Pauling, L. (1970). "General Chemistry." W.H. Freeman &
Co., San Francisco.
Phipps, D.A. (1976). "Metals and Metabolism." Clarendon
Press, Oxford.
26
-------
Table 1.
Periodic table of the elements
P, . ,.jH
|
Is
2
2s ?D
3
3s" D
4
4-;d
5
Sc
K
'"£
7
7s
(5-",
Croup
la
1
H
3
Li
1 1
Na
19
K
17
Rb
V
s/
Cs
87
Fr
•L.'-itria
se-ie
••^c:,r
se-ie
sr
Grojp
111
,
4
Be
12
Mg
20
Ca
38
Sr
k
Ba
88
Ra
Tide
s
:de
s
Group
;,ia
2!
Sc
39
Y
57-
La
89"
Ac
58
Ce
90
Th
IVa
22
Ti
40
Zr
77
Hf
59
Pr
91
Pa
Group
Va
23
V
41
Nb
73
Ta
60
Nd
92
u
Group
Via
24
Cr
42
Mo
74
W
61
Pm
93
Np
Qroup
Mla
25
Mn
43
Tc
/
75*
Re
62
Snn
94
Pu
26
Fe
44
Ru
c
76
Os
63
Eu
f
95 J
Am
Gr-.,p
27
Co
45
Rh
7/
Ir
61
Gd
L
Ts6
Cm
28
Ni
46
Pd
' 78
Pt
65
Tb
97
Bk
Group
Ib
29
Cu
47
79
Au
66
Dy
98
Cf
Group
lit)
30
Zn
48
Cd
80
Hg
67
Ho
99
Es
Gruup
II, 0
5
B
13 I
Al
31
Ga
49
In
/
81 »
TI
68
Er
1 00
Fm
Croup
IVD
6
c
14
Si
32
Ge
50
Sn
>
e:
Pb
69
Tm
IOI
Md
Group
Vo
7
N
15
P
33
As
51
Sb
83
Bi
70
Yb
I0?
No
^r0ur
Vlb
6
0
16
S
34
Se
S2
Te
84
Po
71
Lu
I03
Lr
Group
Vliti
I
H
9
F
I 7
Cl
3S
Br
53
I
85
At
.jr.'up
o
2
He
10
Ne
11
Ar
36
Kr
54
Xe
8-3
Rn
21
-------
r
Figure 1. Shapes of atomic orbitals
28
-------
Electron density
A*
Figure 2. Electron distributions in (a) homopolar covalent
bond (b) heteropolar covalent bond with partial
ionic character (c) ionic bond
29
-------
SAMPLING AND ANALYTICAL METHODS
Magnus Piscator and Velimir Vouk
Dose estimation involves several analytical steps, and the
final result will of course depend on how well both the
sampling and the analysis have been performed. In many well
designed studies from the epidemiological point of view
where great care has been taken to obtain good response
data, the analytical methods used have not been checked for
accuracy. The results of these studies could therefore not
be used for dose-response estimation. There have also been
many instances in which investigators have combined data
obtained in epidemiological studies with data on doses of
metals in different media recorded in previously published
studies. An attempt to construct dose-response curves in
that way may often be disastrous. For example, in a recent
paper cancer mortalities in different countries were related
to assumed dietary intakes of certain metals, and the daily
intake of cadmium in Sweden and USA was assumed to be about
twice as high as that in Japan (Schrauzer et al., 1977). The
data on intake had been published more than 10 years ago.
Recent data (see Friberg et al., 1977) obtained by modern
analytical techniques were completely neglected. The opposite
may also happen, i.e. laboratories with good analytical
experience, capable of making accurate measurements on
metals in different media, may start conducting epidemiological
studies which they are not capable of designing properly.
Results concerning dose-response hereby obtained may be
quite misleading.
This chapter provides some general comments about sampling
and analysis which are intended primarily to aid industrial
physicians, epidemiologists and toxicologists to interpret
analytical data at their disposal. It is not intended to
give details on the advantages and disadvantages of specific
sampling techniques and analytical methods. Such information
is contained in the chapters of Volume II of this series.
metal by metal.
31
-------
Much information on sampling and analysis can be found in
the two volumes "Accuracy in Trace Analysis, Sampling,
Sample Handling, Analysis" edited by LaFleur (1976) and in
the monograph by Winefordner (1976).
1. Sampling
1.1 General considerations
The exposure dose may be estimated from measurement of metal
concentrations in environmental media (air, water, food) as a
function of time. Determination of exposure dose is of course
of great importance since it makes it possible to observe trends in
pollution. Of still greater importance is the assessment of
the dose in tissues or organs. This can be done either by
applying appropriate metabolic models to the exposure dose
measured, or by measuring concentrations in the tissue or organ
of interest directly. Another possibility is to estimate
tissue or organ dose from the measurement of the concentration
of the metal in index media, e.g. blood, urine, hair etc. Here
again we need metabolic models.
Only in a few instances can exposure dose be directly related
to absorbed dose. For example, it is known that methylmercury
is absorbed nearly 100% from the intestine. By measuring
methylmercury in fish meat and at the same time knowing daily
intake of fish, it is thus possible to calculate the dose. For
many other metal compounds, however, accurate data on the rate
of absorption from the lungs or the intestine are lacking
and internal dose measurements must be made.
1.2 Air
When evaluating exposure to metals in air it is important to
know not only the actual concentrations but also the particle
size distribution and the solubility of the metal compounds.
This is especially useful when discussing occupational
exposure, where both highly soluble and highly insoluble
particles of different sizes may occur at the same time.
32
-------
1.2.1 Ambient air
The concentrations of metals in ambient air are generally low,
and intake through air is usually small in relation to intake
via food. Exceptions are areas in the vicinity of plants
emitting large amounts of metals/ e.g. smelters, and heavily
trafficked areas, where the intake of lead through air may
exceed the intake with food and water. In some countries
a continuous measurement of metals in both rural and urban
air has been carried out for many years. Sampling has been
performed at stationary sites often located in such a way that
the exposure of people living in these areas could not be
estimated adequately. The great significance of these continuous
measurements is that they provide information about trends in
general air pollution. Moreover, combined with data on the
fall-out of metals, they may be used for predicting the long-
term accumulation of metals in soil and in certain food chains.
There is an abundance of data from many countries on the
levels of lead in air in areas with heavy traffic as well as
in rural areas. Such measurements have often been made with
stationary samplers. Attempts to correlate the levels of
lead in blood in persons in these areas with the lead levels
in air measured there, have generally shown no or very weak
correlations. These findings stem from the fact that the
subjects under investigation have a great mobility and are
exposed to varying concentrations of lead during the day,
whereas a stationary sampler only determines lead at a
certain point, generally not in the breathing zone. The
varying intake of lead via food also makes it difficult to
define the relationships between lead in ambient air and
lead in blood. There is one study (Azar et al. (1973) where
personal samplers were used for continuous recording of
airborne lead, and where blood lead was repeatedly determined
in several exposed groups. However, as the authors themselves
pointed out, the lack of adequate data on dietary intake
made it difficult to determine the quantitative relationship
between airborne exposure and blood levels of lead. Reconwndations
on sampling systems for ambient air are found in reports by
Katz (1969) and WBO (1977). Statistical problems concerning
air pollution measurements have been treated by EPA (1976).
33
-------
Long-term deposition of metals through fall-out is also of
interest. The fall-out is generally collected in a deposit
gauge of standardized size which makes it easy to compare
data from different countries. In most cases the sample is
collected for one month and the results are expressed as mass
per m and month. The deposit gauge must be protected from conta-
mination by birds and other animals. During its long period of
use, the method has proved to be of great value for monitoring
emissions from plants.
Another valuable means of arriving at the long-term deposition
of airborne metals is to collect and analyze certain types of
mosses growing on rocks or other supports, and taking their
nutrients mainly from the air. These mosses have a great
capacity for absorbing metals. In the Scandinavian countries
(Riihling and Tyler, 1973) and the U.K. (Goodman and Roberts,
1971) this approach has provided valuable information.
1.2.2 Industrial air
Occupational exposure takes place mainly by inhalation and
good sampling techniques and accurate analysis are thus essential
for evaluating the exposure dose in such environments. The
concentrations of metals in air are generally much higher than
in ambient air, which makes it easy to collect sufficiently
large amounts. The analysis will also be easier, since the
interference from other elements is much less than in the
ambient air. One great problem still lies in obtaining informa-
tion on the amount of metal in the breathing zone of the exposed
workers.
Previously, stationary samplers were generally used and the
results have been of value when looking at trends in the
working environment. Such results cannot, however, be used
for estimating individual exposure, except in the case of
workers who have a great mobility in the work shops, e.g.
foremen. During recent years personal samplers have been more
and more in use, and do of course give better estimates of the
amount of a metal actually inhaled. Several points must be
taken into account when interpreting data obtained by personal
sampling.
34
-------
1. Dust on the worker and his clothing may contaminate
the filter, resulting in too high estimates.
2. The personal sampling data for some metals such as
cadmium may give a very inaccurate picture of the
total exposure, if personal habits, especially smoking,
which might increase the exposure substantially are
not considered (Piscator et al., 1976). Adamsson
et al. (to be published) found that workers exoosed
to cadmium oxide dust (<10 ug/m ) had a fecal
elimination of cadmium in some cases of 1-3 ing, only
part of which could be attributed to mucociliary
clearance, the main part probably coming from oral
contacts with contaminated hands or cigarettes.
3. The concentrations of a metal, as well as other
substances, measured in air at a certain point have a
log-normal distribution, for which reason the
distribution should be summarized in terms of the
geometric mean and standard deviation. For
estimating Individual exposure and compliance or
noncompliance with occupational health standards
the arithmetic mean should be used, since that
reflects the actual inhaled amount. Confidence limits
for the arithmetic mean, however, must be calculated
on the basis of log-normal distribution. For
details see Coenen (1966), Sherwood (1971), Leidel
and Busch (1975), and Ulfvarson (1977).
1.3 Water
In areas with running water, cool tap water is generally
used for preparing food, but hot water may also be used/
i.e. for baby food. It might thus be of importance to
determine metals in both the cool tap water and the
hot tap water. The amount of tap water imbibed directly,
versus that used for preparing hot beverages and for cooking
food must be carefully estimated, since the biological
35
-------
availability of metals may be quite different under the
two circumstances. The intake of drinking water may vary
considerably depending on individual and local habits and
on climate (ICRP, 1975; Zoeteman and Brinkmann, 1976) .
Since water pipes may contain materials that can release
certain toxic metals in larger amounts, concentrations of
metals should be examined in the first portion of water tapped
after the faucet has been turned off for some time,
and also in the water after it has been runninq
for some time. Especially lead and cadmium may accumulate
during longer periods when water is not tapped. Estimates of
community exposure based on data from the analysis of metals in
water from waterworks may lead to erroneous conclusions. Here
and otherwise it should be kept in mind that people may often
drink water from different supplies during the day (work-place,
travel etc.). Furthermore, factors like hardness and pH should
be determined, since they might be of importance for exposure
evaluations. Since the concentrations of most of the metals of
interest are low, extremely good care must be taken in cleaning
the vessels for collecting of samples. The glass or plastic
bottles must be acid-cleaned and, before use, checked for leaching
of the metals under study. For more details see EPA (1972) ,
APHA (1976),'and Maienthal and Becker (1976) .
1.4 Food
At least three methods are available at present for estimating
intake of metals via food products. One method is to collect
samples of single foodstuffs, analyze them for the metal
under consideration, and then estimate the amount of metal that
could be ingested by people with different food habits, given
the available food statistics. A second method is to collect
certain classes of foods, i.e. vegetables, dairy products,
fish, and meat products in the amounts that are actually
consumed, analyze each class and make estimates from that.
Prior to analysis, each food should be treated as it would be
normally, e.g. cooked. The so-called market basket studies
in the USA have been based on this method (for details see
Mahaffey et al., 1975). A third method is the duplicate samples
36
-------
method. During a certain period, the people under study put
meals identical to the ones they have eaten into a vessel.
The deposited meals can be homogenized and analyzed, providing
a total intake figure. The first of the three mentioned
methods is the simplest and has been recommended by FAO/WHO
(1977). The disadvantage is that only raw products are
examined. The second method requires larger resources, but
gives more accurate estimates. The third method gives good
information on individual intakes, but can only be performed
on a limited number of subjects. For details on sampling and
analysis of foods, see Crosby (1977).
Another method which can be used on the general population
is to determine the fecal elimination of the metal in
question. Results of studies in Japan, USA, and Sweden show
that the estimates of dietary intakes of a metal like
cadmium with low gastrointestinal absorption thus obtained
are of the same magnitude as those obtained by the more
laborious methods described above (see Friberg et al., 1977).
Occupationally exposed persons must be carefully excluded
from such studies since they might eliminate in feces metals
cleared from the lungs. The same care must obviously be
taken as for water samples in preparing the collecting
vessels.
1.5 Body fluids and tissues
Metal concentrations in blood and urine are generally low
and great care must be taken to avoid contamination from
sampling equipment. Acid-cleaned test tubes must be used for
collection of blood samples. Preventive measures must be
taken lest some materials, like the rubber in stoppers,
release metals into the sample. Blood is usually taken from
the cubital veins and stainless steel needles suitable for
collection of blood for serum-iron determination may be used
to extract samples for most metals. Special care must be
taken when chromium or nickel are to be analyzed, since
these metals may be present in steel. In the industrial
environment caution must also be exercised that the metal
37
-------
under investigation is not transferred from the skin to the
needle. A special problem is the collection of blood from
small children when lead poisoning is suspected. It is in
such cases often impossible to obtain samples from the
cubital vein and instead, capillary blood is drawn from the
finger tips. There is a risk for severe contamination from
the skin when sampling that way.
Urine should be collected in acid-cleaned plastic bottles
and if exposed workers are being examined, urine specimens
should be collected at home to avoid any contamination with
dust from the working environment. Morning samples are
superior to random samples as far as standardization of the
procedure is concerned. Determination of specific gravity or
creatinine in urine is useful when correcting for the great
variations in dilution of urine.
Data on metal concentrations in organs are generally obtained
from autopsy samples. There are very few instances in which it
has been possible to study the metal concentration in organs
from living people.
Generally tissue sections can be obtained with stainless
steel knives. Exceptions are again nickel or chromium for
which special glass knives or other suitable instruments
should be used. When sampling some organs, e.g. kidney, it
is important to cut from the same part every time. Many
metals are unevenly distributed in the different layers of
the organ.
Blood, urine and organ samples can be stored for a long time
before analysis, but if the sample volume is small in relation
to that of the container, the wet weight might change due to
the loss of water. For more details, see Mertz (1975) , Anand
et al. (1975), Maienthal and Becker (1976), and a recent
report by CEC/WHO/EPA (1977).
38
-------
2. Analytical methods
2.1 General aspects
Assuming that correct sampling has been performed, the next
step is quantitative analysis. Regardless of which analytical
method is used, some points must always be carefully considered.
These are accuracy, precision and detection limit. Accuracy
can be defined as a systematic deviation from the "true
value". Precision is generally expressed in terms of the
coefficient of variation, also called RSD (relative standard
deviation), arrived at by analyzing the same sample a number
of times over a certain period. Precision should be measured
under conditions that correspond to the ones prevailing in
the regular analytical work. It should be noted that both
accuracy and precision depend on the concentration of the
metal to be analyzed. Limit of detection is the minimum
amount or concentration that the analytical method is able
to distinguish from a zero concentration. Limit of detection
is of less importance from the toxicological point of view
than the other two measures of analytical performance. It
may often be sufficient to know that a concentration is
below a certain level, provided that accuracy and precision
are good.
Analysis involves many steps, from the moment the sample
arrives at the laboratory until the final reading is obtained.
The main ones are homogenization of samples, ashing, extraction,
separation, and chemical or physical analysis (see e.g.
Crosby, 1977). The media to be examined may exhibit large
differences in matrix composition and metal content. There
are no standard procedures applicable to all types of media.
For example, the determination of cadmium in renal cortex is
a relatively simple procedure, whereas that in blood and
urine is highly complicated due to the low concentration of
cadmium and the complex matrix composition, with many interfering
substances.
The information in the following sections relies to a large
extent on the monographs by Winefordner (1976), and Pecsoc
et al. (1976).
39
-------
2.2 Current methods for analysis of metals
Some twenty years ago, colorimetry, emission spectrography,
and polarography were the dominating methods for the quantitative
determination of metals. Atomic absorption spectrometry was
then in the introductory stages for trace metal analysis. By
the middle of the 1960's, atomic absorption equipment had
become commercially available and this method was being
employed routinely for quantitative determination of metals
by more and more laboratories. Tremendous progress in the
refinement of the atomic absorption method has taken place
during the late 1960's and the 1970's. At present, it is the
most commonly used method. However, many of the earlier
methods are still employed and new methods are being developed
which may become strong competitors to atomic absorption
methods in the future. In the following sections, some of
the most common methods will be discussed mainly with regard
to basic principles, applicability and limitations. Details
pertaining to the use of a given method for a specific metal
are taken up in Volume II of this series.,
2.2.1 Spectroscopic methods
2.2.1.1 Colorimetry and absorption spectrophotometry
Colorimetry, the first modern method to be developed for
metal analysis, is based on the formation of colored complexes
between metals and organic compounds, whose absorbance can
be measured in a colorimeter or spectrometer, after extraction
into a suitable solvent. Using current terminology colorimetry
may also be classified as electronic absorption spectrometry.
The basic principle is simple, i.e. a specific complex is
formed which absorbs light in a given spectral region. Only
a few metals form specific complexes that can be directly
extracted and analyzed, as in e.g. the classic procedure for
determination of serum-iron with bathophenantroline. For
most metals, several separation steps must be taken before
the complex can be extracted. A typical example is the
determination of cadmium with dithizone where cadmium dithizonate
must be separated from e.g. zinc and lead dithizonates.
40
-------
Colorimetric methods still have many applications since the
more modern methods are not suitable, or in some cases, have
not yet been worked out, for some of the metals. NIOSH
(1974) recommended that antimony in urine should be determined
by a colorimetric method, the Rhodamine B method. SCOPE
(1975) recommended a method involving methylene blue for
selenium in air and a dithiocarbamate method for arsenic in
air. Metals of greatest interest at present, e.g. cadmium,
lead, and mercury, can all be determined with colorimetric
methods, especially using dithizone, but atomic absorption
has been shown to be superior. The limitations of the colorimetric
methods are mainly that much work has to be performed in
order to isolate the specific complexes. This work is expensive,
time-consuming and entails a high risk of contamination,
as well as occupational hazards to the analysts.
2.2.1.2 Emission spectrometry
Atomic emission spectrometry is based on the excitation of
metal atoms by flames or by electric discharges. The excited
atoms then lose their acquired energy by emitting photons.
This emission can be recorded on a spectrometer or spectrograph
which has a monochromator that separates the emitted radiation
according to wave-lengths (i.e. the energy of photons). A
spectrum is obtained and a very wide range of metals can be
more or less quantitatively determined. Atomic emission
spectrometry is especially useful for multi-element analysis,
e.g. for soil analysis, geological surveys etc. It was used
by Tipton and coworkers (1963) in extensive investigations
on the metal content in different organs of the human body,
but accurate data could not always be obtained. Atomic
emission spectroscopy is also used in the metal industry for
the quality control of products. In many industries spectroscopic
equipment is also used for monitoring the exposure of workers
to such metals as lead and cadmium. Emission spectrometric
methods are not suited to the determination of easily volatile
substances like selenium, arsenic or mercury.
41
-------
The standards and the samples to be analyzed must have
matrices of the same composition, since the intensity of the
emissions can be affected by the matrix. The electrical
discharge sources may vary. For example, the DC arc source
does not give a better reproducibility than about 20%. The
DC arc enables, however, a low detection limit. Other spark
sources may give a better reproducibility, but a higher
detection limit. For the best type of equipment, the cost is
very high.
2.2.1.3 Atomic absorption spectrometry
The basic principle behind atomic absorption spectrometry is
that metals in the ground state will absorb radiation from a
light beam with the same spectral composition as the light
emitted by the element under consideration. Cadmium atoms
thus absorb radiation from light emitted by a cadmium lamp.
The decrease in intensity as compared to a blank corresponds
to the concentration of the metal. There are two main methods
for atomization of a sample, the flame method and the furnace
method, the latter also being called electrothermal atomic
absorption. The flame methods use different mixtures of
gases for creating a high-temperature atomization flame,
e.g. air-acetylene. Different types of flames may be used
for different metals. Flame methods are generally used for
liquid samples which can be aspirated directly into the
flame. An important modification is the technique developed
by Delves (1970) where a small amount of a liquid sample,
e.g. blood is put into a microcrucible made of nickel and
then inserted in the flame. During the last years, the
flameless methods have undergone a rapid development. Most
common is the graphite tube furnace. The principle is that
after drying in the furnace, the temperature is raised to
atomize the sample which then passes through the light path.
Several milliliters of solution must be used in the conventional
flame method, while only 1-100 ul need be applied to the
furnace. The standard solution, preferably in a similar
matrix, is treated in the same way. Atomic emission spectrometry
has found a number of applications and has been especially
42
-------
useful for the determination of lead, cadmium, and mercury
in biological materials and in exposure media as well as for
the determination of many essential elements, e.g. copper
and zinc. Recently methods have also been developed for
nickel and arsenic in different media. Special modifications
are required for highly volatile metals such as mercury and
arsenic. One limitation is that atomic absorption is not
suited for multi-element analysis. Also, some non-specific
absorption may occur due to the presence of other atoms and
molecules in the flame. Salts such as sodium chloride and
phosphates are especially apt to cause interference. This
may be avoided by background correction, for which a deuterium
lamp is generally used. Many of the modern apparatuses have
such a lamp built in. In some cases, this type of background
correction does not suffice for all the non-specific absorption.
For example, in the direct analysis of cadmium in urine,
values obtained will be too high. The next step is then to
extract the metal according to similar principles as for
colorimetric analysis and to introduce the metal chelate in
an organic solvent into the flame or the graphite furnace.
The most common chelating agent has been APDC (ammonium
pyrolidine dithiocarbamate), which is extracted into MIBK
(methyl-iso-butyl-ketone). This can be used for several
metals, e.g. cadmium, lead, and nickel.
2.2.1.4 X-ray methods
X-ray methods are relatively new and are not a matter of
routine to any large extent. The basic principle is to
subject the sample to high intensity X-rays or still better,
protons (Johansson, 1971), and to measure the radiation
emitted due to electron displacement in the irradiated
atoms. The spectral lines of emitted X-rays can be resolved
by a crystal. A multi-element analysis is thus possible to
perform by this method. A great advantage is that very small
amounts can be analyzed-, for example, a single hair sample
can be subjected to multi-element analysis. More data are
needed to compare these methods with the more established
procedures before a final verdict can be made.
43
-------
2.2.2 Neutron activation
Neutron activation analysis is based on the formation of
radioactive isotopes when a sample is irradiated by neutrons
from a nuclear reactor or from some other neutron source.
The element of interest is then identified and quantitated
by y-spectroscopy. Neutron activation analysis has been used
mainly for research and has only come into routine work for
the determination of total mercury in fish. Even here, its
role is on the decline thanks to the development of atomic
absorption methods. Neutron activation analysis has been
used for multi-element studies, in which cases complicated
extraction schemes have often been carried out. It should be
noted that lead and tin are not suitable for analysis by
that method. Neutron activation is also often used as a
reference method for testing the accuracy of other methods
since, if properly performed, it has a high specificity and
accuracy. Some limitations are that a reactor is needed and
that the cost of analysis is high.
2.2.3 Electro-chemical methods
Electro-chemical methods are based on the measurement of
changes in electro-chemical potentials, which are due to
transfer of electrons from one element to another. These
methods use electrodes of different types. For example, in
polarography, a dropping mercury electrode is applied. By
changing the charge of that electrode, different metals in
solution will form amalgams with the mercury at different
potentials. In this way, so-called polarographic waves are
obtained, from which quantitative estimates can be made of
the concentration of different metals. A further development
of this method is anodic stripping voltametry which is based
upon the reverse process, i.e. metals are first concentrated
on the electrode, and then released, whereafter the changes
in the potentials are measured. This method has proved
suitable for water analysis, but experience is still limited
as to its value for biological materials. Its suitability
for samples in which it is easy to obtain metals in an ionic
form is obvious.
44
-------
2.2.3 Chromatography
Chromatography is based on differential adsorption of dis-
solved substances from a suitable solvent (mobile phase)
which is passed through a stationary adsorbent (paper,
silica gel etc.). Chromatographic methods are of great
importance for separating and identifying metal compounds,
but for final quantitative analysis only gas chromatography
has been routinely used. Chromatographic methods are also
used in the pretreatment procedures, e.g. ion exchange has
been used in preparation for atomic absorption analysis or
for multi-element neutron activation procedures.
In gas liquid Chromatography the stationary phase is a
column filled with a substance in liquid state that adsorbs
organic compounds. The mobile phase is a gas. The sample in
an organic solvent is injected into the system and carried
by the gas to the column, where components are adsorbed and
separated. They will then pass a detector, yielding peak
recordings, which enable their quantification. Methylmercury
and alkyllead especially have been determined by this method.
The determination of alkyllead in gasoline is a relatively
simple procedure, since the organic metal compound is already
in a solvent, whereas several pretreatment steps must be
performed before the methylmercury in fish can be extracted
into a solvent.
Very small amounts of alkyllead and alkylmercury can be
detected by gas Chromatography. It is possible to analyze
for other metals by converting them to organic complexes,
but this has not yet been done on a routine basis.
2.2.4 Spark source mass spectrometry (SSMS)
The basis for spark source mass spectrometry is the formation
of metal ions when metals are subjected to high energy dis-
charges. A beam of metal ions is then directed into a mass
spectrometer where they will separate into a magnetic and
electrostatic field according to their mass and charge. The
mass spectra can be recorded on a photoplate, from which
45
-------
quantitative estimates can be made. All elements from lithium
to uranium may be determined. Non-radioactive isotopes can
be separated by this method and identified, as has been done
in studies on lead metabolism.
Spark source mass spectrometry has been used in several
studies on metals in food, blood, organs etc. Its limitations
are above all the high cost of the equipment and the high
skill required in operating and maintaining it. Like neutron
activation analysis, SSMS will surely serve well in the
future as a reference and special research method.
2.3 Quality control and inter-laboratory comparisons
The precision of a method can be tested by performing repeated
determinations on the same samples, but to ensure accuracy,
it is often necessary to go beyond the resources of one's
own laboratory and, for example, to use so-called reference
samples, among which the most well-known are the NBS (National
Bureau of Standards, U.S.A.) standard reference materials. A
typical example of an NBS standard for a metal is the bovine
liver, which many investigators have used when checking
their methods. While the bovine liver is excellent for
testing methods for the determination of metals in liver and
other tissues with a similar composition, it cannot be used
to guarantee accurate results for e.g. blood. Unfortunately,
many results have been reported where methods for the analysis
of blood have been assessed by using other standards. The
only appropriate references for checking the accuracy of a
method for determination of metals in blood are blood samples
with a known content of the given metal.
Recovery tests are useful when checking for losses etc. in the
different steps in an analysis, but should not be used to prove
accuracy.
Another way to ensure accuracy for a method is to determine
the metal content of a sample with other methods which are
based on different principles. Thus determination with neutron
46
-------
activation, and spark source mass spectrometry, should give
a good idea of the accuracy of a method using atomic absorption
spectrophotometry. When a method has been tested and found
to be accurate it should also be subjected to future checks
through regular analysis of some standard sample.
The need for accurate analytical methods has stimulated an
active interest in collaborative studies, where several
laboratories participate. Examples are the studies by Keppler
et al. (1970) , Lauwerys et al. (1975) , and Kjellstrom et al.
(1975). The results of these studies have been disappointing,
showing very large differences for different types of samples.
Lauwerys et al. (1975) concluded "... several laboratories
that took part in this comparison programme have not yet
adequately developed the technique required for precisely
measuring Pb, Cd and Hg in blood, urine and water." For details
on statistical procedures for use in collaborative tests,
see Youden (1975) and Steiner (1975).
3. Conclusions
During the last few years, the need for standardizing sampling
procedures and for having good programmes for the quality
control of analytical results has been recognized. The
analytical techniques available today are often extremely
sensitive. Before any new method is used in environmental
monitoring, it must be checked for accuracy by comparisons
with other established methods within and outside one's own
laboratory. If the analytical data are not accurate, erroneous
conclusions with regard to dose-response may be drawn and
these may be highly misleading in the evaluation of health
hazards.
47
-------
REFERENCES
Anand, V.D., White, J.M. and Hipolito, V.N. (1975). Clin. Chem.
20., 595-602.
APHA (1976). "Standard Methods for the Examination of Water and
Waste Water." 14th edition, pp 1193. APHA, AWWA and WPCF, American
Public Health Association, Washington, D.C.
Azar, A., Snee, R.D. and Habibi, K. (1973). In: "Environmental
Health Aspects of Lead." pp 581-594. Published by the Commission of
European Communities Directorate General for Dissemination of Knowledge,
Center for Information and Documentation, Luxembourg.
CEC-WHO-EPA (1977). "International Workshop on the Use of Biological
Specimens for the Assessment of Human Exposure to Environmental
Pollutants." To be published.
Coenen, W. (1966). Staub-Reinhalt. Luft 26, 216-221.
Crosby, N.T. (1977). Analyst 102, 225-268.
Delves, H.T. (1970). Analyst 95_, 431.
EPA (1972) . "Handbook for Analytical Quality Control in Water and
Wastewater Laboratories." U.S. Environmental Protection Agency,
Technology Transfer by Analytical Quality Control Laboratory,
National Environmental Research Center, Cincinnati.
EPA (1976). "Quality Assurance Handbook for Air Pollution Management
Systems." U.S. Environmental Protection Agency, Office of Research and
Development, Environmetal Monitoring and Support Laboratory,
EPA 600/9-76-005, Research Triangle Park.
FAO/WHO (1977). Report of the consultation on the Joint FAO/WHO
Food and Animal Feed Contamination Monitoring Programme - Phase
II, Geneva.
Friberg, L., Nordberg, G. and Piscator, M. (1977). In:
"Toxicology of Metals." Volume II, pp 124-163. Environmental
Health Research Series, 600/1-77-022, Environmental Protection
Agency, Research Triangle Park.
Goodman, G.T. and Roberts, T.M. (1971). Nature 231, 287-291.
ICRP (1975). Report of the Task Group on Reference Man, ICRP
Publication No. 23, Pergamon Press, Oxford.
Johansson, T.B., Akselsson, R. and Johansson, S.A.E. (1971).
Nucl. Instrum. Method. 84, 141.
Katz, M. (1969). "Measurement of Air Pollutants." pp 123. World
Health Organization, Geneva.
Keppler, J.F., Maxfield, M.E., Moss, V.D., Tietjen, G. and
Linch, A.L. (1970). Amer. Ind. Hyg. Assoc. J. 3_1, 412-429.
Kjellstrom, T., Tsuchiya, K., Tompkins, E. , Takabatake, E., Lind,
B. and Linnman, L. (1975). In: "Proceedings of the CEC-EPA-WHO
International Symposium on Recent Advances in the Assessment of
Health Effects of Environmental Pollution." pp 2197-2213.
Published by the Commission of European Communities Directorate
General for Dissemination of Knowledge, Center for Information and
Documentation, CID, Luxembourg.
48
-------
La Fleur, P.D., ed. (1976). "Accuracy in Trace Analysis: Sampling,
Sample Handling, Analysis." Proceedings of the 7th Materials
Research Symposium, Vol. I and II, pp 1304. US. Government Printing
Office, Washington, D.C.
Lauwerys,R., Buchet, J.P., Roels, H., Berlin, A. and Smeets, J.
(1975). Clin. Chem. 2_1, 551-557.
Leidel, N.A. and Busch, K.A. (1975). "Statistical Methods for the
Determination of Noncompliance with Occupational Health Standards."
HEW Publication No. (NIOSH) 75-159. National Institute of Occupational
Safety and Health, Cincinnati.
Mahaffey, K.R., Corneliussen, P.E., Jelinek, C.F. and Fiorine,
J.A. (1975). Environ. Health Perspect. 12, 63-69.
Maienthal, E.J. and Becker, O.A. (1976). Interface 5_, 49-61.
Mertz, W. (1975). Clin. Chem. 21, 468-475.
NIOSH (1975). "Manual of Analytical Methods." U.S. Department of
Health, Education and Welfare, HEW Publication No. (NIOSH) 75-1211,
Cincinnati.
Pecsok, R.L., Shields, L.D., Cairns, T. and McWilliam, I.G. (1976).
"Modern Methods of Chemical Analysis." 2nd edition, pp 573.
John Wiley and Sons, New York.
Piscator, M. , Kjellstrom, T. and Lind, B. (1976). Lancet 2_, 587.
Ruhling, A. and Tyler, G. (1973). Water Air Soil Pollut. 2t 445-455.
SCOPE (1975). "Environmental Pollutants - Selected Analytical
Methods." pp 277. Butterworths, London.
Schrauzer, G.N., White, O.H. and Schneider, C.J. (1977). Bioinorg.
Chem. 1_, 35-56.
Sherwood, J. (1971). Amer. Hyg. Assoc. J. 22, 810-816.
Steiner, E.H. (1975). In: "Statistical Manual of the Association
of Official Analytical Chemists." pp 65-88. Washington, D.C.
Tipton, I.H., Cook, M.J., Steiner, R., Boyes, C.A., Perry, H.M.
and Schroeder, H.A. (1963). Health Phys. 9_, 89-102.
Ulfvarson, U. (1977). Scand. J. Work Environ. Health 3, 109-115.
WHO (1977). "Air Monitoring Programme Design for Urban and Industrial
Areas." pp 46. WHO Offset Publication No. 38, World Health Organization,
Geneva.
Winefordner, J.D., ed. (1976). "Trace Analysis - Spectroscopic Methods
for Elements." Chemical Analysis Series, Vol. 46, John Wiley and
Sons, New York.
49
-------
Youden, W.J. (1975). In: "Statistical Manual of the Association of
Official Analytical Chemists." pp 1-63. Washington, D.C.
Zoeteman, B.C.J. and Brinkmann, F.J.J. (1976). In: "Hardness of
Drinking Water and Public Health." pp 173-202. Pergamon Press,
Oxford.
50
-------
SOURCES, TRANSPORT, AND TRANSFORMATION
OF METALS IN THE ENVIRONMENT
Karin Beijer and Arne Jernelov
1. Sources
Some of the most important natural sources of metals in the
atmosphere are the release from surface waters, soils and
vegetation, volcanic activity and forest fires. Metals in
the surface waters originate to a large extent from bedrock,
surface run-off and atmospheric deposition. The major
anthropogenic sources for metals in the environment are the
combustion of fossil fuels, mining and smelting operations,
processing and manufacturing industries, farming and forestry
(fertilizers and pesticidesj increased soil erosion) and waste
disposal (dumping, release of domestic sewage, scrap metal
handling etc).
The world production of metals is shown in Table 1. The data
on the volume and increase/decrease of the production of a
metal may give an indication of the extent of its possible
mobilization in the environment and thereby the pollution
problem involved. Table 2 lists the uses of some metals. A
very large part is used in the production of alloys and in
plating. The chemical industry utilizes an increasing amount
of metals as catalysts. In the production of plastics
considerable amounts are used e.g. as heat stabilizers. Large
quantities of metals also figure in pigments, biocides and
lubricants and in the ceramic and electrical industries.
Fertilizers and pesticides often contain metals.
Table 3 lists the estimated total emission of metals from oil
combustion during 1971 in Sweden. Combustion of fossil fuels
and production of cement give rise to a mobilization of certain
elements, notably mercury and boron, to the oceans via the
atmosphere in amounts comparable to those moving to the oceans
via rivers in the major sedimentary cycle (Goldberg, 1976) . For
51
-------
some metals, viz. arsenic, lead, selenium and zinc, the
contribution from cement production seems to be greater than
that from the burning of fossil fuels. The increase in
European coal production to parallel the increase of heavy
metals in the sediments is found to start at the time of the
Industrial Revolution.
2. Transport
The amount of material moved about the earth's surface due
to the activities of man has been estimated to be about one
tenth of that involved in the major weathering cycle (Gold-
berg, 1972). The major transport routes for metals in the
environment are summarized in Figure 1.
2.1 Atmospheric transport
Metals emitted into the atmosphere from the various sources
will be carried greater or shorter distances with the winds
depending on their state - gaseous, vapor or particulate -
before they are deposited on land or on the surface of the
oceans through dry fallout or washout with precipitations.
In the case of particulates, particle size is the most
decisive factor. The distance a metal will be carried is
also dependent on meteorological conditions such as winds
and precipitation, topography and vegetation. Investigations
indicate that most of the metals associated with particles
are deposited within 10 km from the point of emission.
The gaseous phase is of course of importance for the aerial
transport of mercury but also for that of arsenic, cadmium,
lead, antimony, selenium and zinc (Goldberg, 1974; Lovblad
and Grennfelt, 1977). An enrichment on the finer particles
has been found for the latter elements. Approximately half
of the amount of the lead emitted with automobile exhausts
is deposited within 100 m, and a third, which is emitted in
the form of very fine particles, is transported over long
distances (Folkesson, 1976). Mercury behaves differently from
other elements due to its very high vapor pressure. It has
been shown that the final deposition of about two thirds of
emitted mercury takes place within 200-2,000 km. It has also
52
-------
been shown that airborne mercury strongly interacts with
its surroundings. On the average, less than 20% is finally
deposited while the rest is re-emitted to the atmosphere
(Hogstrom and Enger, 1977; Svedung, 1977).
2.2 Aquatic transport
The greater part of the metal load emitted into the environment
is transported by water (lead is an exception). Most of it
will eventually reach the lakes and the coastal zones of the
oceans via river transport. In a lake the portion of the
total metal load carried on particles of different types
will settle in areas with active sedimentation and thus be
deposited in the sediments from which the metals may be
re-released due to microbial activity and changes in various
physical and chemical parameters, including pH and redox
potential. Very little will ever travel beyond the coastal
waters of the oceans and the metal levels decrease with an
increasing distance from a river mouth. Two mechanisms
caused by increasing salinity and microbial activity may
account for this: 1) Salting out of the large molecular
weight organic fraction - e.g. humic acids of fresh water -
and flocculation of inorganic matter, resulting in increased
particle size, will remove absorbed and incorporated metals
to the sediment (Matson, 1968). 2) A mobilization from
carrier particles by chlorine ions and chelating substances
resulting in an increasing availability to the biota. Aquatic
organisms will concentrate metals from the ambient water to
levels far exceeding the metal levels in the water and thus
to a large extent retain them within the biologically active
coastal waters.
The mechanisms of the transport of metals have been studied
in rivers and estuaries (e.g. Kharkar et al., 1968; Cranstone
and Buckley, 1972, Skei et al., 1973; Gibbs, 1973; Landner
et al., 1977; Duinker and Nolting, 1977). These studies have
clearly shown that a considerable portion of the metals in
the water is associated with suspended particles. The extent
of this association varies greatly with the metals, the
properties of the particles, and the type of water.
53
-------
2.3 Biological transport
The transport of metals via living organisms also plays a
role in the total transport. In Figure 3 the flow of methyl-
mercury and some factors influencing it in a simplified
limnic food-chain are outlined. The figure is included to
illustrate the complexity of an ecological system even when
simplified to this extent. This type of biological transport
mostly does not represent any major relocation of the metal
in space. Naturally migrating organisms transport their
content of metals, but the proportion of the total metal
content in an ecosystem that is removed or introduced this
way is usually small. For an understanding of the biological
response, however, the processes are essential.
One type of situation where "biological transport" does in-
fluence metal transport in a quantitatively important way is
in plankton in coastal areas. High standing crops of phyto-
and zooplankton may absorb a large part of the metals brought
to the coastal zone via rivers. Included in fecal pellets or
in dead organisms, the metals may then settle and become
incorporated in sediments instead of being transported
further out into the oceans.
3. Chemical form and transformation
For an understanding of the effects of metal pollution it is
important to gain a thorough knowledge of the transformation
undergone by the metals from one chemical species to another.
3.1 Physical - chemical transformation
Metals in the aquatic environment may exist in the following
forms: 1) soluble: a) free hydrated ions; b) complex and
_ ""> —
chelate ions with inorganic ligands including OH , CO-,*1 ,
Cl or organic ligands including amines, proteins, humic and
fulvic acids; c) organic molecules. 2) particulate: a)
colloidal complexes or aggregates of e.g. hydrated oxides;
b) adsorbed onto different types of particles; c) precipitated,
e.g. as metal coatings on particles; d) incorporated in
organic particles, plankton etc.; e) held in the structural
54
-------
lattice in crystalline detrital mineral particles. The
speciation is controlled by environmental variables such as
pHv redox potential - dissolved oxygen, ionic strength -
salinity, alkalinity and hardness, the presence of organic
and particulate matter, and biological activity, in addition
to the intrinsic properties of the metals.
The transformation between the different chemical form is
induced by changes in these variables. For example, the
lowering of pH (as will for instance occur when a river with
a neutral pH enters an acidified lake), will cause a release
of metals from complexes and particulate matter; a mobilization
of metals bound in sulfides in anoxic sediments will occur
upon the introduction of oxygen; metals associated with
hydrated iron oxide will be released under reducing conditions.
3.2 Redox transformations
In living organisms enzyme systems have evolved that are
capable of changing the oxidation states of chemical compounds.
In an environment rich in microbial activity each oxidation
state might be made available for metabolic interconversions.
Woolfolk and Whiteley (1962) gave several references to work
dealing with microbial reductions up to 1962. A lot of
research has been done in this field since. Only a few
examples will be mentioned here. At least fifteen strains of
bacteria have been found to oxidize arsenite to arsenate in
cattle dip solutions. Specific enzymatic activity (arsenite
dehydrogenase) was found in purified preparations (Turner
and Legge, 1954). The reduction of arsenate to arsenite
occurs in cultures of the yeast Pichia guillermondi (Bautista
and Alexander, 1972), Chlorella (Blasco et al., 1972) and
marine bacteria (Johnson, 1972) , while the oxidation of
arsenite to arsenate is brought about by Pseudomonas sp.
(Turner, 1949).
55
-------
x)
A plasmid - specified mercury reducing system will reduce
both silver (I) and gold (III) to their elemental state. The
reducing system is induced only by mercury, not by gold or
silver, however (Summers and Sugarman, 1974).
The oxidation of molecular hydrogen by extracts of Microi coccus
lacticytilus at the expense of certain chemical species of
arsenic, bismuth, selenium, tellurium, lead, thallium,
vanadium, manganese, iron, copper, molybdenum, tungsten,
osmium, ruthenium, gold, silver and uranium as well as
molecular oxygen has been reported (Woolfolk and Whiteley,
1962). Theobacillus ferro oxidans is able to oxidize ferrous
ion to ferric at pH 2 in mixed sulfide ores (Tuovinen and
Kelly, 1974) . A culture of Schizosaccarorayces, supplied with
tellurite ions, will carry out a reduction which brings the
metal to its elementary stage (Smith, 1974) .
3.3 Formation and degradation of organic compounds
Many metals and metalloids are released into the environment
in relatively non-toxic forms. They may subsequently acquire
an enhanced toxicity as organometals through environmental
interactions involving both biological and non-biological
processes. Such pathways may potentially combine, serving
then as a means for metal transport, particularly across
sediment-water-organism interfaces.
3.3.1 Methylations of metals and metalloids
That arsenic, selenium and tellurium can undergo biological
methylation has long been recognized. A comprehensive review
of the knowledge concerning biological methylation up to
1945 was made by Challenger (1945). Inorganic compounds of
of arsenic and selenium were shown to be methylated within
Scropulariopsis brevicaulis and tellurium compounds were
x) Plasmids - small ancillary circular DNA molecules carrying
extra chromosomal genes often specifying bacterial resistance
to toxic substances.
56
-------
predicted to follow the same metabolic pathway in the or-
ganisms. Some Penicillum strains were also shown to methylate
selenite and tellurite ions when these ions were present
together in bread cultures. At the time, there was no evidence
of bacterial methylation.
Very little new information on biological methylation of
metal compounds was generated until the mid-sixties. Westoo
at that time demonstrated (1966) that mercury in fish is
predominantly present in the form of methylmercury. It was
soon shown that some microorganisms, unidentified, in natural
organic lake sediment could methylate mercury (Jensen and
Jernelov, 1967). The net result of the process was mono- or
dimethylmercury and the rate of biological methylation of
mercury was found to be well correlated with general micro-
biological activity in the sediment (Jensen and Jernelov,
1969).
Methylvitamin B,,, has been shown to be capable of transferring
methyl groups to mercury (II), thallium (III), platinum (II)
and gold (I) in vitro (Agnes et al., 1971a, 1971b). Two
different pathways seem to be involved. The one, which
mercury (II) and thallium (III) follow, is apparently an
acid-base reaction. The other, which platinum and gold
follow, is probably an oxidation-reduction "redox switch"
reaction, since it requires both platinum (II), and platinum
(IV) or both gold (I) and gold (III) . Cadmium (II) , lead
(II) and indium (III) were found to be ineffective. In each
case the first step of the reaction is the production of the
methylated metal. In the second step the methyl group can
readily be transferred further to a good nucleophile. All
methylated metal compounds except methylmercury readily
decompose to metal ions and methyl groups.
Methylcobalamin has been implicated in the methylation of
mercury, lead, tin, pallacium, gold, platinum and thallium
as well as arsenic, selenium, tellurium and sulfur (Wood et
al., 1977). According to Wood and his co-workers, two
general mechanisms exist for the methylation by
57
-------
methylcobalamin: 1) reactions in which the metal or metalloid
acts as an electrophile; 2) reactions in which the metal or
metalloid acts to abstract a methyl radical. Type 1 reactions
involve a transfer of CH., to the more oxidized state of the
element, and type 2 a transfer of CH • to the reduced member
of a redox couple. It appears that those elements which
react according to the type 1 mechanism (Tl, Pd, Hg) have
redox potentials greater than + 0.8 volts, and that the more
oxidized member of the redox couple, Tl (III), Pd (II) and
Hg (II) , is the species which is methylated by methyl-B,~.
Those elements which are known to react via the type 2
mechanism, e.g. Cr (II), are found at the reducing end of
the redox potential scale. Elements of intermediate redox
potential react via the "redox switch" mechanism described
by Agnes et al. (1971a, 1971b).
3.3.1.1 Methylation of mercury
The biological cycle of mercury is shown In Figure 3. That
microbial activity is a prerequisite for the synthesis of
methylmercury under natural conditions, unless other methylated
metal compounds are present, has been shown in several
experiments. Bacteria isolated from mucous material on the
surface of fish have been shown to be able to methylate
mercury (Jernelov, 1968). Extracellular methylmercury formation
has been demonstrated for several bacteria (Kitamura et al.,
1969; Yamada and Tonomura, 1972 j Vonk and Sijpersteijn,
1973) and intracellular methylmercury formation has been
demonstrated for several fungi (Landner, 1971; Vonk and
Sijpersteijn, 1973). Methylation in vivo was studied in
aerobic cultures of Neurospora crassa. Mercury-tolerant
mutants brought about methylation very effectively when an
excess of cysteine or homocysteine was present in the substrate.
It has been suggested that the biomethylation might be an
"incorrect" synthesis of methionine - normally formed through
methylation of homocysteine (Landner, 1971).
58
-------
The potential for microbial methylation of mercury by fungi
and bacteria has been shown to exist under aerobic as well
as anaerobic conditions. The ecological significance of these
findings, especially of the anaerobic methylation, is difficult
to evaluate, however. For instance, methylcobalamin is known
to be unstable in a natural environment. The transmethylating
activity of methylcobalamin in vitro was found to be inhibited
by cellular proteins and thiol groups (Bertilsson and Neujahr,
1971). Since mercury is hardly present in nature under
anaerobic conditions without the simultaneous presence of
hydrogen sulfide, mercuric sulfide is likely to be formed.
Under these conditions mercury will be effectively prevented
from being methylated.
No formation of methylmercury was found in anaerobic mud in
some experiments by Rissanen (1974), probably due to the
existence of mercuric sulfide. It is true that the sulfide
will be oxidized to sulfate, if aerobic conditions should be
re-established, but this oxidation is probably slow. The
methylation rate was found to be 100-1000 times slower in
aerobic sediments with mercuric sulfide as the mercury donor
as compared with mercuric chloride (Fagerstrom and Jernelov,
1971). In fish allowed to accumulate mercury from sediments,
mercuric sulfide was very slowly mobilized as compared to
mercuric chloride (Gillespie, 1972).
The existence of a demethylating capacity, which was first
described in 1969 (Furukawa et al., 1969; Kitamura et al.,
1969), is a complicating factor when attempts are made to
transfer laboratory results to the ecosystem level. Most
experiments on rates of methylation have been performed in
such a way as not to allow for a distinction between the two
competing processes, methylation and demethylation. Accordingly,
it is possible that most of the results that have been
interpreted as measures of gross methylation rates have in
fact been net methylation rates. The kinetics of the response
of such a competitive system in relation to external stimuli,
such as temperature, is of course more complicated, rendering
the interpretation of data more uncertain than if methylation
alone is presupposed to occur.
59
-------
Bisogni and Lawrence (1973) studied the net methylation with
regard to the effects of redox potential, inorganic mercury
concentration, temperature, microbial activity and sulfide
concentration. A proposed kinetic model showed: 1) Monomethyl-
mercury is the predominant product of methylation of mercury
(near neutral pH); 2) The rate of methylation is higher in
aerobic than in anaerobic systems for a given inorganic
mercury concentration and microbial growth rate; 3) Higher
microbial growth produces higher methylation rates under
aerobic as well as anaerobic conditions; 4) Methylation
rates can be reduced by the addition of sulfide to some
anaerobic systems; 5) Temperature affects methylation rates
in accordance with its effects on the metabolic rate of the
methylating organism.
3.3.1.2 Methylation of other metals and metalloids
The methylation of the three metalloids has been proposed to
follow the same metabolic pathways (Challenger, 1945;
McBride and Wolfe, 1971; Wood, 1974).
Many investigators have reported the microbial formation of
methylated species of arsenic, selenium and tellurium. The
production of dimethylselenide from inorganic selenium
compounds by several types of fungi has been shown (Challenger,
1945; Fleming and Alexander, 1972; Burlees and Flemming,
1974). Anionic forms of selenium and tellurium are subject
to microbial alkylation, the products being dimethylselenide
and dimethyltelluride (Burlees and Flemming, 1974) .
A coryneform bacterium has been reported to convert inorganic
selenium to Me~Se~ (Francis et al., 1974). Three bacterial
isolates from lake sediment were able to convert sodium
selenite to Me~Se, Me-Se,, and an unknown volatile selenium
compound during their growth (Chau et al., 1976). Some fungi
isolated from industrial and agricultural arsenic-containing
sludge, e.g. Candida humicola, are capable of converting
salts of arsenic, arsenous monomethylarsenic or dimethyl-
arsenic acids to trimethylarsine (Cox and Alexander, 1973a,
1973b).
60
-------
Selenium and tellurium as well as arsenic are readily methylated
by the methanogenic bacteria Methanobacterium, strain M.O.H.,
under anaerobic conditions. The inorganic arsenic forms were
reduced and gradually methylated to dimethylarsire. Dimethylarsinic
acid and methylarsonic acid '..ere intermediates (McBride and
Wolfe. 1971). Fresh water, seawater, eggshells and rock
samples were analyzed for arsenic (V) , arsenic(III) , methyl-
arsonic acid and dimethylarsonic acid. The methylated compounds
were found in most biological samples but not in limestone
(Braman and Foreback, 1973).
Evidence has been put forth that microorganisms in lake
sediments can transform certain inorganic and organic lead
compounds into the volatile tetramethyllead (Me.Pb) . Me-.Pb
was converted to Me.Pb by bacterial isolates from Lake
Ontario. Species of Pseudomonas, Alcaligeneus, Actinetobacter,
Flavobacterium and Aeromonas could transform Me-,PbOAc into
Me.Pb. None of the isolates could produce Ne.Pb from inorganic
lead (Wong et al., 1975). Tin has also been demonstrated to
undergo biological methylation (Huey et al., 1974).
3.3.1.3 Phgtoalkylation
In the presence of acetate ion or acetic acid in aqueous
solutions, mercury(II) may be photoalkylated if exposed to
sunlight (Agaki and Takabatake, 1973). Photolysis of mercury
chloride in the presence of sodium acetate under normal
laboratory fluorescent lighting has been found to produce
methylmercury ion, dimethylmercury and metallic mercury
(Jewett et al., 1975). A methyl-chromium bond has been found
to be formed in the long-lived methylpentaaquochromium(III)
ion during the photolysis of tertiary-butoxy radicals and
chromium(II) in aqueous solution (Ardon et al., 1971).
61
-------
3.3.1.4 Transalkylation
Abiotic transalkylation between several metals in aqueous
solution has been demonstrated (Jewett and Brinkman, 1974) and
is shown in Table 4. Example of reactions:
(CH)3Sn+ + Hg2+ -*(CH3)2Sn2+ + CH3Hg+
(CH0)0Sn+ + PtCl 2~ -» {CH.,PtCl "} -» C0H, + Pt° + (CH.J0Sn2+
JO 4 33 ZD J £
Trimethyltin and trimethyllead seem to display the same
transmethylation chemistry. Trimethyllead is a far more
rapid methylator of mercury(II), though, being comparable to
methylcobalamin in its ability to transfer CH3 to mercury.
The methylation of the noble metals palladium(II), gold(III)
and platinum(II) proceeds readily but the reaction pathway
is quite different. Methylmercury can also act as an effective
methyl donor for the noble metals although the reaction rate
is considerably slower.
A species of Pseudomonas which produces a volatile methylated
tin species from tin(IV) has been isolated from the Chesapeake
Bay (Huey et al., 1974). The same organism is also quite active
in forming elemental mercury from phenylmercuric acetate and
mercuric ion. But as was shown, in the presence of both
mercury(II) and tin(IV), methylmercury is formed by this
organism through a mixed biotic and abiotic process. An
abiotic transalkylation process between alkyllead compounds
and inorganic mercury has been demonstrated in the St. Clair
River and in the laboratory, in sterilized sediments (Jernelov
et al., 1972).
In experiments investigating methylation of selenium by sewage
fungi some Penicillum strains were isolated from the sewage.
When tellurium compounds were added (TeCl., HnTeO~ and H,TeO,.},
4 2 2 o o
they were found to be methylated as effectively as selenium
compounds, but only in the presence of selenium. The yield of
dimethyltelluride was proportional to the input of inorganic
selenium and methylated tellurium was found until the input
ratio of selenium and tellurium was about 10:1. This indicates
62
-------
a possible transalkylation, and not a direct biological alkyla-
tion, of tellurium (Fleming and Alexander, 1972).
3.3.2 Demethylation of mercury
The biological stability of organometallic compounds varies
widely. Some of them decompose rapidly through non-enzymatic
processes or through the action of non-specific enzymes. Others
are highly persistent and are degraded only through specific
biochemical processes. Short-chain alkylmercury compounds
belong to the latter group.
Decomposition of organomercurials by bacteria, resulting in the
formation of elemental mercury, has been reported by several
investigators. This process was first observed in Japan when
bacteria isolated from soil heavily contaminated with organo-
mercurxais were found to be capable of converting phenylmercury
and methylmercury into metallic mercury and benzene and methane
respectively (Tonomura et al., 1968). A large number of bacterial
strains, and also yeasts, resistant to and capable of hydrolyzing
and reducing mercury compounds have been found since, some of which
have been investigated more thoroughly (e.g. Spangler et al.,
1973; Drunker and Bott, 1974; Billen et al., 1974; Blair
et al., 1974; Colwell and Nelson, 1975).
Two distinct groups of organisms have been observed: those resistant
to organomercurials as well as mercury(II), and those resistant
to mercuric ion only. The cleavage of the mercury-carbon
bond in organic mercury compounds and the reduction of mercury(II)
have invariably been found to be catalyzed by enzyme systems -
either inducible or constitutive - one or more to catalyze the
cleavage of the mercury-carbon bond of different organic
compounds and one to catalyze the reduction of the mercuric
ion formed in the first reaction or encountered in the environ-
ment. In some cases the work on purifying and characterizing
the different enzymes has progressed very far (Furukawa and
Tonoruura, 1972; Izaki et al., 1974; Tezuka and Tonomura, 1976;
Schottel and Silver, 1976). In one case, two different hydrolases
with different physical properties have been isolated from
bacteria. The differences in their enzymatic properties are
63
-------
not yet fully understood, however {Tezuka and Tonomura,
1976). In another case at least two hydroLases have been
found and purified in a mutant of Escherichia coli, one
active against methylmercury, the other against phenylmercury
(Schottel and Silver, 1976). When the genetic background for
the resistance has been worked out, it has been found to be
associated with plasmids (Summers and Lewis, 1973; Schottel
et al., 1974j Silver et al., 1975).
3.3.3 Combined effects of alkylation and dealkylation
It has been suggested that an equilibrium may be reached
between the production (addition) of methylmercury and its
degradation in mercury polluted environments. Methylmercury
degrading activity by bacteria in sediments from the Baltic
Sea has been demonstrated, as later described by Jernelov et
al. (1975). In the light of these findings an earlier study
on methylation of mercury in the St. Clair system on the
Canada/USA border was re-evaluated. The results were demonstrated
to support the idea of an equilibrium between chemically
alkylated (from methyl- and ethyllead) mercury and biological
degradation (Jernelov et al. , 1972).
The hypothesis can be formulated that in a sediment, under a
constant flow of methylmercury into and/or out of the system,
biological formation and degradation of methylmercury will
result in an equilibrium, with a constant level of methyl-
mercury in the sediment. This is a result of the fact that the
formation is concentration-dependent with regard to inorganic
mercury and the degradation is concentration-dependent with
regard to methylmercury. If input or output of methylmercury
is varying, a disrupted pattern will result, where the
methylmercury level in the sediment tends to return to the
equilibrium as soon as the disburbance has come to an end.
The curve in Figure 4a illustrates such a situation. It is
obvious that as long as no methylmercury is added, the
importance of the demethylating activity will depend upon
the relation between methylating activity and the transport
of methylmercury out of the sediment - i.e. whether or not
64
-------
this relation is such that the equilibrium is approached.
Transport of methylmercury out of the system can be so great
as to prevent methylmercury from building up to concentrations
where demethylation becomes important. Such a situation is
depicted in Figure 4b. Here, the difference between gross
and net methylation rates is of little relevance.
The transport of methylmercury out of the system may be low,
whereupon the demethylating activity assumes importance for
the regulation of the methylmercury concentration in the
sediment (see Figure 4c). In this case the difference between
gross and net methylation rates is large.
In short, the concentration of methylmercury in the sediment
is low when the amount transported out of it is large and
vice versa. Thus, in this example, a negative correlation
exists between methylmercury concentration in the system, in
this case, sediments, and the amount of methylated mercury
released from the system through biological demethylation.
-------
REFERENCES
Agaki, H. and Takabatake, E. (1973). Chemosphere 1., 131-133.
Agnes, G., Bendle, S., Hill, H.A.O., Williams, F.R. and
Williams, R.J.P. (1971a). Chem. Commun. pp 850-851.
Agnes, G., Hill, H.A.O., Pratt, J.M., Ridsdale, S.C.R.,
Kennedy, F.S. and Williams, R.J.P. (1971b). Biochem. Biophys.
Acta 252, 207.
Andersson, G. and Grennfelt, P. (1973). "Determination of heavy
metals in fuel oil and an estimation of emission of heavy
metals from oil combustion." Swedish Water and Air Pollution
Research Institute, Stockholm (In Swedish, abstract in
English).
Ardon, M., Woolmington, K. and Pernick, A. (1971). Inorg.
Chem. ]JO, 2812.
Bautista, E.M. and Alexander, M. (1972). Soil Sci. Soc.
Amer. Proc. 36, 918-920.
Bertilsson, L. and Neujahr, H.Y. (1971). Biochemistry 10,
2805-2808.
Billen, G., Joiris, C. and Wollast, R. (1974). Water Res. J3,
219-225.
Bisogni, J.J. and Lawrence, A.W. (1973). Cornell Univ. Tech.
Rep. No. 63. Ithaca, N.Y.
Blair, W., Iverson, W.P. and Brinkman, F.E. (1974). Chemosphere
3, 167-174.
Blasco, F., Robert, J.J. and Gaudin, C. (1972). C.R. Acad.
Sci. Ser. D. 275, pp 1223-1226.
Braman, R.S. and Foreback, C.C. (1973). Science 182, 1247-
1249.
Brunker, R.L. and Bott, T.L. (1974). Appl. Microbiol. 27,
870-873.
Bryan, G.W. (1977). In: "Marine Pollution." (R. Johnston,
ed) pp 185-302. Academic Press, London.
Burlees, L. and Flemming, R.W. (1974). Bull. Environ. Contain.
Toxicol. 12, 308.
Challenger, F. (1945). Chem. Rev. 36, 315-364.
Chau, Y.K., Wong, P.T.S., Silverberg, B.A., Luxon, P.L. and
Bengert, G.A. (1976). Paper presented at "Microbial Transformation
of Heavy Metals Workshop." May, 1976. National Bureau of
Standards, Washington, D.C.
66
-------
Colwell, R.R. and Nelson, J.O. (1975). Microbiol. Ecol. 2,
191-218.
Cox, D.P. and Alexander, M. (1973a). Bull. Environ. Contam.
Toxicol. 9_, 84-88.
Cox, D.P. and Alexander, M. (1973b). Appl. Microbiol. 25,
408-413.
Cranstone, R.E. and Buckley, D.E. (1972). Environ. Sci.
Technol. 6_, 274-278.
Duinker, J.C. and Nolting, R.F. (1977). Mar. Pollut. Bull.
8_, 65-71.
Fayerstrom, T. and Jernelov, A. (1971). Water Res. 5^, 121-
122.
Fagerstrom, T. and Jernelov, A. (1972). Water Res. 6_, 1193-
1202.
Fleming, J.F. and Alexander, L.T. (1972). Anpl. Microbiol.
2_4, 424-429.
Folkeson, L. (1976). "Lead, Especially Exhaust Lead, in the
Terrestrial Environment: Accumulation and Ecological Effects,
a Literature Survey." 103 pp. Report No. 794, National
Swedish Environmental Protection Board, Solna (In Swedish
with English summary).
Francis, A.J., Duxburg, J.M. and Alexander, M. (1974). Appl.
Microbiol. 28, 248.
Furukawa, K. and Tonomura, K. (1972). Agric. Biol. Chem.
36_, 217-226.
Furukawa, K., Suzuki, T. and Tonomura, K. (1969). Agric.
Biol. Chem. J33, 128-130.
Gibbs, R.J. (1973). Science 180, 6-7.
Gillespie, D.C. (1972). J. Fish Res. Board Can. 29_, 1035-
1041.
Goldberg, E.D. (1972). "The Changing Chemistry of the Oceans."
pp 267-288. Wiley-Interscience, New York.
Goldberg, E.D. (chairman). (1974). In: "Ecological Toxicology
Research." (A.D. Mclntyre and C.F. Mills, eds) pp 233-285.
Plenum Press, New York.
Goldberg, E.D. (1976). "Health of the Oceans." UNESCO Press.
Huey, C., Brinkman, F.E., Grim, S. and Iverson, W.P. (1974).
In: "Proceedings of the International Conference on Transport
of Persistant Chemicals in Aquatic Ecosystems." pp 11:73-78,
May 1-3, 1974. Ottawa.
67
-------
Hogstrom, U. and Enger, L. (1977). Unpublished data.
Izaki, K., Tashiro, Y. and Funaba, T. (1974). J. Biochem.
Tokyo 75>, 591-599.
Jensen, S. and Jernelov, A. (1967). Nord. Forsk. Bioci. Inf.
10, 4 (In Swedish).
Jensen, S. and Jernelov, A. (1969). Nature 223, 753-754.
Jernelov, A. (1968). Vatten 24, 360 (In Swedish).
Jernelov, A., Landner, L. and Larsson, T. (1975). J. Water
Pollut. Control Fed. 47, 810-822.
Jernelov, A., Lann, H., Wennergren, G., Fagerstrom, T.,
Asell, B. and Andersson, R. (1972). Analysis of methyl-
mercury concentrations in sediment from the St. Clair River
system. Swedish Water and Air Pollution Research Institute,
Stockholm (In Swedish with abstract in English). Unpublished
report.
Jewett, K.L. and Brinkman, F.E. (1974). Preprints of papers.
Div. Environ. Chem., Amer. Chem. Soc. 1_4, 218-225.
Jewett, K.L., Brinkman, F.E. and Bellama, J.H. (1975). In:
"Symposium on Marine Chemistry in the Coastal Environment."
(T. Church, ed) No. 21. American Chemical Society, Washington,
D.C.
Johnson, D.L. (1972). Nature 240, 44-45.
Khakar, D.P., Turekian, K.K. and Bertine, K.K. (1968).
Geochim. Cosmochim. Acta 32, 285-298.
Kitamura, S., Sumino, K. and Taina, M. (1969). Jap. J. Hyg.
2A, 76-77.
Landner, L. (1971). Nature 230, 452-454.
Landner, L., Grahn, O. and Sjoo, L. (1977). Metal transport
in three Swedish rivers. Swedish Water and Air Pollution
Research Institute, Stockholm (To be published).
Lovblad, G. and Grennfelt, P. (1977). Heavy metals and other
trace elements in coal and their atmospheric emission from
the combustion of coal - a literature survey. Swedish Water
and Air Pollution Research Institute, Stockholm (In Swedish).
Matson, W.R. (1968). "Trace Metals, Equilibrium and Kinetics
of Trace Metal Complexes in Natural Media." Diss., Massachusetts
Institute of Technology, Cambridge.
McBride, B.C. and Wolfe, R.S. (1971) . Biochemistry 10, 4312-
4319.
68
-------
"Minerals Yearbook." (1971). U.S. Bureau of Mines, Washington,
D.C.
Rissanen, K. (1974). Personal communication.
Schottel, J., Mandal, A., Toth, K. , Clark, D. and Silver, S.
(1974). In: "Proceedings of the International Conference on
Transport of Persistant Chemicals in Aquatic Ecosystems." pp
11:65-71, May 1-3, 1974, Ottawa.
Schottel, J. and Silver, S. (1976). Personal communication.
Silver, S., Schottel, J. and Weiss, A. (1975). In: Proceedings
of the 3rd International Biodegradation Symposium." (J.M.
Scharpley, ed) Aug. 1975. Kingston, R.I.
Skei, J.M., Price, N.B. and Calvert, S.E. (1973). Ambio 2L,
122-124.
Smith, D.G. (1974). J. Gen. Microbiol. 83, 389-392.
Spangler, W.J., Spiragelli, J.L., Rose, J.M., Plippin, R.R.
and Miller, H.H. (1973). Appl. Microbiol. 25, 488-493.
Summers, A.O. and Lewis, E. (1973). J. Bacteriol. 113, 1070-
1072.
Summers, A.O. and Sugarman, L.I. (1974). J. Bacteriol. 119,
242-249.
Svedung, I. (1977). Unpublished data.
Tezuka, T. and Tonomura, K. (1976). J. Biochem. 80, 79-87.
Tonomura, K., Nakagami, T., Futai, P., Maeda, K. and Yamada,
M. (1968). Nature 217, 644-646.
Tuovinen, O.K. and Kelly, D.P. (1974). Int. Metall. Rev. 19,
21.
Turner, A.W. (1949). Nature 164, 76-77.
Turner, A.W. and Legge, J.W. (1954). Aust. J. Biol. Sci. _7,
452.
Vonk, J.W. and Kaars Sijpersteijn, A. (1973). Antonie van
Leeuwenhoek 39, 505-513.
WestSo, G. (1966). Acta Chem. Scand. 20, 2131-2137.
Wong, P.T.S., Chau, Y.K. and Luxon, P.L. (1975). Nature 253,
263-264.
Wood, J.M. (1974). Science 183, 1049-1054.
Wood, J.M., Kennedy, S. and Rosen, G.G. (1968). Nature 2 2_0,
173-174.
69
-------
Wood, J.M., Ridley, W.P., Dizikes, L.J., Cheh, A.M. and
Frick, T.D. (1977). Paper presented at the International
Symposium on "Factors Influencing Susceptibility to Metal
Toxicity." Scientific Committee on the Toxicology of Metals,
Stockholm, July 17-22, 1977.
Woolfolk, C.A. and Whiteley, H.R. (1962). J. Bacteriol. 84,
647-658,
Yamada, M. and Tonomura, K. (1972). J. Ferment. Technol. 50,
159-166.
70
-------
Table 1. World production (metric tons) of metals (From
Minerals
Metal
Ag
Al
As
Au
B
Ba
Be
Bi
Ca
Cd
Ce
Co
Cr
Cs
Cu
Dy
Er
Eu
Fe
Ga
Gd
Ge
Hf
Hg
Ho
In
Ir
K
La
Li
1965
Production
8
6 300
59
1
3 500
5
3
12
7
18
4 800
5 000
5 500
620 000
9
000
000
000
400
000
600
000
000
200
000
000
000
000
000
200
Yearbook,
1969
Production
9
9 000
52
1
700
3 800
7
3
560
17
11
20
5 100
6 000
6 600
720 000
1
3
0
9
8
000
000
000
400
000
000
200
800
000
000
700
000
000
242
000
000
000
.21
.56
.2
900
.81
1971) .
Increase
12
44
-11
0
7
29
29
42
62
9
6
19
19
16
6
.1 %
.2 %
.4 %
%
.3 %
.5 %
.7 %
.9 %
.7 %
.8 %
.2 %
.3 %
.3 %
.4 %
.6 %
Comment
Al
Irregular
Irregular
Estimated
BaSO4
Per annum
production
production
as B-O,
USA
Monazite, CePO. incl
lanthanides
Chromite
Import USA
Ore
Metal production
Fe in prod
Import USA
Import USA
Import USA
Import USA
. ore
See platinum group
14 000
8
000
600
17 000
000
21
.2 %
Argentina,
Australia
Brazil and South
Africa only
71
-------
Table 1. (Cont'd).
Metal
Lu
Mg
Mn
1965
Production
160
17 000
000
000
1969
Production
200
18 000
000
000
Increase
24
5
.1 %
.9 %
Comment
(1967)
Mo
Na
Nb
Nd
Ni
Os
Pb
Pd
Pr
Pt
Rb
Re
Rh
Ru
Sb
Sc
Se
Sm
Sn
Sr
Ta
Tb
Te
Th
45
120 000
430
2 700
2 600
63
200
200
10
6
000
000
000
000
000
92
000
820
000
000
000
600
150
110
65
130 000
480
3 200
3 300
6
65
1
230
230
27
13
000
000
000
000
000
105
.95
9.1
000
250
000
000
000
300
190
230
44
12
13
19
26
13
3
53
11
14
170
100
32
110
.9 %
.6 %
.2 %
.0 %
.0 %
.8 %
.8 %
.0 %
• J "6
.0 %
.0 %
.0 %
.9 %
.0 %
USSR, Eastern Europe,
China excluded
As Nad
See Ta
Ore
Metal production
Platinum group
Consumption per year
est. 1972 USA
See Platinum group
Irregular increase
Ore
Metal production
As SrSO., SrCO3 etc.
West Germany, Poland,
Spain and USSR excluded.
Increased demand expected
Including Nb
Import USA in 3 800 tons
monazite as ThO,
72
-------
Table 1. (Cont'd).
Metal Production Production Increase
Comment
Ti
Tl
Tm
U
V
W
Y
Yb
Zn
Zr
2 500 000
220 000
17 000
8 900
27 000
4 300 000
290 000
3 200 000
380 000
0.27
19 000
10 300
33 000
5 300 000
380 000
30 .6
69.4
12.0 %
15.5
20.7
23.7
29.0
Ilmenite
Rutile
Import USA, unprocessed
Czechoslovakia, East
and West Germany,
Hungary, India, and
USSR excluded
Scheelite
USSR, Eastern Europe
and China excluded
73
-------
Table 2. Uses of metals (From Bryan, 1977).
Uses of metals and compounds with importance
Metal generally decreasing from left to right.
Percentages are for USA
Ag Photography; electric conductors; sterling ware;
solders; coinage; electroplating; catalysts; batteries
Al Building and construction; transportation; electrical
industry; packaging; consumer durables; machinery;
glass; pharmaceuticals.
As Pb and Ca arsenate and sodium arsenite pesticides;
glass and enamels; alloys; electronics.
Cd Electroplating (50 %); pigments (22 %); thermoplastic
stabilizers, e.g. in PVC (22 %); batteries; low
melting point alloys.
Co Alloys mainly; catalysts; pigments; enamels; glazes;
electroplating.
Cr Metallurgy-alloys (64 %); refractory bricks (21 %);
electroplating; tanning; paint; wood preservatives.
Cu Electrical industry; alloys, e.g. brass; chemical
catalysts; antifouling paint; algicides; wood
preservatives.
Fe Iron and steel industry.
Hg Chlor-alkali production; electrical apparatus;
anti-mildew paint; instruments; catalyst; pesticides;
Pharmaceuticals; antifouling paint.
a) Information taken from Minerals Yearbook (1971)
74
-------
Table 2. (Cont'd).
Uses of metals and compounds with importance
Metal generally decreasing from left to right.
Percentages are for USA
Mn Metallurgy-largely steel alloys; dry batteries;
chemical industry, e.g. permanganate; glass;
ceramic coloring.
Mo Metallurgy-steel alloys; catalysts; pigments;
glass; lubricant and oil additive.
Ni Metallurgy-steel and other alloys; electroplating;
catalysts.
Pb Storage batteries (42 %); leaded gasoline (19.5 %);
pigments, e.g. red lead paint; ammunition; solders;
cable covering; anti-fouling paint.
Sb Antimonial lead (hardened lead); plastics; ceramics
and glass; flame-proof chemicals; bearing metal;
pigments; type metal.
Se Glass industry; electrical industry, e.g. rectifiers,
photocells; paints; rubber; insecticide-sodium
selenite.
Sn Tinplate; solders; bronze; white metal; chemical
reducing agent; fungicide-triphenyl tin acetate;
anti-fouling paint.
V Metallurgy-steel and other alloys; catalyst; pigments.
Zn Zn-based alloys; brass and bronze; galvanizing;
rolled Zn; paints; batteries; rubber.
a) Information taken from Minerals Yearbook (1971).
75
-------
Table 3. Estimated total emission of metals (metric tons/
year) from oil combustion (from approximately
30 x 10 tons) in Sweden 1971 (From Andersson and
Grennfelt, 1973).
As
Cd
Co
Cr
Cu
Fe
Hg
In
Mn
Mo
Ni
Pb
Sb
Sc
Se
V
Zn
0
0
4
0
3
67
0
0
1
2
180
10
0
0
2
580
10
.7
.2
.4
.82
.9
.075
.001
.4
.075
.007
.5
76
-------
4-1
4->
Q)
IS
Q)
t^
e
o
n
ED
—
^
(0
in
G
o
•H
-U
3
•-I
o
in
to
3
o
0)
p
D1
rd
C
-H
13
0)
-U
a
g
(D
4J
-U
03
in
G
o
•H
4J
O
tO •
Q) -^
S-l '*
r^
G iS
.£ .x
4-> G
0) -H
g n
in «
G
rd 15
H G
E-i rt
•
^*
a)
M
X!
fO
EH
CM
iH
U
'O
U
(N
^H
U
G
W
OO
(D
s
1-1
o
r-H
EH
CM
QJ
^-i
^^
T3
CM
rH
O
-P
CM
^p
^H
U
4-1
Cu
fN
W
•^
,-H
O
'd
PJ
CM
n3
Z
»*
T— 1
O
3
<<
(0
2
CM
,-H
U
tn
ffi
SH
O
4-1
d
rH
>i
^
4-1
QJ
s
o
o o
o
o
+ + +
+ + + + 0
+ + +
-•— ^ , v s
ft £> 0
+ 4- 000
+
+ + + +
G A + iH G 4-
w ft tn EH cn co
ro n ;u CM CM n
0) 0) CU (U 0) 0)
S S S S 2 S
CD M
U X
G C
rd O 0
G -H 0
O -P iH
Ul U
0) rd rd
K 0)
M M
O 0)
•H 0) >
4J M O
0) 0)
C .G T!
tP S Q)
rd >
E in M
0) 0)
H 05 Ul
rd rd /2
CD o o
rH
O CU Cfl
3 4-i rd
G rd 5
U
— -H C
M T! o
0) C -H
4J -H 4-1
Q) O
g W rd
rd G 0)
-H tn M
15 -H
in O
S-i C
(D en
4-1 D 0)
3 -H VM
O ft 0)
x^^. (— *
• s
6 -P
E C tn
rd CD
in 4-1 w
U rd
C rd 0
-H Q)
>-i 0
4-> 4-1
3 £ rd
O O U
rd -H
T3 Q) 13
(U G
-H M -H
M O
M UH U)
rd O
U S M
0
df-J N
S-H
CD 0 13
S -H G
2 1
XJ
13
QJ
13
Jn
rd
4-1
CD
^
C
O
-H
4-1
03
i— 1
>1
-C
4-1
OJ
E
en
G
rd
M
4-1
m
o
0)
4-1
rd
Di
.
£!
•
0
O
o
o
rH
o
4->
a.
d
G
0
-r-\
4-1
0
rd
Q)
M
0
2
„ .
O
.
J-l
Q)
4-1
rd
?
13
QJ
in
en
rd
tn
0)
13
G
•H
T3
0)
4->
O
3
13
G
O
U
,
13
77
-------
r
METAL
EMISSION
i
ri\ i nuir
,1
F
A
tJ
nt
w
A
S
'H
TERRESTRIAL RUN-OFF
SYSTEMS
LAKES
IRRIGATION
\0
FLOW
ESTUARIES MIXING ^ OCEANS
SEDIMENTS
SEDIMENTS
Figure 1. Routes for trace elements in the environment.
78
-------
Figure 2. Mercury transport in a simplified limnic food chain.
A. Composition of sedi- G.
ment H.
B. Temperature I.
C. pH K.
D. Concentration of
oxygen L.
E. Sediment M.
F. Sediment feeding
bottom organisms
Methylmercury in water
Suspended organic material
Predatory bottom organisms
Detritus feeding bottom
organisms
Roach
Pike
Each "box" represents the amount of methylmercury
present in a trophic level at a given time. Boxes
surrounded with hexagons indicate living organism
populations.
The broad arrows represent flow of mercury and the
narrow ones flow of information. The sources of
information can be either "outside the system",
indicated by circles, which means that they have an
influence on the system without being influenced by
it, or "inside the system", indicated by roofed
circles, which means that they affect and are affected
by the system itself; feedback mechanisms.
79
-------
Figure 3. The mercury cycle.
80
-------
a)
input of CH^Hg
b)
c)
output of CH-.Hg
Theoretical equilibr.
between demethylating
and methylating activity
Figure 4. The combined effect of input and output, methylation
and demethylation of mercury.
81
-------
EFFECTS - GENERAL PRINCIPLES
UNDERLYING THE TOXIC ACTION OF METALS
Thomas W. Clarkson
Metals produce a wide range of effects in man resulting
from their action at molecular, cellular, tissue and organ
levels. Depending upon the particular metal involved,
the metal action may manifest itself as a local effect
such as on the skin, pulmonary membranes or gastrointestinal
tract or it may manifest itself as a systemic effect that
could potentially involve any tissue or organ in the body.
Furthermore, metals may act as allergens, mutagens, teratogens
or carcinogens. No group or class of effects is unique to
the metals. Indeed it is the diversity of effects that, if
anything, is characteristic of the metals.
A comprehensive, all-inclusive description of the biological
effects of metals would amount to a textbook of general
pathology. Instead this chapter will attempt to discuss the
general principles underlying the toxic action of metals.
Effects associated with a particular metal are described in
the chapters of this book dealing with specific metals.
1. Metals as elements
1.1 Cumulative action of metals
Unlike the majority of organic chemicals that can be
eliminated from tissues by metabolic degradation the metals as
elements are indestructable and therefore have the potential
for accumulation in the body leading to chronic effects. The
only way metals can be eliminated from the body is by
excretion. The rate and pathways of excretion vary greatly from
one metal to another. For example, the biological half-time of
methylmercury in man is about seventy days versus an estimated
half-time of ten to twenty years for cadmium. The biological
half-time for one and the same metal may also vary between
different tissues. The biological half-time of lead in tissues
83
-------
is in the order of a few weeks versus an estimated biological
half-time of about ten years in bone.
Accumulation in tissues does not necessarily imply the advent
of toxic effects. In the case of certain metals inactive
complexes or storage depots are formed. Lead is stored in
bone normally in an inert form. Only on the release from
bone that may occur under certain physiological conditions
will lead poisoning result. As will be discussed in detail
elsewhere in this book cadmium and other metals bind to a
small molecular weight protein, metallothionein, that appears
to form inert complexes. Only when such complexes reach a
critical concentration in the cell do toxic effects appear.
Inorganic mercury, cadmium and other metals may form inert
complexes with selenium compounds. The formation of such
complexes results usually in the retention of the metal in
body tissues. The long-term sequelae of storage of these
complexes over a human lifespan is not yet fully understood.
It is clear from animal experiments that the formation of
these complexes can prevent the acute and short-term toxic
effects of these metals. However, it is not clear to what
extent storage of these complexes over long periods of time
is without hazard. For example, cadmium is believed
to accumulate in human kidney tissue over virtually the entire
lifespan of the individual. In members of the normal popula-
tion such an accumulation of cadmium can attain values of up
to 50 mg/kg. The estimated critical concentration for effects
is around 200 mg/kg. This long-term accumulation of cadmium
is a matter of concern given that levels in later life are
approaching those associated with tissue damage.
1.2 Metals as essential and non-essential trace elements
The presence of many of the metals in the earth's crust can
no doubt be interpreted as showing that they are essential
for life processes. Metals are usually essential because they
are an integral part of at least one enzyme. Since the primary
action of an essential metal is as catalysts only trace amounts
of the metals are necessary for optimal cellular function.
84
-------
The dose-response relationships for the essential metals
therefore must take into account the fact that at very
low intakes of the metal biological effects may appear due
to deficiencies whereas at high intakes effects may be due to
an overdosage.
The consequences of deficiency states of essential trace
metals are described in great detail in a variety of nutritional
textbooks. In general, deficiency results in retarded growth
along with specific effects associated with each metal. In
certain cases the deficiency state may not be associated
with reduced amounts of metal in the diet. Zinc deficiency may
be caused by reduced reabsorption of the metal from the diet
and also by elimination of the metal in perspiration.
Deficiency states may be produced by competitive action of
other metals as shall be discussed below. Thus the fact that
certain trace elements are essential to life leads to the
possibility of interaction with other metals in producing
biological effects.
The body has developed a variety of homeostatic mechanisms
with regard to the essential trace metals. These mechanisms
usually take the form of a controlled absorption from the
diet allowing enhanced absorption with low dietary intakes
and diminished absorption with high dietary intakes. One of the
most studied is that of iron whereby at low intakes or in
the face of major biochemical demands for this element, ab-
sorption across the intestinal barrier is increased. When
body stores are adequate, iron is stored in the mucosal cells
as ferritin and is eliminated from the body when the
intestinal cells exfoliate. Specialized carrier proteins
have been identified for the essential trace metals such as
transferrin which carries in plasma such metals as iron, cobalt
and zinc. \ calcium transport protein has been identified in
kidney tissue and in mitochondria! membranes. Non-essential
metals may participate in some of these specialized processes,
resulting in toxic effects directly or by interaction with
the essential trace metals.
85
-------
The essential trace metals are less likely to produce toxic
effects because of the presence of homeostatic mechanisms
that control the body burden. Nevertheless there are a
variety of ways in which the control mechanisms have been
overwhelmed or circumvented. The actions of metals on skin or
pulmonary membranes are independent of homeostatic controls. For
example when the trace metal is present as a corrosive or
acid forming salt, effects on skin and the mucosal membranes
may appear. Although chromium is an essential trace element,
chromate salts can produce severe effects in both the skin,
nasal septum and in the lung. The homeostatic mechanisms
usually operate to control gastrointestinal absorption so that
entry into the body by other pathways circumvents this control.
For example, the inhalation of zinc oxicle fumes can give rise
to an allergic response resulting in metal fume fever. In-
halation of manganese oxides or salts can result in chronic
toxicity. The homeostatic control can be: overwhelmed by
very high doses of the essential metal such as is the case
with iron salts. Above a certain dosage, destruction of the
gastrointestinal tract is so severe that the iron enters the
bloodstream in toxic quantities.
The toxic effects may take forms which are not clearly unique
for the essential metals except in the case where the action
of the enzyme involving the metal is enhanced. For example
excessive doses of molybdenum may increase the activity of the
molybdenum-requiring enzyme xanthine oxidase resulting in the
increased production of uric acid. As discussed in the chapter
on molybdenum, this may account for reports of gout-like
symptoms in people having excessive exposure to this metal.
Some of the toxic effects of the non-essential trace metals
may be a consequence of a blockage of the availability or
activity of essential metals. Xanthine oxidase is inhibited
by tungstate salts due to the displacement of molybdenum.
Lead blocks the utilization of iron in heme synthesis by
inhibition of the enzyme ferro chelatase. There is evidence
86
-------
that cadmium can block the entry of zinc into the fetus thereby
causing a variety of teratogenlc effects.
The effect ,:f K ad in mitochondria may be a direct consequence
of its carriage into the sub-cellular particle by the cadmium
transport system. In fact the calcium carrier is believed to
be relative^' non-specific for calcium and capable of trans-
porting a broad spectrum of divalent cations.
Competition effects probably occur between essential metals them-
selves. The toxicity of copper is attenuated by that of molybdenum
and vice versa. Other examples of competitive interaction
between specific metals can be found in the chapters on specific
metals and in the chapter on factors affecting dose-effects and
dose-response relationships.
2. Oxidation-reduction reactions of metals
All metals can undergo conversion from their atomic
metallic form to a corresponding cation by the loss of
one of more electrons. The opposite process (reduction)
may also occur in tissues in a limited number of cases. Each
metal has its characteristic oxidation-reduction or redox potential,
The toxicity of each metal depends both quantitatively and
qualitatively on its oxidation state. These oxidation
changes may be brought about either by chemical action,
depending upon the local redox potential in the cell or bio-
logical fluid, or by enzymic action. In general the removal
or addition of electrons to the metal atom will influence
the chemical activity and therefore the potential of the metal
to interact with tissue ligandt as will be discussed in
section^ be Low.
Examples where the toxicity of certain metals is greatly
influenced by the oxidation state are given in the relevant
metals chapters but one illustration, mercury, will be given
here as well. Mercury may exist in three oxidation states as
illustrated in the general equation:
-------
Hg° 2 Hg2++ * Hg++
Metallic mercury (Hg ) may be oxidized to mercurous
mercury (Hg? ) which may in turn undergo oxidation to
mercuric mercury (Hg ). The toxicity of these three
oxidation states of mercury differs considerably. Metallic
mercury gives rise to symptoms and signs in man associated
almost exclusively with damage to the central nervous
system. Mercurous mercury forms a limited number of salts.
The most common of these is mercurous chloride or calomel
which is highly insoluble. Its toxic properties are mainly
that of an antiseptic or local irritant probably due to the
release of a small amount of ionized mercury. The divalent
form of mercury forms a variety of inorganic mercuric salts
most of which have a high acute toxicity. The signs and
symptoms resulting from these forms of mercury relate in
part to its local action as a corrosive chemical, such as
damage to the gastrointestinal tract and in part to systemic
action, mainly kidney damage. Hemorrhage in the lower gastro-
intestinal tract is also a possible systemic consequence of
ingestion of inorganic salts of mercury. The reason for the
qualitative differences in toxicity between metallic mercury
and inorganic mercury salts is not fully understood but
studies in recent years indicate the Importance of oxidation
reactions in explaining some of these.
Mercury vapor inhaled into the pulmonary spaces of the lung
is believed to cross the alveolar membranes rapidly due to
its high diffusibility (mercury vapor is a monatomic gas) and
high lipid solubility (the heptane to water ratio is about 20).
On entering the plasma where it dissolves in the metallic atomic
form it rapidly diffuses into the red blood cells to undergo
oxidation to divalent ionic mercury. The latter binds to ligands
in the red cells (Figure 1). However, a certain amount of the
dissolved vapor persists in the bloodstream for a sufficient
length of time (about 15 seconds) to reach the blood-brain
barrier which it easily crosses. Once in the brain tissue it
-------
undergoes oxidation to ionic mercury and again is available
for binding to tissue ligands. Inorganic mercury in the
ionic form is not available for transport across the blood-
brain barrier in substantial amounts because it is present
in plasma in a protein-bound form. In this metabolic scheme
it is clear that the rate of oxidation of the inhaled vapor
is a critical process controlling the amount of mercury
available for diffusion across the blood-brain barrier.
A considerable amount of work has now been carried out on the
mechanisms of the oxidation step. Work by Stock (1934) showing
that mercury vapor was oxidized to mercuric oxide in aqueous
solutions in the presence of oxygen and work by Clarkson et
al. (1961) showing that oxygen accelerated the in vitro
uptake of mercury vapor by blood suggested that a purely chemical
reaction was involved. However, the finding that uptake by red
cells was much more rapid that that by plasma suggested the
involvement of metabolic factors. This was confirmed by the
findings of Nielsen-Kudsk (1965) that ethanol depressed the
retention of mercury vapor by man and that ethanol could
depress mercury vapor uptake in vitro by blood. He also
demonstrated that the generation of hydrogen peroxide inside
the red cells was an important factor in the oxidation of mercury
vapor. Subsequent studies have pointed to the hydrogen
peroxide catalase system as the primary pathway of oxidation
of mercury vapor not only in red cells but in liver and probably
in other tissues (Magos et al., 1977) .
The oxidation action has important implications for the meta-
bolism of inhaled mercury vapor in the body and therefore for the
determination of health effects. Not only will the oxidation
reaction determine the residence time of the inhaled vapor
in the bloodstream and therefore its potential of reaching such
sensitive sites as the central nervous system and the fetus but
it also will allow the possibility of interaction with other
chemicals or genetic states which affect the activity of catalase,
The interaction between alcohol and mercury vapor has been
mentioned. Population groups which are genetically deficient
89
-------
in catalase have been described in several countries. Even
in the normal population, catalase activity in red cells has
a wide range of values.
That the hydrogen peroxide catalase system is responsible for the
oxidation of mercury vapor has important implications with
regard to the mechanism of action at the sub-cellular level.
It would seem likely that the ionic species is produced within
the cell wherever sufficient amounts of hydrogen peroxide and
catalase occur. In many cells this could be the peroxisomes and
lysosomes. Norseth and Brendeford (1971) have pointed to the
possibility that lysosomes damaged by mercury may release
hydrolytic enzymes. This could be the primary mechanism of
injury to the cell. The evidence that mercury vapor is probably
oxidized to ionic mercury within the peroxisome would lend
further support to Norseth and Brendeford's ideas.
3. Organometallic compounds
Many metals are capable of forming covalent bonds with the
carbon atom to create organometallic compounds. In general the
toxic properties of these compounds differ drastically
from those of the inorganic forms of these metals.
Tetraethyl lead, triethyl tin, trimethyl bismuth and methyl-
mercury, in contrast to the inorganic forms, can produce severe
damage to the central nervous system. Inorganic lead does
give rise to nervous effects but they are different from those
produced by the organic form of lead. In general the toxicity
of the longer chain alkyl derivatives of these metals is
less than that of the short chain compounds. Their toxicity to
the central nervous system is undoubtedly associated with their
ability to penetrate the blood-brain barrier rapidly.
The alkylation of lead, tin and bismuth does not occur in
mammalian tissues. In the case of tellurium and trivalent
arsenic, methyl derivatives are formed in the body but these
appear to be relatively non-toxic. Dimethyl telluride is
rapidly exhaled from the body and dimethylarsonic acid and
dimethylarsinic acid appear to have a low toxicity.
90
-------
Cleavage of the metal carbon bond occurs in the body in the
case of many of these organometallic compounds. In the case
of tetraethyl lead, the trialkyl metabolite is the toxic
species. In other cases, such as that of the carbon mercury
bond, cleavage may serve as a detoxification pathway. The
role of biotransformation in the toxicity of organometallic
compounds has been most extensively studied and is best
illustrated in the case of mercury, which forms a wide variety of
organomercuric compounds. A brief discussion will be given
below of three classes of carbon mercury bonds in terms of the
way in which the stability of the bond determines the toxic
properties of these compounds.
The short chain alkyl mercurials (ethyl, methyl and propyl
mercury) have probably the most stable mercury bonds of
all the organomercurial compounds. The stability of the carbon
mercury bond in methylmercury compounds is illustrated in
Figure 2. This records the blood concentration of methyl-
mercury upon hospital admission and during the subsequent period
of almost one year in a person who had been exposed to methyl-
mercury in an outbreak in Iraq. It may be seen that although
the methylmercury concentrations in blood eventually declined
to background values, the organic form of mercury persisted in
this patient for at least one year. Inorganic mercury levels
were also elevated at the time of exposure and shortly afterwards.
This was probably due to a slow conversion of methyl to inorganic
mercury. However, to explain the persistence of methylmercury in
the body for one year this conversion must be less than 1% per
day.
Conversion to inorganic mercury in the body probably does not
play a direct role in determining the damage inflicted on the
central nervous system in man or animal. At one time it was
suspected that the latency period between the end of
exposure and the appearance of signs or symptoms might be
explained by a slow accumulation in the brain of inorganic
mercury split off from the carbon atom. However, studies in
animals indicate that no such accumulation of inorganic
91
-------
mercury takes place. The opposite is true for the liver and
kidney where there is evidence from animal experiments to
indicate that effects on these organs may be due to inorganic
mercury split off from the parent molecule. The accumulation of
inorganic mercury in kidneys of rats given methyl- and
ethylmercury compounds is shown in Figure 3. After repeated
doses the increase in the concentration of inorganic mercury
in the kidneys of these animals is substantial. The kidney level
associated with acute damage following a single dose of
mercuric chloride is usually between 20 and 40 mg/kg wet
weight. It may be seen that on chronic dosing with methyl-
or phenylmercury compounds the level of inorganic mercury
can far exceed levels associated with acute damage. This
high level of inorganic mercury probably causes the ultra-
structural and biochemical changes seen in rats. Even so,
no gross structural damage is observed. These observations
suggest that the inorganic mercury released or accumulated in
the kidney slowly over a period of time due to the cleavage of
the carbon mercury bond is better tolerated by the tissue than a
single acute dose. The inorganic mercury is perhaps able to
stimulate the synthesis of metallothionein in this tissue and
it might then serve in a protective role. Studies of these
possibilities have not been reported.
Although the role of conversion to inorganic mercury in de-
termining the toxic effects of methylmercury seems to be
minimal, it is important in excretion of methylmercury from
the body. Most of the elimination of mercurv from the bodv in
animals or humans given methylmercury takes place via the
feces. Inorganic mercury is at least 50% of the total mercury
in the feces of experimental animals and in humans it may
amount to 100% (Turner et al., 1975). These data suggest that
conversion to inorganic mercury may be the rate-determining
step in the excretion of methylmercury from the body. Where this
conversion to inorganic mercury is occurring is not satisfactorily
established. In animals given methylmercury some inorganic
mercury is excreted in the bile but not enough to explain the
fecal excretion of inorganic mercury. It has been
92
-------
suggested (Norseth and Clarkson, 1971) that microflora
in the intestine might be able to convert methylmercury
to the inorganic form and thereby contribute to fecal
excretion. This methylmercury might originate as methylmercury
secreted from the bile or as methylmercury exfoliated from
intestinal cells. Some preliminary experiments indicated that
microflora in vitro were able to convert methyl to inorganic
mercury. The situation is not completely clear since further
experiments by Norseth (1971) indicated no difference in
excretion rate or biological half-time of methylmercury
in germ-free animals as compared to normal animals. This
is an area of considerable research interest as we are in
great need of understanding the mechanisms whereby this
hazardous form of mercury is eliminated from the body as
well as the very large individual variations in the biological
half-time of methylmercury in man.
The aryl and alkoxyalkyl mercurials form another distinct
class of organomercurials that are less toxic than the alkyl
compounds. One reason for their low toxicity is their rapid
conversion to inorganic mercury. There is evidence that the
conversion takes place in the liver and is mediated by an
enzyme in the cytosol of the liver cell. The mechanism is
not understood in detail except that it is believed to occur
by reductive cleavage of the carbon mercury bond.
The organomercurial diuretics form an important class of
mercury compounds. Considerable investigations have been
carried out as to their biological action and the mechanisms
thereof. Despite the introduction of a variety of new
types of diuretic agents in recent years, the organomercurial
diuretics still provide one of the most selective ways in
which to inhibit active sodium transport in the renal tubule.
This in itself is surprising insomuch as organic mercury cations
can react with a variety of tissue ligands and would therefore
not be expected to be so selective in their biological action.
Investigations have been carried out to try to discover how
compounds that are so unselective in their actions in vitro
93
-------
could produce such selective and graded damage in vivo.
An important landmark in these investigations was the
publication of the mercuric ion hypothesis by Weiner
et al. (1^62). The assumptions of this hypothesis are
depicted diagrammatically in Figure 4. The organomercurial
diuretic is present in plasma bound mainly to plasma proteins
but a small fraction is bound to a diffusible amino acid presumed
to be cysteine. The diffusible and non-diffusible compounds are in
equilibrium. The cysteine-bound form may enter the tubular
fluid via filtration through the glomerulus and via excretion
across the tubular cells. The organomercurial is reabsorbed in
the lower proximal tubule. At this point the mercurial compound
may bind to non-specific sites in the cell, resulting in no
biological effects. Alternatively the organomercury compound
may undergo cleavage of the carbon mercury bond releasing
ionic inorganic mercury to bind to a specific receptor site
leading to inhibition of sodium transport. The receptor site
is bidentate. The divalent inorganic mercury forms a ring
structure with the two ligands. Formation of this ring is the
essential prerequisite for diuretic action.
This hypothesis has some support by virtue of the fact that
in vitro all organomercurial diuretics are unstable and readily
release inorganic mercury under acid conditions in the presence
of a complexing agent such as cysteine. The mercuric ion
hypothesis also offered a simple explanation for the well-
known fact that mercurial diuresis is potentiated by acidosis
as well as for the fact that certain organic mercury compounds
such as parachloromercury benzoate and paramercuribenzene
sulphonate are able to block the diuretic action of organo-
mercurials. Both of these compounds proved to have stable
carbon mercury bonds in the in vitro test. It was argued that
they would attach to only one ligand in the receptor and
thereby either displace inorganic mercury from the ring structure
or prevent inorganic mercury from forming the ring structure
with the receptor (Millar and Farah, 1962). Indirect support
was given to the mercuric ion hypothesis by the observation
94
-------
that mercurial diuretics do liberate inorganic mercury in
vivo in kidneys of rats and dogs. At the same time, it was
observed that non-mercurial diuretics also released inorganic
mercury and that the levels of inorganic mercury in the kidney
tissue were not increased during acidosis (Vostal and Clarkson,
1973). At the present moment therefore the mercuric ion
hypothesis cannot be said to be definitively established but
it has many attractive aspects in the context of trying to
understand the selective toxicity of metals at the molecular
level.
The formation of chelate ring structures by polyvalent metals
may play a role in their selective toxicity. An example is
the case of trivalent arsenic which forms a ring compound
with lypoic acid, resulting in the inhibition of keto
acid dehydrogenases. This is believed to be the biochemical
lesion behind the biological effects.
4 • Reaction of metals with ligands
Metal cations may form coordinate-covalent bonds with a
wide variety of electron donating groups (ligands) . The
overall .reaction is represented as:
M + L £ ML
where M is the metal cation and L is the monodentate ligand.
The equilibrium constant or affinity constant (K) is expressed
by the Mass law relationship
K =
Most ligands of biological importance contain the oxygen, nitrogen
or sulfur atom.
The biological consequences of metal interaction with ligands
have recently been discussed in detail by Webb (1977) . Some of
the immediate ones are breaking of hydrogen bonds, displacement
95
-------
of other metals from a ligand site and change in the tertiary
structure of a protein leading to inhibition of catalytic activity.
The attachment of metals to ligands in membrane structures
can lead to inhibition of active transport or to alteration in
the passive permeability properties of the membrane. Inter-
action of metals with ligands of nucleic acids potentially
could affect both transcriptional and translation processes
and may underlie the mutagenic and carcinogenic action of
certain metals. Inhibition of DNA repair enzymes might also
contribute to the mutagenic and carcinogenic potential of a
metal.
Each metal possesses its own spectra of affinity constants for
different tissue ligands. These are illustrated for methyl-
mercury and mercuric ion cations in Table 1. The wide range of
values for the affinity constants expressed in logarithmic
units is obvious. Significantly high affinity constants are
seen even with such simple ligands as hydroxyl and chloride
anions. In the case of methylmercury, the highest values are
seen in molecules containing the sulfhydryl group. Other
metal cations will usually exhibit a different spectra of
affinity constants. Nevertheless, it appears to be a general and
implicative rule that most metal cations can form stable
complexes with a wide variety of ligands. This raises the
question as to how metals that are so reactive to such a wide
variety of tissue ligands can produce such selective actions
in the whole animal. This question is by no means fully
answered at this time but this section will try to deal with
the general principles of metal interaction with ligands and to
describe a few cases where selective actions of a metal may
be understood.
4. i interaction of metals with diffusible ligands
Following absorption from the gastrointestinal tract or
the lung the metal may enter the bloodstream as the free
cation (M) or in a complexed form (ML) and may undergo
reactions with diffusible ligands in plasma (L ) as follows:
96
-------
M + L1 ^ ML1
or ML
+ L1 ? ML1 +
Examples are known where these liqands modify the toxicity of
the metal and play a decisive role in determining the site
of action in the body.
The beryllium cation (Be ) has a high affinity for
oxygen-containing ligands such as hydroxyl and phosphate.
Immediately upon entering the plasma, beryllium is believed
to form insoluble compounds involving hydroxyl and phosphate
anions, leading to the formation of colloidal beryllium
salts. This accounts for the observation that beryllium is
rapidly taken up by the reticulo-endothelial system following
its absorption into the bloodstream. Subsequent removal of
beryllium from the RE cells in the liver and its transport
to bone is believed to be the result of a slow solubilization of
the beryllium cation. The carrier molecule in this case is
unknown.
Ionic lead rapidly forms insoluble precipitates with inorganic
phosphate anions. The reaction is highly pH dependent and
involves the interaction of lead with the trivalent phosphate
3_
anion (PO. ). Early studies by Aub and his
co-workers (1926) clearly identified the role of phosphate
in limiting the solubility of lead in plasma. Lead phosphate
colloids will carry a net negative charge in the presence of
excess phosphate in plasma. These colloids have the potential
for absorbing organic molecules such as amino acids or
proteins to form'a highly dispersed colloid in the form of
peptised sol. It has been proposed that the initial step in
lead uptake by red cells involves the adsorption of the small
colloidal particles on the surface of the red cell. There is
some electron microscopical evidence to support this pathway
of uptake (for review see Passow et al., 1961). Eventually
lead may penetrate into the interior of the cell where it is
available to interact with internal cellular ligands, resulting
97
-------
in the inhibition of the enzyme ALA-dehydratase. Lead may also
enter the red cells during hematopoiesis. At the present moment
it is not clear what fraction of the lead in the interior of the
cell originates from lead uptake during development of the
cell and what fraction represents lead penetration across the
red cell membrane.
The interaction of the uranyl ion (UO- ) with bicarbonate
anions and the carboxylate groups of serum proteins was
studied intensively in the 1940 "s and became a classic example
of the importance of metal ligand interaction in determining the
selective action of a metal cation. It was found that uranium
absorbed into the bloodstream as UO~ rapidly formed
complexes with bicarbonate and with carboxylate anions on
serum protein. The bicarbonate complexes were readily
filterable in the kidney and entered the proximal tubular
fluid. Acidification processes in the proximal tubule removed
the bicarbonate anions, liberating the free uranyl cation.
Thus UOp became available for reaction with cellular
ligands in the proximal tubule of the kidney. This mechanism
is believed to be the explanation for the fact that UO2
damages primarily the cells of the proximal tubule. Moreover
such a mechanism gave an explanation for the dramatic effects
of acidosis and alkalosis on the excretion and toxicity of
uranyl salts. In acidotic animals the bicarbonate removal
from urine is complete so that most of the uranium filtered as
the bicarbonate complex is deposited into the renal cortex. On
the other hand in animals made alkalotic by treatment with
sodium bicarbonate a substantial amount of the bicarbonate
passes into the urine, carrying with it an appropriate quantity
of U0_ . Alkalotic animals thus have considerably less
deposition in the kidney tubules and show fewer toxic effects
of uranium.
The interaction of metals with organic diffusible molecules
such as amino acids is less well-documented in vivo. Most metals
are either protein-bound or attached to the red cells in blood
so that the amount actually in a diffusible form is quite small.
98
-------
The identification of the chemical nature of these diffusible
complexes therefore presents considerable analytical problems
while at the same time these complexes may be quite important
in terms of the toxicity of the metal, as has been well
demonstrated in in vitro studies.
Vander (1962) has shown that the infusion of a cadmium
cysteine complex into the arterial circulation of the kidney
in dogs leads to an enhanced reabsorption of sodium in the
proximal tubule. This might be important for some of the
cardiovascular effects of cadmium. In vitro studies on the
isolated intestine indicated that silver complexes of cysteine
also stimulate active ion transport (Clarkson and Toole, 1964).
It is not clear from the available data whether the cysteine
complex itself interacts with the membrane to produce these
transport changes or whether the cysteine modifies the
chemical potential of the metal, allowing only small amounts
of the free metal cation to react with the membrane sites.
One matter that is clear is that the cysteine complexes of
certain metals differ from the inorganic salts in terms of
binding activity. For example, the chloride salts of inorganic
mercury and a variety of organomercurial diuretics are potent
membrane poisons that can both inhibit active transport and
increase membrane permeability. An attempt has been made to
explain the selective localization of the effects of organomer-
curial diuretics in the renal tubule in terms of a supposed
equilibrium between protein amino acid and chloride complexes of
these mercurial compounds (Clarkson and Vostal, 1973). It is
assumed that the mercurial diuretic enters the tubular fluid
either by filtration at the glomerulus or by active transport
across the tubular cells. Following similar arguments as
were used for bicarbonate complexes of uranium, it is then
assumed that the organomercurial cysteine complex undergoes
dissociation as cysteine is reabsorbed in the proximal tubules
of the kidney. This allows the formation of the chloride
complexes of the organomercurials since the chloride concentration
is maintained, if not increased, as the tubular fluid moves
along the nephrons.
99
-------
The reactions occurring within the renal tubular fluid were
assumed to be:
R-Hg+ + Cl~ * R-HgCl
R-Hg+ + Cys" ^ R-HgCys
Cys" + H+ £ CysH
where R-Hg is the organomercurial cation and CysH and
Cys are the cysteine molecules with undissociated and
ionized sulfhydryl groups, respectively.
It may be seen that increasing the acidity as well as
increasing the chloride concentration will facilitate the
formation of the chloride complex as the amino acid is
reabsorbed. Making reasonable assumptions concerning the
affinity constants of the organomercurial for the sulfhydryl
group and for the chloride anion and the pK of the sulfhvdryl
group, it is possible to estimate that a small fraction
of the mercurial would indeed be present as the chloride
complex in the renal tubular fluid. This may be the species
that penetrates into the cell where it may undergo biotrans-
formation, allowing inorganic mercury to react with the bidentate
receptor as discussed previously.
The formation of amino acid complexes with metals may in
some cases be enzyme-mediated. Methylmercury is excreted in
the bile of mice and rats as a small molecular complex with a
sulfhydryl compound, which may be the amino acid glutathione.
The process may be catalyzed by conjugation enzymes in the
liver such as glutathione transferase, although evidence on
this is still not conclusive. Such complexes secreted into
the bile undergo reabsorption in the gastrointestinal tract.
This is known to be true for methylmercury and possibly for
other cations. Such small molecular weight complexes, formed
within the liver, secreted in bile and reabsorbed into the
bloodstream may represent an important mobile form of the
metal in the body, allowing the metal to reach its site of
action (Norseth, 1973).
100
-------
4.2 Interaction of metals with macromolecules
The interaction of metals with nucleic acids and proteins
has been studied extensively in vitro. Unfortunately in
most of these types of experiments, the conditions were un-
physiological. The presence of small diffusible anions such
as chloride, bicarbonate or amino acids could substantially
change the interaction with ligands and macromolecules. In
most cases it is impossible to extrapolate in vitro physio-
chemical observations to the situation in the whole animal.
For theoretical reasons it is clear that interactions
with macromolecules are probably responsible for the toxic
effects of most metals. The receptor for the toxic action is
probably in most cases a macromolecule functioning catalytically
or as a structural or transport component of a membrane.
Webb (1977) has discussed the interaction of heavy metals
with nucleic acids and has pointed to phosphate groups and
nitrogen atoms in heterocyclic bases as likely sites for co-
ordinate covalent bonding with metal cations. Such metals as
cadmium and copper probably bind to the heterocyclic bases,
resulting in the breaking of hydrogen bonds and the desta-
bilizing of the DNA structure. Metals may also be essential
for enzymes that are involved in the synthesis and repair of
nucleic acids. Zinc is necessary for the catalytic activity
of thymidine kinase. Its replacement by other metals such
as cadmium may affect the functioning of this enzyme and
may therefore have effects on DNA synthesis. The mutagenic
activity of certain arsenic compounds has recently been
attributed to the inhibition of DNA repair enzymes by this
metal (Rossman et al., 1977).
The interaction of metals with proteins and particularly with
enzymes has been subjected to more extensive study than any
other area of metal interaction on the molecular level. Many
factors influence this interaction, as demonstrated by a
variety of in vitro studies. For example, the pH of the medium
is important since hydrogen ions compete with the metals for
101
-------
binding to most of the ligands. Diffusible anions such as
chloride and hydroxyl will limit the reactivity of the metal
cation with the macromolecular ligand. Some of the ligands
in the macromolecule may be buried within the tertiary
structure and not readily available for interaction with
the metal cation. Reaction of the metal cation with external
ligands on the protein may lead to the unfolding of the protein
chains, leading to time-dependent reactions. Neighboring groups
may influence the reactivity of the metal with a particular
ligand. These factors considerably influence the type of
inhibition that metals can exert on an enzyme. Direct inter-
action of the metal with the active site follows the kinetics
of competitive inhibition. For example, mercury inhibits
the enzyme invertase but does so competitively with the
substrate sucrose. A metal may react with sites adjacent
to the active site of the enzyme, resulting in partial and
non-competitive inhibition. The sites remote from the active
center may also interact with the metal, which may produce
either no effect or, in certain cases, quite dramatic effects
possibly due to changes in the tertiary protein structure.
An example of such a reaction is that between inorganic mercury
and hemogloblin. Riggs (1952) has shown that although
mercury does not react with the active oxygen binding sites
in hemogloblin it does substantially change the oxygen dissociation
curve of this molecule by changing the tertiary structure.
An interesting example of the selective action at the molecular
level is the case of ionic inorganic mercury (Hg++) and
methylmercury (CH-Hg ) interaction with the receptor for
acetyl chlorine. Shamoo et al. (1976) have contrasted the
considerably greater ability of the CH3Hg as compared to
Hg to block the activity of the isolated acetyl choline
receptor. The proposed explanation is that CH.,Hg because
of its higher lipid solubility can penetrate regions of the
acetyl choline receptor not available to Hg . Such a
difference in ionic action may help to explain the findings
that neuromuscular blockage can be produced by methylmercury
in phrenic nerve diaphragm preparations and that muscular
102
-------
weakness can occur in patients poisoned by methylmercury
but not in those poisoned by inorganic mercury (Rustam et al.,
1975) .
Some examples of enzyme inhibition in vivo by metals are
well documented. The effect of lead on heme synthesis is well-
known. The enzyme delta amino levulinic acid dehydratase (ALA/D)
probably is the most sensitive enzyme with regard to lead.
Its inhibition leads to decreased synthesis and to negative
feedback control, which may in turn lead to the derepression
of the enzyme delta amino levulinic acid synthetase and thus
increase synthesis of delta amino levulinic acid. This appears
in the urine as an early warning of the toxic effects of
lead.
The inhibition of an enzyme may not always be the critical step
for the biological or toxic effects of the metal for the whole
animal. A classic example is the inhibition of the enzyme
ALA dehydratase in red blood cells, which apparently plays
little role in the biological effects of lead. The enzyme may
have been important for the early development of the red cell
but ceases to perform any useful physiological function in the
mature erythrocyte.
The availability of macromolecular binding sites can be
modified by the diet under certain circumstances and this may
in turn affect the toxicity of the metal. Surtshin (1957)
demonstrated that the sucrose feeding protected rats against
the renal toxic effects of mercurial salts. It was subsequently
shown that such a diet increased the protein bound SH in kidney.
The protective effect probably could be attributed to the
binding of mercury to those non-specific sites (Surtshin and
Yagi, 1958).
Reaction with proteins is a necessary step in the antigenic
action of metals. The pulmonary tissue reaction to beryllium
is believed to be produced by cell-mediated immunity, probably
involving protein complexes of beryllium. Metal fume fever seen
103
-------
after inhalation of zinc is also believed to be a result of
reaction to antigenic protein complexes of zinc.
4.3 Interaction of metals with membrane-bound ligands
Rothstein (1973) has pointed out that even at toxic doses
of a metal only a small fraction of tissue ligands is
involved in heavy metal complexes. For example, assuming
that methylmercury is attached to sulfhydryl groups in
the red cells, a lethal concentration of 5000 ,ug/kg would
involve the binding of only one sulfhydryl group in a thousand.
Thus the important point to bear in mind is not the average pro-
perties of all the sulfhydryl groups but the properties of those
few sulfhydryls that are involved in metal binding. The same
principle applies to other abundantly available ligands such as
amino and carboxylate groupings. Part of the problem in inter-
preting whole animal effects from our knowledge of molecular level
effects lies in our difficulties of studying the properties
of a small number of ligands which may be involved in metal
binding.
An excellent example of the importance of a critical sub-group
of ligands of the same type is the case of metal binding
to membrane-bound ligands. The ability of the metal to
interact with such ligands is determined not only by chemical
specificity but also by what Rothstein has described as
"geographical" specificity. Chemical specificity may be equated
to the affinity constant of the metal for the ligand. Geographi-
cal specificity is the ability of the cation to diffuse
to ligands buried within the membrane.
Several principles govern the interaction of metals with
cellular ligands. The outer surface of the cell membrane will
be the first part of the cell to come into contact with the
metal. Heavy metals that penetrate the cell membrane very
quickly, such as methylmercury cations, probably undergo
very little reaction with membrane components. On the other
hand, those metals which pass through the membranes slowly
and are able to form tight bonds with ligands might seriously
104
-------
affect membrane function. In general, many more metal binding
sites will be found in the interior of the cell than in the
membrane. The possibility of. protection in terms of metal
binding to non-specific sites is less in the case of interaction
with the membrane than in the case of interaction with components
within the cell.
An important example of geographical specificity has been
illustrated by Rothstein in the interaction of organomercurial
cations with the sulfhydryl ligands in the red cell membrane
(for review see Rothstein, 1973). The organomercurial chlor-
merodrin labelled with a radioisotope of mercury was allowed
to react for a brief period of time with a suspension of red
cells at a low temperature. The organomercurial proved to
react only with sulfhydryls on the outer surface of the cell
membrane. Quantitative aspects of this finding are indicated
on the Scatchard plot given in Figure 5. A graph of the bound over
the free organomercurial versus the bound mercurial gives a
linear relationship over most of the concentration range. The
Scatchard equation has found extensive use in studies of metal
binding to protein and, more recently, of metal binding to
cell membranes. The slope of the line is a measure of the affinity
constant of the metal for the ligand and the horizontal
intercept is the measure of the number of ligands per protein
or per red blood cell. The number of SH groups bound to chlor-
merodrin in this example is less than 1% of the total SH groups
inside the red cell and represents only a fraction of those
SH groups in the entire membrane. Reaction of chlormerodrin
with this small number of SH surface groups results in the
inhibition of sugar transport across the red cell as shown
in Figure 6.
Reaction with the surface groups has no effect on the potassium
transport properties of the red cell membrane. Only in
the case where an organomercurial penetrates within the structure
of the membrane itself is potassium permeability affected.
The point of these studies is to illustrate that geographical
specificity may be just as important as chemical specificity in
105
-------
determining the selective action of a metal. They also
indicate that there may be time-dependent effects of metals
as the metal cation diffuses from the outside to the interior
of the cell.
The interaction of metals with membranes within the cell
such as those in the mitochondria and microsomes is undoubtedly
governed by similar principles as those described by Rothstein
for the plasma membranes of cells. Aldridge (1970) was able
to show that in mitochondria the cation triethyltin was
probably bound to histidine residues. The interaction of
triethyltin with mitochondrial membranes leads to uncoupling
of oxidative phosphorylation.
Webb (1977) has discussed examples where divalent metal
cations can displace calcium from the nerve membranes at
neuromuscular junctions. Heavy metals may also react with
SH groups on the nerve membrane, thus affecting the transport
of calcium. The effects of many of the metals on the central
nervous system could well be mediated through their affecting
the activity of receptors to hormonal transmitters such as
acetyl choline, as has been demonstrated in the case of
methyImercury.
Acknowledgement
This paper is based on work performed under a Center grant
(ES 01247) and Program Project grant (ES 01248) from the
National Institute of Environmental Health Sciences and under
contract with the U.S. Energy Research and Development Ad-
ministration as a part of the University of Rochester
Biomedical and Environmental Research Project. It has been
assigned report no. UR-3490-1253.
106
-------
REFERENCES
Aldridge, W.N. (1970). In: "Effects of Metals on Cells, Sub-
cellular Elements and Macromolecules." (J. Maniloff, J.R.
Coleman and M.W. Miller, eds) pp 255-271. Charles C. Thomas,
Springfield, 111.
Aub, J.C., Fairhall, L.T., Minot, A. and Reznikoff, P.
(1926). "Lead Poisoning." Williams & Wilkins, Baltimore.
Clarkson, T.W. (1972). "Recent Advances in the Toxicology of
Mercury with Emphasis on the Alkylmercurials." Vol. II, pp
203-234. CRT, CRC Press, Cleveland.
Clarkson, T.W. (1976). In: "Trace Elements and Human Disease."
(A. Prasad, ed) pp 453. Academic Press, New York.
Clarkson, T.W. and Toole, S.R. (1964). Amer. J. Physiol.
206, 658-668.
Clarkson, T.W. and Vostal, J.J. (1973). In: "Modern Diuretic
Therapy in the Treatment of Cardiovascular Disease." (G.
Wilson and A.F. Lant, eds) pp 221-232. Excerpta Medica
Foundation, Amsterdam.
Clarkson, T.W., Gatzy, J. and Dalton, C. (1961). Studies on
the equilibration of mercury vapor with blood. University of
Rochester Atomic Energy Project, Rochester, N.Y., Report No.
UR-582.
Gage, J.C. (1964). Brit. J. Ind. Med. 21, 197-202.
Magos, L. and Butler, W.H. (1976). Arch. Toxicol. 35, 25-
39.
Magos, L., Halbach, S. and Clarkson, T.W. (1977). The role
of catalase in the oxidation of mercury vapor. Biochem.
Pharmacol. (In press).
Miller, T.B. and Farah, A.E. (1962). J. Pharmacol. Exp.
Therap. 135, 102-111.
Nielsen-Kudsk, F. (1965). Acta Pharmacol. Toxicol. 23, 263-
274.
Norseth, T. (1971). Acta Pharmacol. Toxicol. 30, 172-176.
Norseth, T. (1973). In: "Mercury, Mercurials and Mercaptans."
(M.W. Miller and T.W. Clarkson, eds) pp 264-272. Charles C.
Thomas, Springfield, 111.
Norseth, T. and Brendeford, M. (1971). Acta Pharmacol.
Toxicol. 29, 375-384.
Norseth, T. and Clarkson, T.W. (1971) . Arch. Environ. Health
£2, 568-577.
Passow, H., Rothstein, A. and Clarkson, T.W. (1961). Pharmacol.
Rev. .13, 185-224.
107
-------
Riggs, A.F. (1952). J. Gen. Physiol. 36, 1-16.
Rossman, T.M., Meyn, S. and Troll, W. (1977). Effects of
arsenite and DNA repair in Escheria Coli. Environ. Health
Perspect. (In press).
Rothstein, A. (1973). In: "Mercury, Mercurials, and Mercaptans."
(M.W. Miller and T.W. Clarkson, eds) pp 68-92. Charles C.
Thomas, Springfield, 111.
Rustam, H., Von Burg, R., Amin-Zaki, L. and Elhassani, S.
(1975). Arch. Environ. Health 3_£, 190-195.
Shamoo, A.E., MacLennan, D.A. and Eldefrawi, M.E. (1976).
Chem. Biol. Int. 12., 41-52.
Stock, A. (1934). Z. Anorg, Chem. 217, 241-253.
Surtshin, A. (1957). Amer. J. Physiol. 190, 271-277.
Surtshin, A. and Yagi, K. (1958). Amer. J. Physiol. 192,
405-409.
Turner, M.D., Smith, J.C., Kilpper, R.W., Forbes, G.B. and
Clarkson, T.W. (1975). Paper presented at the International
Conference on Heavy Metals in the Environment. Toronto,
Canada, Oct. 27-31.
Vander, A.J. (1962). Amer. J. Physiol. 203, 1005-1007.
Vostal, J.J. and Clarkson, T.W. (1973). In: "Mercury, Mercurials,
and Mercaptans." (M.W. Miller and T.W. Clarkson, eds) pp
139-155. Charles C. Thomas, Springfield, 111.
Webb, M. (1977). In: "Clinical Chemistry and Chemical Toxicology
of Metals." (S.S. Brown, ed) pp 51-64. Elsevier, Amsterdam.
Weiner, I.M., Levy, R.I. and Mudge, G.H. (1962). J. Pharmacol.
Exp. Therap. 138, 96-112.
108
-------
Table 1. Dissociation constants (negative logarithm) for
complexes of methylmercuric and mercuric ions
(From Clarkson, 1972).
Ligand CH - Hg+ Hg+
Cl 5.4 6.7
OH 9.5 10.3
Histidine (NH ) 8.8 10
Cysteine 15.7 14
Albumin 22.0 13
109
-------
Figure 1. A schematic representation of the metabolism of
inhaled mercury vapor. The vapor rapidly crosses
the pulmonary membranes and enters the plasma
compartment. From this location it may enter the
red blood cells where it undergoes oxidation to
form divalent ionic mercury. The latter binds to
cellular ligands. Elemental mercury may also cross
the blood-brain barrier where it is oxidized,
resulting in tissue binding of ionic mercury.
Mercury vapor dissolved in plasma is believed to
persist for sufficient time to allow diffusion
into all tissues in the body including passage
across the placenta and uptake by the fetus (From
Clarkson, 1976).
110
-------
1000
ORGANIC
INORGANIC
200 300
DAYS AFTER EXPOSURE
Figure 2. The concentration in whole blood of organic and
inorganic mercury in a person after dietary
exposure to methylmercury.
Ill
-------
100-
80-
3.5 mg Hg/Kg/WEEK
4.2 mg Hg/Kg/WEEK
• PHENYL
• METHYL
O METHYL
3.5mg Hg/Kg/WEEK
4 8 12
WEEKS OF EXPOSURE
Figure 3. The accumulation of inorganic mercury in rat
kidneys following repeated doses of phenyl or
methylmercury compounds (Data from Gage, 1964, and
Magos and Butler, 1976) .
112
-------
Plasma
Kidney
R Hg S Protein '
RHgSCyst
1
-------
MOLES BOUND/RBC * 10
18
Figure 5. A mass law (Schatchard) plot of the binding of
chlormerodrin by human red blood cells (From
Rothstein, 1973).
114
-------
lOCh
CHLORMERODRIN (6°C)
0 10 20 30 40 50 60
TIME IN MINUTES
Figure 6. The reversal of the inhibition of chlormerodrin on
glucose efflux from human red blood cells as
related to the binding and desorption of the agent
(From Rothstein, 1973).
115
-------
FACTORS INFLUENCING EFFECTS
AND DOSE-RESPONSE RELATIONSHIPS OF METALS
Gunnar Nordberg, Jiri Parizek and Magnus Piscator
There is growing evidence that absorption, metabolism
and/or effects of both essential and non-essential metals can
be substantially influenced by various factors. The presence
or absence of such factors may sometimes modify the character
of effects and in other instances change dose-effect and
dose-response relationships. These modifying factors include
certain characteristics of the exposed organism, e.g. constitu-
tional factors like sex and age, which often are called host
factors. Absorption, metabolism, and effects of metals can also
be modified by simultaneous or previous exposure to certain
environmental agents. In some cases such agents exert their
effect through a direct interaction with the metal in question
at the molecular level. In other cases an indirect metabolic
change is induced by the interfering environmental factor which
gives rise to the change in dose-response relationship. Such
indirect interactions are sometimes of considerable interest.
However, it should be recognized that in certain cases, even
where the influence of a certain environmental factor is well
established, the corresponding mechanism may not be known.
The term "interaction" may in such instances be used in a
broad sense. Factors effectuating modifications of metal
toxicity may range from direct interactions between a metal
and another metal, e.g. cadraium-zinc interaction, to the influence
of socio-economic and behavioral factors on toxic effects of
metals.
For several decades chemical and physical interactions of metals
have been studied in experiments both in vitro and in vivo in
animals. There is an abundance of data on such interactions
and no attempt will be made in this chapter to review all
these data. Many results from such studies have found practical
application in the field of animal nutrition, but their signifi-
cance for imbalance of trace elements and for toxic effects of
117
-------
some metals in the human being has been difficult to
evaluate. Only such data that are of general interest and
data that are thought to have a bearing on the human situation
will be taken up. The experimental work has usually been
performed on healthy animals of the same strain in carefully
controlled environments. Often the doses or exposure
routes have been of little relevance to the human situation.
Nevertheless such studies may sometimes give indications about
what might happen in human beings and may be a basis for
further studies with more relevance to human beings.
At a meeting in Tokyo, 1974 (Nordberg, 1976) some results
from such animal studies were discussed and several interactions
of possible importance for human beings were recognized.
Further considerations concerning interactions among metals
and other factors of importance in metal toxicology were
given during a meeting held in Stockholm, 1977, by the
Task Group on Metal Interactions (Nordberg,, 1978). At the
Tokyo meeting interaction was defined as a process by which
metals in their various forms, or other factors, change the
critical concentration or a critical effect of a metal under
consideration. At the Stockholm meeting this definition
was accepted with the addition that the term interaction
should also be used to describe the influence of other
substances and factors on the metabolism and toxicity of
metals (Task Group on Metal Interaction, 1978).
1. Definitions
As mentioned above the factors that will be discussed in
this chapter and which can influence metal metabolism and
toxicity include both host factors and external factors
like climate etc., but the main part of the
document will be concerned with interactions between metals
and other metals or chemicals. Such interactions may be
both direct and indirect, in which case a distinction
may be made between "non-interactive" and "interactive"
joint action.
118
-------
"Independent action" is one type of non-interactive joint
action. This is by far the most common type and occurs when
two metals have different sites of action or different modes of
action, and when they do not influence each other. Another
example of non-interactive joint action is when two metals
exert action at the same site and their modes of action
are similar but they do not influence the action of each
other. The result of such joint action is an addition of
effects which can be estimated from the dose-effect and
dose-response relationships of the constituents of a mixture.
Such a type of joint action is the one of arsenic and lead in
relation to coproporphyrin excretion (Fowler, 1978).
In addition to these non-interactive types of joint action
of chemicals there are other types of interactions where
the dose-effect or dose-response curve of the combination
cannot be assessed from those of the individual chemicals.
Synergism takes place when the effect or response of the
combined exposure is greater than additive. Examples are the
synergistic effects of cigarette smoking and asbestos in
inducing lung cancer. An example involving metals is the
greatly increased teratogenic effect seen in animals from
the combined injection of cadmium and lead in relation to
the effects of each of the metals separately.
Antagonism takes place when one factor reduces the effect of
another. An example of relevance for this chapter is the
interaction between selenium and various forms of mercury
which will be discussed in more detail later (section 2.4).
With regard to the joint action of metals or metals and
other chemicals, the type of interaction will be classified
as additive, synergistic or antagonistic, which are common
concepts in pharmacology, whenever possible in the text to
follow.
119
-------
Some of the mentioned types of combined action do not apply
exclusively to chemicals but also to other exposure factors,
such as ionizing radiation, climatological factors, and
infectious agents. A synergistic action of smoking and
exposure to radon has been known to increase the risk for
lung cancer. In animal experiments, variations in temperature
have been shown to cause great differences in the LD of
one and the same chemical.
When it comes to such host factors as age, sex, race and other
genetic factors it is not appropriate to talk about an inter-
action. Nevertheless, it is well recognized that such factors
influence dose-effect and dose-response carves of metals.
Such factors will therefore be included in this chapter.
Since data on interactions of other factors in principle
can be found in most papers dealing with metal toxicity, it is
obvious that it would be an enormous task to extract all the
relevant data. Only in a few instances can it be seen directly
that a report is concerned with interaction. This chapter will
therefore focus on metal-metal interactions and take up
several of the above mentioned factors rather briefly, even
though they are probably of very great importance.
The specific information given in this chapter relies to a
large extent on data and conclusions presented at the meeting
in Stockholm and reported by the Task Group on Metal Inter-
action (Nordberg, 1978). That group also discussed some
special cases of interactions like those resulting from
suboptimal intakes of certain nutrients, but this will not
be regarded as a separate category here.
2. Metal-metal interactions
These interactions have long been a subject of study in the
field of animal nutrition. It was recognized early that livestock
had to have a carefully controlled intake of essential elements
in order to ensure high food standards for humans (see e.g.
120
-------
Underwood, 1977) . During later years the need has arisen to
extend the observations in this field to non-essential
elements since the increase in pollution in many areas of the
world has caused an unwanted exposure to toxic metals. Studies
on effects of metals on human beings, both in the general and
the industrial environment, have given us an awareness that
simultaneous exposure to several toxic elements might occur
and that the intake of essential elements might at the same
time be marked by deficiencies. Presently of special concern
with regard to the human situation are arsenic, cadmium, lead
and mercury, which were discussed in some detail at the
Stockholm meeting (Nordberg, 1978). In the following the
present experience with regard to these metals and some other
metals of interest will be summarized. Data on specific inter-
actions will also be found in the chapters on the specific
metals in Volume II of this series.
2.1 Interactions between arsenic and other metals
Special attention has been devoted to selenium since both
arsenic and selenium can form similar types of ions, for which
reason antagonism could be expected. Both metals can be
methylated in vivo (Braman and Foreback, 1973; Francis et al.,
1974) which may complicate the understanding of possible
interactions. It has been shown that arsenic can decrease the
toxicity of selenium, one mechanism being that arsenic increases
the excretion of selenium via bile (Levander and Baumann, 1966a,
1966b). On the other hand, injection of arsenic compounds was
found to increase the toxicity of methylated selenium
compounds (Obermeyer et al., 1971). It is also known that
selenium in turn will increase the biliary excretion of
arsenic (Levander and Baumann, 1966a) and that high dietary
levels of selenium cause decreases in organ concentrations of
arsenic (Whanger, 1976).
There are, however, no human data which can support these
findings in animals. It is thus not known at present to what
extent these interactions are of importance in human beings.
Simultaneous exposure to selenium and arsenic may occur in
121
-------
industrial environments. Certain groups of the general
population are exposed to large amounts of arsenic from
drinking water or to methylated arsenic compounds in marine
organisms. The selenium intake of such groups could be of
importance for dose-response relationships for arsenic.
The influence of cadmium and lead on the toxicity of arsenic
has been investigated in a few animal experiments. It was found
that lead additively increased protoporphyrin excretion in
arsenic-treated animals, whereas cadmium had no effect (Mahaffey
and Fowler, 1977) . Since the doses were high it is impossible
to draw any conclusions from these studies with regard to
the human situation. A combination of cadmium, lead,
and arsenic is not uncommon in certain industrial environments.
2.2 Interactions between cadmium and other metals
That zinc can counteract many of the toxic effects of cadmium
has been shown in a large number of studies. Cadmium and
zinc belong to the same group in the periodic table. It has
been speculated that cadmium can displace or replace zinc in
some essential systems in the organism, thus causing functional
changes. Especially some parts of the reproductive system
have very high concentrations of zinc and many of the sexual
functions require zinc. Large amounts of cadmium in single
doses may cause testicular destruction, but this damage may
be prevented by injection of large amounts of zinc (Parizek,
1957) . In many of the earlier studies on relationships
between zinc and cadmium (see e.g. review by NAS, in press),
animals exposed to cadmium were on diets with extremely
high concentrations of zinc, which might have prevented some
of the expected effects of cadmium from appearing. Of greater
interest are the studies where animals have been exposed to
cadmium, but where the dietary concentrations of zinc have
been just adequate or suboptimal. In such experiments it has
been shown that exposure to small amounts of cadmium may
cause a redistribution of zinc in the organism (Petering et
122
-------
al., 1971). Zinc concentrations increase in liver and kidney
and might decrease in other organs where cadmium does not
accumulate, e.g. testes. When the intake of zinc was increased,
these decreases in organs were not seen. It has also been
shown that if pregnant animals are exposed to cadmium, the
fetal concentrations of zinc as well as of copper may be
lowered (Pond and Walker, 1975; Choudhury et al. , 1977). It
could be expected that cadmium would cause changes in the
activities of some enzymes which require zinc. It has been
shown that the renal activity of leucine aminopeptidase,
which is a zinc-requiring enzyme, will be reduced in animals
exposed to cadmium (Cousins et al., 1973). It has, however,
not been shown that cadmium actually replaces zinc in any
enzyme in vivo.
That the increase in cadmium concentrations in the kidney
with age is accompanied by an equimolar increase in zinc
concentrations has been shown in studies on human beings
(Piscator and Lind, 1972; Blinder et al., 1977). In other
words, for each atom of cadmium that enters the kidney, one
atom of zinc will also be stored. There are indications that
above a certain level, corresponding to about 60 mg Cd/kg wet
weight in renal cortex, this equimolar increase of zinc
does not go on. A change in the cadmium-zinc relationships
at higher levels of cadmium is thus suggested. This might
be a factor of importance for development of renal damage
in cadmium-exposed people. This change in the equimolar
relationship has also been found in studies on horses
(Piscator, 1974; Blinder and Piscator, 1977). Since the
intake of zinc in human beings is generally regarded as
being only optimal, or in certain population groups
suboptimal, it is obvious that excessive exposure to cadmium
may cause changes in zinc metabolism and perhaps also a
lowering of zinc concentrations in some organ systems. There
might also be a risk for lower zinc concentrations in the
fetus if pregnant women are exposed to cadmium.
With regard to copper, several animal experiments, mainly on
the rat, have given data (Petering, 1974; Stonard and Webb,
123
-------
1976; Bremner and Campbell, 1978) which show that cadmium
can cause changes in its distribution. Thus, exposure to
cadmium caused increases in the renal concentrations of copper,
which could be attributed entirely to an increased content
of copper in the cadmium binding protein, metallothionein
(Stonard and Webb, 1976). It has been shown in animals that
the copper concentrations in the fetus may be reduced if
the mother is exposed to cadmium. At present there are no
data that show that cadmium has a profound influence on
copper metabolism in human beings (Bremner and Campbell, 1978).
In contrast to the zinc levels in kidney, which are related to
the cadmium levels, the copper levels in the human kidney
are fairly constant during a life-time and do not seem to be
related to the cadmium exposure (Schroeder et al., 1966;
Piscator and Lind, 1972; Tsuchiya and Iwao, 1978).
Cadmium can cause anemia, one mechanism for this action being
a decreased absoprtion of iron from the gut. This anemia
can be reversed by injections of iron compounds (Berlin and
Friberg, 1960). It has also been shown that iron deficiency
may increase the absorption of cadmium from the gut
(Hamilton and Valberg, 1974; Valberg et al., 1976). There are,
however, no indications of direct interactions between cadmium
and iron in the body. An indirect interaction is the change
in iron metabolism caused by hemolysis in cadmium-exposed
rabbits (Axelsson and Piscator, 1966).
Interactions between cadmium and calcium have attracted
great interest during later years, one reason being the
osteomalacia occurring in cadmium exposed women in Japan,
the so-called Itai-itai disease (see Friberg et al., 1974).
In animals on calcium deficient diets, the absorption of
cadmium will increase, and vice versa, i.e. in animals with
high dietary levels of calcium, the absorption of cadmium as
well as of several other metals will decrease. It has been
proposed that cadmium may be bound to the calcium-binding
protein which is responsible for the uptake of calcium from
the mucosal wall. With regard to systemic action, it has not
124
-------
been possible to show any direct effect of cadmium on bone
or on calcium metabolism. It has, however, been suggested
that at high renal concentrations of cadmium, cadmium might
interfere with the synthesis of vitamin D-. in kidney, thus
indirectly influencing bone metabolism.
With regard to interactions between cadmium and some other
toxic metals of interest, especially arsenic, lead, and
mercury, very few attempts have been made to study these
problems. It is known that mercury can bind to the cadmium-
binding protein metallothionein, and that interactions may
thus occur in the kidney between these two metals (Magos et
al., 1974). Mercury has a stronger binding capacity and
could displace cadmium from metallothionein. Lead does not
bind to metallothionein in vivo. There are no data showing
any direct interactions between lead and cadmium, or between
arsenic and cadmium. No data have yet been put forth which
would indicate any important relationships between these
metals in human beings even though these metals often occur
simultaneously in the environment.
In summary, the present data suggest that the most important
interactions in which cadmium is involved in human beings concern
zinc. Animal data suggest that important interactions may
occur between cadmium and copper in certain animals. It is
recommended that whenever cadmium is analyzed for evaluating
human exposure conditions, primarily zinc but also copper should
be analyzed.
2.3 Interactions between lead and other metals
A decrease in the activity of the enzyme ALA-D, which is a
zinc-requiring enzyme, in the presence of a high exposure to lead
has been shown in both humans and animals. Simultaneous
administration of zinc can prevent this decrease in activity.
Horses exposed to lead through industrial pollution showed
less severe responses when there was also excessive
exposure to zinc (Willoughby et al., 1972). Important
interactions, the most so in the gut, occur between lead ana
calcium, the similarity between these metals making bone the
125
-------
main storage organ for lead. Calcium deficiency has been
demonstrated to increase the absorption of lead in several
experiments. It has also been indicated in human studies
that especially people with high absorption of calcium, e.g.
children, have a higher absorption of lead (Alexander et al.,
1973). It may be noted that some of the neuro-muscular
actions of lead might be due to a direct interference with
binding sites usually occupied by calcium (Silbergeld et
al., 1974).
Iron deficiency has been shown to cause an increase in the
absorption of lead. Lead can cause anemia, but this is due
to the interference of lead with the synthesis of hemo-
globin and to some extent to intravascular destruction of
red cells. No direct interactions between lead and iron
in the body have been demonstrated.
There are no data on any interactions between lead and mercury..
With regard to interactions between lead and arsenic, reference
is made to the section on arsenic (section 2.1) where some
animal data are presented. In summary, the absorption of lead
ingested with food might be influenced to a great extent by
the calcium and iron status. There are also data indicating
important interactions between zinc and lead.
2.4 Interactions between mercury and its compounds and other
metals
It has been recognized that data are limited with regard to
the influence of metals and metalloids on the toxicity
of inorganic mercury and that definite conclusions
with regard to humans cannot be given (Task Group on Metal
Interaction, 1978) . Some interesting animal data (review:
Parizek, 1972; Ganther et al., 1973; Parizek et al., 1974;
Levander, 1978) will be mentioned in the following since
they may have possible relevance for humans as well.
126
-------
Selenite and other compounds that are metabolized to selenite
have been shown to reduce the lethal effect of inorganic
mercury in rats, preventing the development of renal tubular
and intestinal necrosis (Parizek and Ostadalova, 1967;
Parizek et al., 1971). Dietary selenate prevents the depression
of growth in rats receiving mercuric chloride by mouth
(Potter and Matrone, 1974). In long-term experiments dietary
selenate leduces the chronic renal tubular damage produced
by oral mercuric chloride in rats (Groth et al., 1973,
1976) .
Parizek et al. (1969, 1974) demonstrated that administration
of selenite gave rise to a marked increase in mercury in
blood and a reduced excretion of mercury. Both mercury and
selenium were bound to a plasma protein with an atomic
ratio of one (Burk et al., 1974). The formation of a
mercury-selenium-protein complex may explain these alterations
in the kinetics and toxicity of mercury (Parizek et al.,
1971, 1974, Parizek, 1976) and may also be related to the
observation by Groth et al. (1976) of intranuclear inclusion
bodies containing Hg and Se. The decreased passage of mercury
and selenium across the placenta and into milk when inorganic
mercury and selenite are administered simultaneously may also
be explained by the formation of an Hg-protein-Se-complex.
An indication that similar processes as those demonstrated
in animals might occur in man is provided by the observations
by Kosta et al. (1975) of elevated concentrations of mercury
and selenium in organs of mercury miners many years after
cessation of exposure.
The influence of other metals on inorganic mercury toxicity has
not been documented to any noteworthy degree in the literature.
One observation is that pretreatment of animals with cadmium
protected (male) rats from the nephrotoxic effect of inorganic
mercury (Magos et al., 1974). This action is regarded to be
mediated through an induction of metallothionein.
127
-------
With regard to organic mercury compounds, the Task Group on
Metal Interaction (1978) considered that the toxic effects
of methylmercury had been shown to be influenced by selenite
in several animal species. Thus 0.5-8,0 mg/kg of selenite
in the diet protected chicks, quail and rats from the toxic
effects of up to 0.2 mmoles/kg of methylmercury in the diet.
Although it has been suggested that selenium present in
marine fish might protect against the toxicity of dietary
methylmercury (Ganther et al., 1972; Ganther and Sunde
1974; Ohi et al., 1976) the evidence for this is inconclusive;
for instance the difference in toxicity could be explained
by differences in methylmercury intake, or in quantity ard
quality of protein in the diet (Stillings et al., 1974).
Dietary selenite and vitamin E diminished the mortality in
methylmercury exposed quail (Welsh and Soares, 1976). In rats
Welsh (1976) showed that vitamin E and other antioxidants had a
protective effect.
In a study of humans exposed to methylmercury, multiple
regression statistical analysis revealed an inhibition of
ALA-D in red cells related to concentrations of both mercury
and lead in blood (Schiitz and Skerfving, 1975) . The molar
effects of mercury and lead were similar.
On basis of the evidence available, the Task Group on Metal
Interaction (1978) found that selenite alters the kinetics
and reduces the neurotoxicity of methylmercury in several
animal species, at least at high levels of methylmercury
exposure. Methylmercury also alters selenium kinetics in
rats.
2.5 Interactions between copper and other metals
The influence of cadmium on copper metabolism has already been
mentioned. Another important relationship is that between
molybdenum and copper. This is of enormous importance for
ruminants where very small changes in the concentration of
any of these metals may cause disease. An excessive intake
128
-------
of molybdenum may cause copper deficiency. Data on human
beings are scarce but Kovalskit et al. (1961) have reported
that a high dietary intake of molybdenum caused increases in
the urinary excretion of copper and a lowering of blood
copper levels. In an experimental study on human beings it
was shown that an increase in the dietary intake of molybdenum
caused an increase in the excretion of copper (Deosthale and
Gopalan, 1974) . Animal data indicate that important relationships
might exist between copper and zinc. It has been proposed
that the zinc/copper ratio in the diet could be related to
cardio-vascular diseases in human beings (Klevay, 1975),
Oral therapy with zinc sulfate in large doses (150 mg Zn/day)
has recently been shown to cause copper deficiency in a
patient treated for coeliac disease (Porter et al., 1977).
2.6 Interactions between thallium and potassium
Thallium and potassium seem to have some common receptor sites
where they compete and thus undergo direct interaction. Potassium
has been proposed for use in the treatment of thallium poisoning
since the administration of potassium increases excretion of
thallium. For further data see Kazantzis (1977).
3. Interactions between metals and nutritional factors
In addition to the nutritional factors which have already been
mentioned, e.g. the intake of some metals, other nutrition
components might have a profound influence on metal metabolism.
It is well-known that phytate in certain diets causes a
reduction in the absorption of zinc and renders zinc less
available from vegetable diets than from diets based mainly
on meat. The protein content of the diet has been shown to
play a role in cadmium and lead toxicity, i.e. the less
protein, the higher absorption and the more severe the signs
of toxicity. Of special interest is that milk will increase the
absorption of several metals considerably. In neonatal and
adult animals Kostial and co-workers (e.g. Kello and
Kostial, 1973; Kostial et al., 1975), have found that
a milk diet causes a very high absorption of certain elements
like lead and cadmium and that the proportion of milk in the
129
-------
diet can be directly related to the absorption of certain
metals. The factor mainly responsible for this high absorption
has not yet been identified but the lipid content has been
put forth in several contexts as one such factor.
Among the vitamins, vitamin C has been shown to reduce lead
(Clegg and Rylands, 1966) and cadmium toxicity (Fox, 1975),
but on the other hand vitamin C increases the retention of
mercury in the tissues (Blackstone et al., 1974). Human data
are scarce, but it has been indicated that one reason for
the Itai-itai disease was low intakes of calcium, vitamin D,
and protein (Friberg et al., 1974).
4. Influence of drugs on metal metabolism and toxicity
Among the drugs taken by healthy people at large are oral
contraceptives. Intake of such synthetic hormones is known
to cause changes in copper metabolism, influencing the
synthesis of ceruloplasmin and bringing about higher serum
concentrations of copper. Zinc metabolism is also recognized
to be influenced by oral contraceptives. It has not been
established as to whether exposure to toxic metals like
lead and cadmium might need to be evaluated by different
dose-response curves in women taking such pills. Some of
the drugs for treatment of hypertension, which essentially
have a chelating action, also affect metal metabolism.
Prolonged treatment with sor.e of these drugs was noted to
increase zinc excretion, but not cadmium excretion (Wester,
1974) . Such treatment may thus in the long run cause changes in
the cadmium-zinc relationship which may be of importance especially
in people with a high body burden of cadmium. To what
extent antibiotics might change metabolism of both essential
and non-essential metals has not been documented in detail,
but one animal study has indicated that penicilline micrht cause
changes in the distribution of cadmiun (Deauidt et al., 1973).
5. Influence of other chemicals on metal toxicity
Ethyl alcohol has been shown to cause an increase in the
absorption of lead. This might be of weight when evaluating
130
-------
dose-effect relationships in people exposed to lead in
illicitly distilled alcohol. In cases of lead poisoning
caused by such consumption, the symptoms from the CNS seem
to be more severe than what could be expected from lead
exposure alone. Some synergistic effects between alcohol and
lead might thus be suspected.
Consumption of large amounts of alcohol can temporarily
depress ALAD activity, which is also a sign of lead toxicity.
However, it has also been shown that ALAD activity may
increase in animals exposed to both lead and alcohol. The
depression caused by either of the substances alone is
partially reversed when the two are combined (Moore, 1975).
Alcohol influences the metabolism of mercury after exposure
to mercury vapor. This is thought to be due to the inter-
ference of the alcohol with some of the enzymes taking part
in the conversion of mercury vapor to bivalent mercury
(Nielsen-Kudsk, 1965). Simultaneous exposure to cobalt and
alcohol has caused the so-called beerdrinker's cardio—
myopathy, where the effect was greater than what could be
expected from the single action of the compounds (see
Blinder and Friberg, 1977).
Exposure to 50 ppm carbon monoxide, a level not unusual in
industrial operations, caused changes in the distribution of
some essential metals in animal experiments (Mazaleski et al.,
1970). It is not known to what extent this might apply to
human beings and to exposure to toxic metals.
6. Influence of age on metal toxicity
It has long been claimed that certain age groups are subject
to special risks, i.e. young children and elderly. However,
not until recently have quantitative data been available to
document these claims. Several investigations have dealt with
young animals and children, whereas data on elderly people are
131
-------
still lacking. The fetus is regarded as specially vulnerable
since many functions are not fully developed until after
birth, and several possibilities exist; for contact between
the fetus and toxic metals. Especially methylmercury is
known to cross the placental barrier and accumulate in the
fetus. That fetal exposure to methylmercury can cause intrauterine
methylmercury poisoning is well documented. Lead crosses the
placental barrier, but it is still a subject of discussion
as to what extent pre-natal lead exposure plays a role in
later development. It is suspected that the fetal brain does
not tolerate lead to the same extent as does the adult
brain. The data available at present do not permit any
certain conclusions.
Extensive studies by Kostial and co-workers have shown that
the neonate animal has a higher absorption of certain
toxic metals like lead and cadmium due to its dependence on the
milk diet (see section 3). The neonate also has a higher whole
body retention, higher blood levels, and a much higher
accumulation in the brain compared to the adult exposed to
similar doses. Exposure to lead in the early neonatal stages
in animals can cause changes in emotional behavior, learning
deficits and increased motor-actovity. There are also data
suggesting that children develop symptoms at lower blood
lead concentrations and have a higher absorption of lead than
do adults. The behavioral effects noted in animals have also
been observed in children. With regard to the age group
10-20 years, there are not data available. Since the greatest
calorie intake is required during that period, it follows that
the highest amounts of e.g. cadmium will be accumulated at the
same time.
7. Influence of sex on metal toxicity
When women are discussed as a risk group what is often actually
meant is that through them the fetus may be exposed. The sex
differences to be discussed here concern only those directly
applicable to sex itself.
132
-------
Data from animal experiments are to some extent conflicting.
Lead acetate (Kostial et al., 1974) and cadmium chloride
(Engstrom and Nordberg, 1978) have been shown to have a
higher acute toxicity in male than in female animals. On the
other hand, female rats were shown to absorb twice as much
cadmium after a single oral dose as male rats (Kello et al.,
1978). Since castrated males absorbed as much as the females,
it was then concluded that male sex hormones were responsible
for the decreased retention of cadmium in males.
8. Influence of physical factors on _metal metabolism and
toxicity
Since sweat is an important excretion route for some metals,
e.g. zinc and nickel, it can be expected that in hot climates
or in hot industrial environments metal losses via sweat might
influence metabolism and toxicity. Workers exposed to
nickel in a hot environment had considerably lower nickel
levels in blood than expected (Szadkowski et al., 1969).
Zinc losses via sweat would be of special importance in
persons also exposed to cadmium.
High noise levels are common in many industrial environments and
prolonged exposure to noise might indirectly cause some
physiological changes which in turn might have an impact on
metal metabolism.
9. Influence of irritant gases and lung function on metal
metabolism and toxicity
The clearance of inhaled particles from the lung is dependent
on the intact function of the mucociliary transport system.
Long-term cigarette smoking and bacterial and viral diseases
can impair this transport system. A decreased clearance
of inhaled particles might increase the absorption of certain
metals or prolong their retention time in the lungs. It is
not known to what extent simultaneous exposure to irritant
gases, like SO9, and metals will affect the distribution and
133
-------
toxicity of these metals. Another part of the pulmonary
defence system consists of the alveolar macrophage, the
function of which might be depressed by exposure to nitrogen
dioxide. This might result in prolonged retention time and
thus a greater absorption of slowly soluble metallo-compounds
in the alveoli. The metals themselves might have an effect
on the macrophage function and thereby disturb the mucociliary
transport system.
Smoking may affect the uptake and metabolism of metals in
several ways. Long-term smoking might cause an impairment of the
mucociliary transport system, as said above. Carbon monoxide
might affect metal distribution, as also mentioned earlier. Since
the smoker will inhale large amounts of carbon monoxide this
could be another factor of importance. However, there are at
present no data available on this matter in human beings.
Finally it should be mentioned that the cigarette smoke in
itself contains certain metals of toxicological interest, e.g.
cadmium, nickel, mercury, lead, and thus smoking itself might
contribute to the metal burdens (see Volume II of this series
under the specific metals for details).
10. Influence of personal hygiene on metal metabolism and
toxicity
Both the industrial and the general environments present
great possibilities for contamination of hands and clothing
with metal compounds. In the industrial environment large
amounts of metallic dust will accumulate and may be
transferred to the human body through smoking, i.e. if
cigarettes or pipe tobacco are contaminated, or to the mouth
by direct contact with contaminated hands. Such circumstances
might cause a considerable extra exposure, making it extremely
difficult to arrive at dose-response relationships based on
measurements of the actual metal in air. In a study at a
Swedish cadmium factory, it was recently found that the
134
-------
cadmium levels in air measured by personal samplers varied
on an average between 2,ug and 10/ug/m in a group of 15
workers, while the fecal elimination of cadmium varied on an
average between 30 and 2 , 800 ,ug/day, indicating a considerable
exposure via the gastrointestinal tract in some cases (Adamsson
et al., to be published). The highest values were found
among smokers, and among people who were generally classified
as having bad personal hygiene habits in their work. Children
playing in areas with lead-containing dust or soil may
easily contaminate their hands and spread their exposure
considerably by licking them. Exposure to mercury vapor in
many laboratories, dentists' offices etc. can be avoided by
good hygiene and by care in handling mercury drops.
11. Conclusions
A multitude of factors must be taken into account when
evaluations of toxic hazards of metals are to be made. Every
human being has a special set of external and internal
factors which might influence the distribution and effects
of metals in the body. As far as essential metals are concerned,
homeostatic mechanisms in absorption and excretion maintain
them at their physiological levels regardless of relatively
large variations in exposure. External factors which are not
part of man's normal environment may cause profound changes,
however, a widespread example being the changes in metabolism
of certain essential metals caused by oral contraceptives.
The exposure to toxic metals, today often to several simultaneously,
in man is a result of great changes in the external environ-
ment due to industrial processes and to the general pollution
of the environment. The extent to which these metals can
interact with each other and, in the long run, cause changes
in each other's metabolism and toxicity is not known. To
which extent the combination of such metals with other
environmental pollutants, i.e. certain gases, can give other
dose-effect relationships than the expected is also a largely
unanswered question.
135
-------
REFERENCES
Alexander, F.W., Delves, H.T. and Clayton, B.E. (1973). In:
"Environmental Health Aspects of Lead." pp 319-331. Published
by the Commission of European Communities Directorate General
for Dissemination of Knowledge, Center for Information and
Documentation CID, Luxembourg.
Axelsson, B. and Piscator, M. (1966). Arch. Environ. Health
12., 374.
Berlin, M. and Friberg, L. (1960). Arch. Environ. Health !L,
478.
Blackstone, S., Hurley, R.J. and Hughes, R.E. (1974).
Food Cosmet. Toxicol. 12, 511-516.
Braman, R.S. and Foreback, C.C. (1973). Science 182, 2147.
Bremner, I. and Campbell, J.K. (1978). In: "Factors Influencing
Metabolism and Toxicity of Metals." (G.F. Nordberg, ed) Op. cit.
Choudhury, H., et al. (1977). Dietary cadmium: Embryotoxicity and
neonatal behavioral effects. Annual Report of the Center for the
Study of the Human Environment. University of Cincinnati,
Cincinnati.
Cousins, R.J., Barber, A.K. and Trout, J.R. (1973).
J. Nutr. 103, 964-972.
Deostahle, Y.G. and Gopalan, C. (1974). Brit. J. Nutr. 31,
351-355. —
Dequidt, J., Haguenoer, J.M. and Fromont, B. (1973). Arch.
Mai. Prof. Med. Trav. 34., 427-436.
Blinder, C-G. and Friberg, L. (1977). In: "Toxicology of
Metals." Volume II, pp 188-205. Environmental Health Research
Series, 600/1-77-022, Environmental Protection Agency,
Research Triangle Park.
Blinder, C.G. and Piscator, M. (1978) . In: "Factors Influencing
Metabolism and Toxicity of Metals." (G.F. Nordberg, ed) Op. cit.
Blinder, C.G., Piscator, M. and Linnman, L. (1977). Environ.
Res. 13, 432.
Engstrom, B. and Nordberg, G.F. (1978). Effects of detergen ;
formula chelating agents on the metabolism and toxicity of
cadmium in mice. Toxicology (In press).
Fowler, B. (1978) . In: "Factors Influencing Metabolism and
Toxicity of Metals." (G.F. Nordberg, ed) Op. cit.
Fox, M.R.S. (1975). Ann. N.Y. Acad. Sci. 258, 144.
Francis, A.J., Duxbury, J.M. and Alexander, M. (1974).
Appl. Microbiol. 28, 248.
136
-------
Friberg, L., Piscator, M., Nordberg, G. and Kjellstrom, T.
(1974). "Cadmium in the Environment." CRC Press, Cleveland.
Ganther, H.E. and Sunde, M.L. (1974). J. Food Sci. ^9, 1.
Ganther, H.E., Goudie, C., Sunde, M.L., Kopecky, M.J., Wagner, P.,
Sand-Hwan Oh and Hoekstra, W.G. (1972). Science 175, 1122-1124.
Ganther, H.E., Wagner, P.A., Sunde, M.L. and Hoekstra, W.G.
^1973) . In: "Trace Substances in Environmental Health - VI."
(D.D. Hemphill, ed) pp 247-252. University of Missouri,
Columbia.
Groth, D.H., Vignati, L., Lowry, L., MacKay, G. and Stokinger,
H.E. (1973). In: "Trace Substances in Environmental Health -
VI." (D.D. Hemphill, ed) pp 187-189. University of Missouri
Press, Columbia.
Groth, D.H., Stettler, L. and Mackay, G. (1976). In:
"Effects and Dose-Response Relationships of Toxic Metals."
(G.F. Nordberg, ed) pp b27-b43. Elsevier, Amsterdam.
Hamilton, D.L. and Valberg, L.S. (1974). Amer. J. Physiol.
227, 1033.
Kazantzis, G. (1977). In: "Toxicology of Metals." Volume II,
pp 388-404. Environmental Health Research Series, 600/1-77-022,
Environmental Protection Agency, Research Triangle Park.
Kello, D. and Kostial, K. (1973). Environ. Res. 6_, 355.
Kello, D., Dekanic, D. and Kostial, K. (1978). Influence of sex
and dietary calcium on intestinal cadmium absorption in rats.
Arch. Environ. Health (In press).
Klevay, L.M. (1975). Perspect Nutr. 28, 764.
Kosta, L., Byrne, A.R. and Zelenko, V. (1975). In: "Proceedings
of a CEC-EPA-WHO International Symposium on Renal Advances
in the Assessment of Health Effects of Environmental Pollution."
pp 245. Commission of the European Communities Directorate
General for Dissemination of Knowledge, Center for Information
and Documentation CID, Luxembourg.
Kostial, K., Maljkovic, T. and Jugo, S. (1974). Arch. Toxicol.
!31, 265.
Kostial, K. et al. (1975X In: "Proceedings of the 18th
International Congress on Occupational Health." p 290.
Kovalskii, V.V., Yarovaya, G.E. and Shmavonyan, D.M. (1961).
Z. Obsc. Biol. 22, 179 (In Russian; translation available
through the Department of Environmental Hygiene, the
Karolinska Institute, S-104 01 Stockholm 60, Sweden).
Levander, O.A. (1978). In: "Factors Influencing Metabolism
and Toxicity of Metals." (G.F. Nordberg, ed) Op. cit.
137
-------
Levander, O.A. and Baumann, C.A. (1966a).Toxicol. Appl.
Pharmacol. 9_, 98.
Levander, O.A. and Baumann, C.A. (1966b) Toxicol. Appl.
Pharmacol. 9_, 106.
Magos, L., Webb, M. and Butler, W.H. (1974). Brit. J. Exp.
Pathol. 55_, 589.
Mahaffey, K.R. and Fowler, B.A. (1977) . Environ. Health
Perspect. 19, 165.
Mazaleski, S.C., Coleman, R.L., Duncan, R.C. and Nau, C.A.
(1970). Amer. Ind. Hyg. Assoc. J. 3_1, 183-188.
Moore, W. (1975). In: "Proceedings of a CEC-EPA-WHO International
Symposium on RecentAdvances in the Assessment of Health Effects
of Environmental Pollution." pp 1171-1176. Commission of the
European Communities Directorate General for Dissemination of
Knowledge, Center for Information and Documentation CID,
Luxembourg.
NAS (1972). "Airborne Lead in Perspective." Committee on Medical
and Biological Effects of Environmental Pollution, National
Academy of Sciences, National Research Council, Washington, D.C.
NAS (1978). "Zinc." Committee on Medical and Biological Effects
of Environmental Pollution, National Academy of Sciences,
National Research Council, Washington, D.C.
Nielsen-Kudsk, F. (1965). Acta Pharmacol. 23, 263.
Nordberg, G.F., ed (1976). "Effect and Dose-Response Relation-
ships of Toxic Metals." Elsevier, Amsterdam.
Nordberg, G.F., ed (1978). "Factors Influencing Metabolism
and Toxicity of Metals." Environ. Health Perspect. (In press).
Obermeyer, B.D. et al. (1971). Appl. Pharmacol. 20, 135.
Ohi, G. et al. (1976). Environ. Res. 12, 49.
Parizek, J. (1957). J. Endocrinol. 15, 56.
Parizek, J. (1972) . In: "Nuclear Activabion Techniques in the
Life Sciences." I.A.E.A.-SM-157/82.
Parizek, J. (1976). In: "Effects and Dose-Response
Relationships of Toxic Metals." (G.F. Nordberg, ed) PP 4y«-siu.
Elsevier, Amsterdam.
Parizek, J. and Ostadalova, I. (1967). Experientia 23, 142.
Parizek, J., Benes, I., Ostadalova, I., Babicky, A., Benes, J.
and Lener, J. (1969). Physiol. Bohemoslov. 18, 95-103.
138
-------
Parizek, J., Ostadalova, I., Kalouskova, J., Babicky, A. and
Benes, J. (1971). In: "Newer Trace Elements in Nutrition."
(W. Mertz and W.E. Cornatzer, eds)Pp 85-122. Marcel Dekker,
New York.
Parizek, J., Kalouskova, J., Babicky, A., Benes, J. and Pavlik, L,
(1974). In: "Trace Element Metabolism in Animals - 2." (W.G.
Hoekstra, J.W. Suttie, H. Ganther and W. Mertz, eds) PP liy-131.
University Park Press, Baltimore.
Petering, F.G. (1974). In: "Trace Element Metabolism in Animals
- 2." (W.G. Hoekstra, J.W. Suttie, H. Ganther and W. Mertz, eds)
p 311. University Park Press, Baltimore.
Petering, H.G., Johnson, M.A. and Stemmer, K.L. (1971).
Arch. Environ. Health 23, 93.
Piscator, M. (1974) . In: "Proceedings of a CEC-EPA-WHO
International Symposium on Recent Advances in the Assessment
of Health Effects of Environmental Pollution." pp 951.
Commission of the European Communities Directorate General
for Dissemination of Knowledge, Center for Information and
Documentation CID, Luxembourg.
Piscator, M. and Lind, B. (1972). Arch. Environ. Health 24, 426.
Pond, W.G. and Walker, E.F. (1975) Proc. Soc. Exp. Biol. Med.
148,665.
Porter, K.G., McMaster, D. Elmes, M.E. and Love, A.H.G.
(1977) . Lancet 2, 774.
Potter, S. and Matrone, G. (1974). J. Nutr. 104, 638.
Schroeder, H.A. et al. (1966). J. Chron. Dis. 19, 1007.
Schiitz, A. and Skerfving, S. (1975). Scand. J. Work Environ.
Health 1, 54.
Silbergeld, E.K., Fales, J.T. and Goldberg, A.M. (1974).
Nature 247 ,49.
Stonard, M.D. and Webb, M. (1976). Chem-Biol. Interac. 15, 349.
Stillings, B.R. et al. (1974). Toxicol. Appl. Pharmacol. 30,
243. —
Szadkowski, D., Kohler, G. and Lehnert, G. (1969). Arztl.
Forsch. 23, 271.
139
-------
Task Group on Metal Interaction (1978), In: "Factors Influencing
Metabolism and Toxicity of Metals." (G.F. Nordberg, ed) Op. cit.
Tsuchiya, K. and Iwao, G. (1978). In: "Factors Influencing
Metabolism and Toxicity of Metals." (G.F. Nordberg, ed) Op. cit.
Underwood, E.J. (1977). "Trace Elements in Human and Animal
Nutrition." Academic Press, New York.
Valberg, L.S., Sorbie, Y. and Hamilton, D.L. (1976). Amer. J.
Physiol. 231, 462.
Welsh, S.O. (1976). Fed. Proc. 35, 761.
Welsh, S.O. and Scares, J.H., Jr. (1976). Nutr. Rep. Int.
13_, 43.
Wester, P.O. (1974). Acta Med. Scand. 194, 505.
Whanger, P.O. (1976). In: "Selenium-tellurium in the
Environment." p 234. Industrial Health Foundation, Pittsburgh.
Willoughby, R.A., MacDonald, E., McSherry, B.J. and Brown, G.
(1972). Can. J. Comp. Med. 36, 348.
140
-------
EPIDEMIOLOGICAL ASPECTS OF ASSESSMENT OF DOSE-RESPONSE AND
DOSE-EFFECT RELATIONSHIPS
Tord Kjellstrom
1. The epidemiological approach
With the term epidemiology we mean the science of the occurr-
ence of health and illness in human populations. In clinical
medicine the individual and changes in the individual's
health status constitute the focal point. In epidemiology
the interest is extended to groups of individuals.
Epidemiological studies are essential for assessing dose-
response or dose-effect relationships for toxic metals,
because of the difference in the metabolism and toxicity of
most metals between animals and humans. Results of animal
experiments can seldom be used directly for quantitative
conclusions about human effects and responses.
One of the main concepts of epidemiology is the population
at risk. If, for instance, congenital malformations are
under study they can only occur among newborn children and
these newborn children could be defined as the population at
risk. If spontaneous abortions are under study, the women
who may become pregnant constitute the population at risk.
In studies of metal toxicology the population at risk would
further be limited to the group that was exposed to the
metal. For example, in New Zealand there have been on the
average about 4 cases of occupational lead poisoning re-
ported every year during the 1970's (NZ Department of Health,
annual reports). In a total population of 3 million (and
765,000 males in the age group 20-64, in whom most lead
poisoning cases occur) this is not a common condition or a
major public health problem. However, when we consider that
the number of workers exposed to lead (the population at
risk) was only 1740 on an average during the 1970's (NZ
Department of Health, annual reports), the picture is dif-
141
-------
ferent (Fig. 1). The incidence of reported lead poisoning in
the population at risk was 2.3 cases (1000 person-years).
This is similar to the incidence of hospitalization for
cancer (2.4 cases/1000 person-years) and higher than the
incidence of hospitalization for myocardial infarction and
coronary heart disease (1.1 cases/1000 person-years) (NZDH,
1975).
The occurrence of disease or toxic effects is measured
either as incidence, which is the number of newly developed
cases during a defined time period, or as prevalence, which
is the number of existing cases at a defLned point of time.
Incidence and prevalence are calculated as rates in the
population at risk (e.g., number of cases per 1000 persons
at risk). Depending on the type of disease or effect that is
measured, the time unit used for incidence rates may vary.
The most common for chronic effects is years. For acute
effects days or even hours may be more appropriate. The time
unit is included in the unit for incidence rate (e.g.,
cases/1000 person-years). The distinction between incidence
and prevalence is important in studies of metal toxicology,
because whenever the occurrence of effects is measured
strictly in terms of prevalence or incidence, the population
at risk, the time unit for effect measurements, etc., have
to be more clearly defined than in studies where occurrence
is measured only as proportion or percentage of an exposed
group. There is room for a lot of improvement in this area.
Even reputable journals like the Bulletin of the WHO publish
articles on metal toxicology where the term incidence is
used for prevalence data (Al-Mufti et al., 1976).
Incidence and prevalence rates are interdependent because
naturally the number of existing cases depends on how many
new cases occur. The duration of the disease and the time
unit for incidence are the factors on which this interde-
pendence is based. A low incidence may give a high preva-
lence if the duration of the disease is longer than the
time unit for incidence. If the time unit for incidence is
142
-------
made very long, the calculation should be corrected for the
increasing number of prevalent cases that can not get the
effect again. The correction is made by calculating the
cumulative incidence rate (Miettinen, 1977). This measure is
particularly valuable in studies of the role of toxic metal
exposure in the etiology of chronic diseases like cancer.
Then, we are often interested in the cumulative incidence
over a 30- or 40-year period and need to take into considera-
tion increasing incidence rates with age. The cumulative
incidence in the population at risk can for an individual be
described as the risk (or probability) of getting the disease.
The 40-year cumulative incidence rate for lead poisoning in
New Zealand would be about 2/10,000 persons. The 40-year
risk for an average lead exposed worker to get lead poison-
ing would be 0.01, or 1%.
In mortality and cancer-morbidity studies the life-table
method for calculating individual risks has been used (Lemen
et al., 1976; see section 5). From death registers and
cancer registers the age-specific average incidence rates of
dying from a particular disease or of getting cancer are
calculated for a whole country or province. For each indi-
vidual in the population at risk, these rates are used to
estimate the risk of dying or getting cancer within the time
period under study. The expected number of_ cases is calcu-
lated as the sum of the individual risks and this number is
compared with the observed number of case_s to calculate the
risk ratio and the risk difference (Miettinen, 1977).
Epidemiological studies for assessing dose-effect and dose-
response relationships of metal toxicity have the advantage
that we are usually dealing with only one specific exposure
- that to the metal compound. The problems lie in the measure-
ment of dose. It is often necessary to estimate long-term
doses due to long half-times of the metal compounds in the
critical organ. This may be quite difficult (see section
4.2) because direct measurements are seldom possible. On the
other hand, many of the toxic metal compounds are persistent
143
-------
in nature and this makes retrospective studies of e.g.
museum specimens of food possible.
Measurements of effects may pose no problems particular to
metal toxicity studies (see section 4.1) and the measure-
ments of relatively high response rates can be made in
studies of limited size. From the standpoint of prevention
of effects it may be of greatest interest to measure very
low response rates (e.g. 1 case/1000 persons). This neces-
sitates large scale studies that are not always feasible
because of cost or because the population at risk is not
large enough. This is a feature of most environmental health
epidemiology studies, but as will be discussed in the next
section, so-called association studies make it possible to
overcome this problem.
2. Observation versus association
Dose-effect and dose-response relationships for humans can
be ascertained by epidemiological studies in basically two
different ways. Observation studies involve measurement of
average dose in groups within the population at risk and
measurement of occurrence of effects and response in the
same groups. Association studies involve the use of meta-
bolic models, statistical distributions of the occurrence of
the toxic metal in tissues or the daily intake, as well as
estimates of critical concentrations in the critical organ
to calculate the proportions of groups in the population at
risk that are likely to have the effect (Figs. 2 and 3).
Observation studies are the most straightforward and they do
not involve making assumptions about the metabolism of the
metal compound, which may insert uncertainties in the dose-
response evaluations. The disadvantage with this type of
study is that it is often difficult to measure low response
rates of effects directly when they are not routinely re-
corded. Exceptions are, for instance, deaths or cancers
where these effects can often be ascertained from existing
records.
144
-------
There are many examples of dose-effect and dose-response
evaluations of toxic metals using observation studies.
Different degrees of effects of lead on hemoglobin synthesis
(Roels et al., 1975) as well as different degrees of renal
tubular damage by cadmium (Kjellstrom et al., 1977) have
been studied in this way. A combined dose-effect and dose-
response evaluation of neurological effects of methylmercury
(Fig. 4) based on observations on 93 persons in Iraq (Bakir
et al., 1973) is one of the most detailed studies of this
type available, but still no observation data on response
rates under 5% were recorded.
Association studies may be used as an extension of data from
observation studies into parts of the dose-response rela-
tionship where no observation data exist or as a complete
substitute for observation studies. The metabolic model is
the key. The models may be one-compartmental as the one for
methylmercury proposed by Berglund et al. (1971) or may take
on one of many degrees of complexity. The model is a way of
linking together data on metal intake, metal concentrations
in different tissues and metal excretion. Data from groups
with relatively low exposure levels can be used to quantify
rates of transfer between compartments in the model. If it
can be assumed that these rates are independent of the
exposure levels (within the range of interest) critical
organ concentrations under particular exposure conditions
can be calculated and anticipated. Individual variation in
dose within the population at risk as well as the individual
variation of half-times of the metal compound in the main
body compartments can be expressed as statistical distribu-
tions. In the case of cadmium, for instance, it has been
shown that both daily intake and renal cortex cadmium con-
centrations follow log-normal distributions, which coincide
remarkably well with calculations using a metabolic model
(Fig. 5; Kjellstrom, 1977). The distribution in renal cortex
can be used to calculate how large a proportion of the popu-
lation with a particular average renal cortex cadmium concen-
tration would have concentrations above a certain level.
145
-------
The critical organ concentration is not defined as a thre-
shold level above which pathological changes necessarily
occur in the critical organ (Task Group on Metal Toxicity,
1976) in all individuals. For most metals it can be assumed
that there is a considerable individual variation in the
sensitivity of the critical organ. However, the best esti-
mate of the metal concentration in the critical organ that
is necessary to cause the critical effect would be the cri-
tical organ concentration.
By using metabolic models based on epidemiological data on
metal intake, tissue distribution and excretion, as well as
statistical distributions and estimates of critical organ
concentration, the proportion of the population at risk who
get the critical effect (=response rate) can be calculated.
This can be done before human effects occur, which is an
advantage with association studies. By definition observa-
tion studies can only be done when effects can actually be
observed. Association studies have the disadvantage that the
findings must always be judged as approximate due to the
need for assumptions and due to limitations of the data
collected. These drawbacks also apply when association
studies are used to expand dose-response relationships docu-
mented by observation studies down to lower response rates,
because slight errors in the middle part of the dose-response
curve can become quite large at the extremes. Even so, such
calculated dose-response relationships as those for methyl-
mercury (Nordberg and Strangert, 1976) and for cadmium
(Kjellstrom, 1977) can be of value to indicate what response
rates can be expected at different doses. This can be used
to decide on sample size in observation studies and to
decide on provisional hygienic standards for primary preven-
tion.
3. Study design
Epidemiological studies can be of three types. Descriptive
studies provide data on the occurrence of exposures, dose or
effects but give no direct comparison between dose and
146
-------
effects or dose and response. Studies of metal intake via
food in a population or of the incidence of stomach cancer
in different age and sex groups would be descriptive studies.
Analytical studies put data on exposures or doses together
with data on effects or responses in order to estimate the
causative role for a particular exposure. Intervention
studies involve manipulation (or intervention) of metal
exposure in a randomized group in the population and follow-
up of the influence on the occurrence of effects. Interven-
tion studies are in a sense equivalent to human experiments
and can only be used for the study of reversible effects
that are not injurious to the persons exposed.
Analytical studies are the most common type of observation
studies for assessment of dose-effect or dose-response rela-
tionships. Descriptive studies supply basic data for asso-
ciation studies.
The direction in which data are collected is of importance
in analytical studies. The study can focus on persons who
have the effect under study and compare them with a refer-
ence group without the effect, or it can focus on persons
with the metal exposure and compare them with people without
exposure. The first type of study is called a case-reference
study (Table 1). The cases and non-cases are compared as to
their metal exposures in the past, or as to the present
metal dose based on analysis of indicator tissues. Higher
doses in the group of cases would be interpreted as the
metal being associated with this effect. Absolute occurrence
can not be measured with a case-reference study alone, but
the relative occurrence (risk ratio) between a high dose
group and a low dose group can be calculated. If both the
cases and the referents are stratified into groups according
to metal dose, the risk ratio for each strata could be
calculated and these ratios could be expressed as a dose-
response relationship.
In order to measure the absolute incidence or prevalence
rates in different dose groups, which must be done to
147
-------
establish complete dose-effect or dose-response relation-
ships, a follow-up study would be more suitable. In this
type of study an exposed group and a non-exposed group
(referents) are selected. The occurrence of effects of the
metal is measured in each group. The exposed group can be
stratified into groups according to dose level and the
measurement of incidence or prevalence of effects in each
group forms the basis for dose-effect and dose-response
relationships.
If there is a good indicator of dose, like for instance
blood-lead for inorganic lead exposure (see Tsuchiya, 1977)
or hair-mercury for methylmercury exposure (see Berlin,
1977), the data on dose and effects may be collected at the
same time (cross-sectional study). Usually the data have to
be collected longitudinally in time because of lack of good
indicator material and because of changes of exposure with
time which may not be reflected in the indicator material.
The data can be collected either retrospectively or prospec-
tively. Prospective data collection can of course be done in
a more standardized, elaborate and controlled manner. When
the half-time in the critical organ is very long, as for
cadmium (see Friberg et al., 1977) , retrospective exposure
estimates are essential for arriving at results reasonably
quickly.
One of the most important steps in an epidemiological study
is the selection of the groups to be studied. In both descrip-
tive and analytical studies it should be made certain that
the people in the "exposed group" have really been exposed
and belonged to the population at risk. Ideally the whole
group that is and has been exposed should be studied or, it
it is too large, a sample from this group should be randomly
selected. In an industrial situation this means that all
workers who ever were exposed in the factory under study
should be included. An operational definition for "exposed"
may be used such that an exposure duration limit is set
(e.g. all workers who worked in the metal exposed area of
148
-------
the factory one year or longer). If only workers who are
employed at the time of the study are included, those with
the effect may be selectively excluded, because they had to
stop working when they got the effect. In a dose-response
study of cadmium-induced tubular proteinuria (Kjellstrom,
1977) it was shown that the possible underestimation of
response due to selective exclusion may be considerable.
Similar bias may occur in studies based on population
screenings where only those healthy enough to come to the
local school can participate. Selection bias may also be
introduced in studies of industrial workers by the procedure
of pre-employment selection of workers. Mainly healthy young
men would be selected for the work that involves metal
exposure and the results of such a study may not be app-
licable to the general population. This is called the
"healthy worker effect" and it is a problem in e.g. retro-
spective follow-up studies of mortality (Kitagawa and Hansen,
1973).
The selection of the reference group in analytical studies
is equally important as the selection of the exposed group.
Selection bias caused by e.g. sick people at home not being
able to participate, should be avoided. If any sampling of
an exposed group takes place, the same criteria for sampling
should be used for the reference group. Confounding factors
can interfere with the analysis. A confounding factor is
something that is associated with metal exposure and also
associated with the effect; e.g. if the exposed group in a
study of molybdenum and gout had a greater proportion of
heavy drinkers than the reference group and heavy drinking
was associated with gout, the drinking habit would be a
confounding factor. Confounding can be controlled already at
the stage of selection of target groups for the study by
restriction of the target group to a sub-group of the popula-
tion at risk. If drinking is a confounding factor, the study
can be limited to non-drinking people only. If sex is a
confounding factor, the study could be limited to women
only, etc. Confounding can also be controlled for at the
analysis stage of the study (see section 5).
149
-------
4. Data collection
4.1 Measurement of effect and response
Effects can be classified into subjective symptoms, social
or behavioral changes, biochemical or physiological changes,
clinical disease, and death. Subjective symptoms can be e.g.
the shortness of breath associated with beryllium exposure
(Reeves, 1977} or the weakness, fatigue and anorexia asso-
ciated with mercury vapor exposure (Berlin, 1977). In this
group of effects we can also include subjective psychologi-
cal and mood changes and annoyance reactions to dust, mucous
membrane or skin irritation and smell. Data may be collected
via standardized interviews or self-administered question-
naires. When interviews are carried out it is important to
avoid observer bias caused e.g. by having different inter-
viewers in the exposed group and the reference group. The
way questions are asked may affect the answers. Ideally the
interviews should be blind, which means that the interviewer
should not know to which group a particular person belonged.
A method for collection of subjective symptom data that has
been used in air pollution effect studies is the panel
method. Groups of people in an area with anticipated varia-
tions in exposure to air pollutants are asked to keep a
diary of symptoms as they occur each day. During the same
time period exposure levels are measured and any association
between exposure and effects is analyzed. The panel method
is only useful for studies of acute effects.
With social and behavioral effects we may mean family strife
caused by behavioral changes in the affected person, de-
creased work efficiency, lagging behind in school work,
absenteeism, and workmen's compensation or other insurance
claims. The latter two are usually based on diagnosed clini-
cal disease but this is not always the case and absenteeism
may be an unspecific "annoyance" effect of metal exposures
in the workplace. The data collection can be carried out in
a similar way as for subjective symptoms or by using exist-
ing records of school work results, insurance claims, etc.
Using this method it was found that high arsenic exposure
150
-------
during early life affected the children's school results 15
years later (Ohira and Aoyama, 1973). The effects mentioned
here are quite unspecific and the risk of confounding may be
greater than for other effects.
Critical and sub-critical effects of metal compounds are
probably of the type biochemical and physiological changes
but only for s few metals have they been established, e.g.
decreased erythrocyte ALA-D activity after lead exposure
(see Tsuchiya, 1977) and increased urinary 3 --microglobulin
excretion after cadmium exposure (see Friberg et al., 1977).
Histopathological changes may also be included in this type
of effect. Measurement problems are related to finding the
specific type of change caused by a metal compound. Animal
studies can be of great value here, but species differences
have been shown for instance for the type of proteins ex-
creted in urine as an effect of cadmium poisoning. Another
problem is to be able to take samples of the tissue where
the effect occurs. The biochemical change we measure may be
an indirect measure of the effect, like cadmium-induced
urinary (3,.,-microglobulin excretion, or a direct measure,
like lead-induced ALA-D inhibition.
When the effect is measured as changes in the concentrations
of the metal compound normally occurring in tissue or in
physiological variables, the dose-effect relationship can be
evaluated from the quantitative increase in the variable with
dose. For calculations of dose-response relationships it is
necessary to arrive at an operational definition of what
level of the variable constitutes an effect. The distribu-
tion of the variable in the reference group may be used to
calculate a certain percentile as the "cut-off" level. When
urinary concentrations of specific compounds are used for
measuring effects it is necessary to correct the individual
concentrations for variations in urinary dilutions. When
compounds in whole blood are measured it may be necessary to
correct for hematocrit.
151
-------
Another problem with biochemical, physiological and histo-
pathological changes as measures of effects is whether the
changes measured should be integrated as adaptation to the
metal or just physiological variations influenced by the
metal. We do not have detailed knowledge about all the che-
mical steps in toxicity from the first changes to the severe
clinical disease for any single metal.. Any change that
occurs in the critical organ that may be related to more
severe effects in this organ should be considered at least
as a sub-critical effect (Task Group on Metal Toxicity,
1976) .
Effects of metals can also be measured as clinically diag-
nosed disease. Doctors' records, hospital records, data
registers like a cancer register may be used as a source of
data. Such records or registers may not be complete, how-
ever, and this may cause response to be underestimated.
There may be factors that cause a biased selection of the
people in the records or the register, like economic bar-
riers or social class dependent attitudes to medical care. A
complete clinical examination of the whole group under study
may overcome such shortcomings of existing records. Cancer
registers or congenital malformation registers are becoming
more and more complete and accurate in industrialized coun-
tries where reporting systems are good and autopsy rates
high. Cancers with a relatively short and characteristic
course, like lung cancer or breast cancer, are more com-
pletely recorded than e.g. prostatic cancers which occur at
older ages and may be overshadowed by other diseases that
cause death. When this type of effect is measured it is just as
necessary as in other cases to collect data for the exposed
group and the reference group in an identical way. If a
cancer register is used for calculating national reference
rates for a "life-table study" (see section 5) all cancer
cases in the exposed group should also be picked up via the
cancer register. It must also be considered whether the fact
that the exposed group (e.g. metal exposed factory workers)
was exposed to carcinogenic metal compounds in itself makes
152
-------
it more likely for their cancer to be diagnosed and recorded
in the cancer register.
Finally, the type of effect that is most easily defined,
death, has been used in some studies as an indicator of
long-term prognosis after metal exposure (Cooper and Gaffey,
1975). After very high exposures to metal compounds death
can be one of the main effects (Bakir et al., 1973). Depend-
ing on the character of the recording system and autopsy
rates, data on causes of death vary in quality. Mortality
studies need long-term follow-up which poses problems of
tracing the place of death of the target group and problems
of identification. Case-reference studies of mortality
involve finding data on exposures of people who can not be
interviewed. The major difficulty in mortality studies is
that individual dose data have to be linked up with the
effect data since the effect itself cannot be measured.
When the target group is well defined and the effect has
been measured accurately the measurement of response is
straightforward: the proportion of the group who gets the
effect in a specific time period (incidence response) or the
proportion who has the effect at a specific point in time
(prevalence response). Selection bias (see section 3) would
be the most likely cause for errors in response measurement.
4.2 Measurement of dose
Ideally the metal dose should be measured in the tissue
where the metal elicits its effect (Task Group on Metal
Toxicity, 1976). In many cases this is not possible and
indirect measurements of varying degrees of accuracy must be
made instead (Fig. 6). The critical organ is often an inter-
nal organ that is not directly accessible for metal measure-
ments. At autopsy, samples may be collected and analyzed and
in some cases biopsy specimens have been used (e.g. renal
biopsies from cadmium workers; see Axelsson and Piscator,
1971). Autopsy data may be compared with earlier data on
effects (Nomiyama, 1973). For certain metals the recently
153
-------
developed in vivo neutron activation technique (Harvey et
al., 1976) seems promising. No epidemiological studies using
this technique have yet been published.
Indicator tissues like blood (Pb), urine (Cd), hair (methyl-
Hg) or nails (As) can be used to estimate the dose in the
critical organ. In order to translate indicator tissue con-
centrations into critical organ dose, metabolic models (see
section 2) are useful. Feces amount of a metal is a good in-
dicator of daily intake via ingestion when the gastroin-
testinal absorption is low. Comparison between cadmium body
burden and daily fecal and urinary amounts of cadmium shows
a good agreement between body burden and urine cadmium on a
group basis after long-term low level exposure. Body burden
and kidney burden have also been shown to be well correlated
(Kjellstrom, 1977). Hair mercury is a good indicator of body
burden of methylmercury (Friberg and Vostal, 1972) and
because of the relatively even distribution of methylmercury
in the body (see Berlin, 1977) it is also a good indicator
of the amount of methylmercury in the brain. Hair may give
us retrospective dose data because the metals stay in the
hair as it grows out. It was shown in the methylmercury
poisoning outbreak in Iraq (Bakir et al., 1973) that sectional
analysis of hair could pinpoint the time when mercury consump-
tion started and the time when it reached its maximum (Fig.
7) .
The measurement of dose by analyzing indicator tissues or
the critical organ itself involves chemical analysis of the
metal compound in the tissue. The concentrations are usually
very low, which makes accurate analysis difficult. The
problems of chemical analysis have been highlighted by
interlaboratory comparisons (Berlin et al., 1975) where
differences up to two orders of magnitude were found. In
order to make the data reliable they should always be sup-
ported by appropriate method studies, preferably including
comparisons between two completely different types of analy-
sis (like atomic absorption spectrophotometry and neutron
activation analysis).
154
-------
The dose may be estimated from exposure data if uptake
kinetics and half-time in critical organ are known. The
product of exposure level and exposure duration corrected
for the ratio between duration and half-time and also for
the absorption rate, etc., can give an estimate of dose. The
calculation of dose should be based on the exposure that is
relevant for uptake. Measurements of exposure to metals via
air should include an estimate of particle size distribution
and the metal concentration in each size fraction. The
absorption of metals in the intestines can be affected by
other elements in the diet (see e.g. Friberg et al., 1977;
Tsuchiya, 1977). Therefore, additional information on diet
content would be necessary to calculate dose from exposure
levels.
Most metals, because they are elements, occur naturally in
air, food and drinking water. Whatever particular exposure
conditions are under study we usually have to consider the
total exposure by adding all possible uptakes. Skin exposure
may be significant in occupational environments.
Smoking can increase exposure both through the natural
content of metals in cigarettes and pipe tobacco and through
external contamination of the smoking gear from hands,
overalls or the workroom air (Piscator et al., 1976). A
smoker may of course be sensitive to airborne metal expo-
sures in other ways. Smoking is associated with chronic
bronchitis and emphysema and it slows down lung clearance
(Albert et al., 1969).
We can assume that for certain effects of certain metals the
dose may not be a function of exposure level times the ex-
posure duration, but rather a function of the frequency of
peak exposures above a certain level (impulse exposures).
When carcinogenic effects oi netals are studied, the dose is
calculated in three different ways depending on the assumed
mechanisms for carcinogenicity (Task Group on Air Pollution
and Cancer, 1977).
155
-------
5. Data analysis
This is the stage in an epidemiologicai study where the data
are arranged so that conclusions can be drawn. In descrip-
tive studies the conclusion would be in the form of a des-
cription of the occurrence of a variable in the population
group studied and in an analytical or Intervention study it
would be in the form of rejecting or accepting a hypothesis
about an association between an exposure and an effect. If
the study had been designed for ascertaining a dose-effect
or dose-response relationship, the conclusion may be that
this was or was not found. In drawing these conclusions a
number of standard statistical techniques are used.
Quantitative dose data can be expressed as distributions
(Fig. 8) of individual doses. From the distribution it can
be calculated how large a proportion of a population can be
expected to have a dose above levels associated with certain
effects. We can also express the dose as population dose,
which means the sum of all individual doses in a population.
This is the approach often employed in radiation protection
and basically it assumes a linear dose-response relationship
down to very low doses. A certain population dose is assumed
to be associated with a certain number of cases with effect.
If the dose data are in the form of environmental measure-
ments of emissions, it may also be relevant to estimate the
dose commitment. This is the estimated total future popula-
tion dose that will arise from the emissions.
Quantitative effect data can also be expressed as distribu-
tions. A difference between the distributions is most easily
ascertained with the Kolmogorov-Smirnoffs test.
When the effects are measured as the number of cases of a
particular disease (e.g. primary liver cancer), and inci-
dence data based on a total population are available, the
life-table technique can be used to estimate incidence
ratios between exposed group and total population. A start- .
ing point in time for the calculation is defined. This may
156
-------
be when the persons started exposure, 20 years after they
started exposure, or some other particular date (like the
day when the cancer register became operational). For each
exposed individual the age of the starting point and the
calendar year of the starting point are used to find, in the
register data, the incidence of the disease in the total
population for that age group at that time. The incidence
value is interpreted as that person's risk of developing the
disease that particular year. By adding up all the indivi-
dual risks for each year after the starting point, the
yearly expected number of cases is calculated. This is
compared with the actual observed number to yield the inci-
dence ratio (observed cases/expected cases).
In case-reference studies a Venn-diagram may be useful for
figuring out the incidence ratios (Fig. 9). When more than
one exposure variable are evaluated at the same time (e.g.
cadmium and smoking exposure and local effects on the lung)
a Venn-diagram may show clearly what type of additional data
is needed when some data are already available.
The shape of the dose-response curve is often sigmoidal like
a cumulative normal distribution. Such a curve can be trans-
formed to a straight line by probit-transformation. When the
effect is not specific for the metal exposure but also
occurs at a low level in a non-exposed population, the
response rates should be adjusted before the probit-transfor-
mation. This and other aspects of probit analysis of dose-
response data for toxic metals have been discussed by Kjell-
strom (1976). In the case of cancer, it has now been pro-
posed (Task Group on Air Pollution and Cancer, 1977) that at
low doses the dose-response curve can be approximated by a
straight line. The implication is that there is no safe
threshold for carcinogenicity. This has great influence on
the assessment of hygienic standards to protect people from
cancer-inducing metals. If the cancer-risk at dose A is
1/1000 then the risk would be 1/100,000 at dose A/100, which
may be a common dose in large population groups, and public
health protection measures may be needed.
157
-------
REFERENCES
Albert, R.E., Lippman, M. and Briscoe, W. (1969). Arch.
Environ. Health 18, 738-755.
Al-Mufti, A.W., Copplestone, J.F., Kazantzis, G., et al.
(1976). Bull. WHO 53_, (suppl.), 23-57.
Axelsson, B. and Piscator, M. (1971). Unpublished data. In:
"Cadmium in the Environment." (L. Friberg, M. Piscator and
G. Nordberg) 1st edition. CRC Press, Cleveland.
Bakir, F., Damluji, S.F., Amin-Zaki, L., Murtadha, M.,
Khalidi, A., Al-Rawi, N.Y., Tikriti, S., Dhahir, H.I.,
Clarkson, T.W., Smith, J.C. and Doherty, R.A. (1973).
Science 181, 230-241.
Berglund, F., Berlin, M., Birke, G., Cederlof, R., von
Euler, U., Friberg, L., Holmstedt, B., Jonsson, E., Luning,
K.C., Ramel, C., Skerfving, S., Swensson, A. and Tejning,
S., (1971). Nord. Hyg. Tidskr., Suppl. 4.
Berlin, M. (1977). In:"Toxicology of Metals." Volume II,
pp 301-344. Environmental Health Research Series, 600/1-
77-022, Environmental Protection Agency, Research Triangle
Park.
Berlin, A., Lauwerys, R., Buchet, J.P., Roels, H., Del
Castilho, P. and Smeets, J. (1975) . In: "Proceedings of a
CEC-EPA-WHO International Symposium on Recent Advances in
the Assessment of Health Effects of Environmental Pollution."
pp 2185-2213. Commission of the European Communities Direc-
torate General for Dissemination of Knowledge, Center for
Information and Documentation CID, Luxembourg.
Cooper, W.C. and Gaffey, W.R. (1975). J. Occup. Med. 17,
100-107.
158
-------
Friberg, L. and Vostal, J. (1972). "Mercury in the Environ-
ment." CRC Press, Cleveland.
Friberg, L., Nordberg, G. and Piscator, M. (1977). In:
"Toxicology of Metals." Volume II, pp 124-163. Environmental
Health Research Series, 600/1-77-022, Environmental Protec-
tion Agency, Research Triangle Park.
Harvey, T.C., McLellan, J.S., Thomas, J.J., et al. (1976).
2nd East Kilbride Conference on Progress and problems of in
vivo activation analysis. Scottish University Research and
Reactor Centre, Glasgow.
Kitagawa, G.M. and Hansen, P.M. (1973). "Differential Morta-
lity in the United States." Harvard University Press, Boston.
Kjellstrom, T. (1976). In: Nordberg, G.F., ed. "Effects and
Dose-response Relationships of Toxic Metals." pp. 147-156.
Elsevier, Amsterdam.
Kjellstrom, T. (1977). Accumulation and renal effects of
cadmium in man. A dose-response study. Diss., Karolinska
Institute, Stockholm.
Kjellstrom, T., Evrin, P.-E. and Rahnster, B. (1977). Environ.
Res. 13, 303-317.
Lemen, R.A., Lee, J.S., Wagoner, J.K and Blejer, H.P. (1976).
Ann. NY Acad. Sci. 271, 273-279.
Miettinen, O.S. (1977). "Principles of Epidemiologic Research."
Duxburg Press, Massachusetts (In press).
Nomiyama, K. (1973). Jap. J. Hyg. 28, 45 (In Japanese).
Nordberg, G.F. and Strangert, P. (1976). In: Nordberg, G.F.,
ed., "Effects and Dose-response Relationship of Toxic Metals."
pp. 273-282. Elsevier, Amsterdam.
159
-------
NZDH (1975). Trends in health and health care. New Zealand
Department of Health.
Ohira, M. and Aoyama, H. (1973) . Jap. J. Hyg. 27, 500-531
(In Japanese).
Piscator, M., Kjellstrom, T. and Lind, B. (1976). Lancet 2_,
587.
Reeves, A. (1977). In: "Toxicology of Metals." Volume II, pp
85-109. Environmental Health Research Series, 600/1-77-022,
Environmental Protection Agency, Research Triangle Park.
Roels, H.A., Lauwerys, R.R., Buchet, J.P. and Vrelust, M.T.
(1975). Int. Arch. Arbeitsmed. 34_, 97-108.
Task Group on Air Pollution and Cancer - Risk Assessment
Methodology and Epidemiological Evidence. (1978) . Report
from an International Symposium at the Karolinska Institute,
Stockholm, 1977. Environ. Health Perspect. (In press, 1978).
Task Group on Metal Toxicity (1976). In: Nordberg, G.F.,
ed., "Effects and Dose-response Relationships of Toxic
Metals." pp. 7-114. Elsevier, Amsterdam.
Tsuchiya, K. (1977). In: "Toxicology of Metals." Volume II,
pp 342-400. Environmental Health Research Series, 600/1-77-
022, Environmental Protection Agency, Research Triangle
Park.
160
-------
Table 1. Classification of epidemiological studies
Type: Descriptive (just listing observations)
Analytical (testing associations between exposure
and effects)
Intervention (the "experiment")
Direction;Case-reference (start from cases with effect and
probe into their exposures)
Follow-up (start from exposed people and see if
they get the effect)
Time: Cross-sectional (all data collected referring to
one point in time)
Longitudinal:
Retrospective (data collected regarding the
present and the past)
Prospective (data collected regarding the present
and the future)
161
-------
Definition
of
Population
at risk
well
poisoned
at risk
Figure 1. Diagram indicating the relevance of defining a
population at risk. Without such a definition
the response estimate will be "diluted".
162
-------
Response
100%
50%
D Dose
Direct measurement of D and R
Figure 2. Observation studies. The target population is
stratified into groups according to dose. The
mean dose and the response are measured for each
strata giving one point each on the dose-response
curve.
163
-------
Concentration in
indicator tissue
Concentration in
critical organ
Frequency
Concentration
in indicator
tissue
Half time
in critical
organ
JjCalculatlon
Response
100%
50%
-•-Dose
Figure 3. Association studies. Epidemiological data on human
metabolism of the metal and distribution of half-
times, etc., are used to establish a metabolic
model. The model calculates the proportion of a
group with a particular dose that would have metal
concentrations in critical organs above an estimated
critical concentration. This is used as an estimate
of response rate in each dose group.
164
-------
• PARESTHES1A
• ATAXIA
A DYSARTHR1A
O DEAFNESS
+ DEATH
6 16 40 78 156 312
Estimated body burden of mercury
(mg)
Figure 4. Dose-response relationship for different symptoms
and signs in the poisoning incident in Iraq. Dose
estimates from blood mercury levels.
A = at the time of onset of symptoms;
B = at the time of cessation of exposure
(from Bakir et al., 1973).
165
-------
CADMIUM IN KIDNEY CORTEX
INTAKE
ug Cd/day
Figure 5. Relationship between distribution of daily cadmium
intake (B) and distribution of cadmium concentra-
tions in kidney cortex in the age group 30-59 in
Sweden (A). • = calculated relationship from a
metabolic model (C) (from Kjellstrom, 1977).
166
-------
Metal concentration in environmental
indicator material (metal in soil,
plants, etc.) I
Metal concentration in direct
exposure - related environmental
material (metal in ambient air, food,
drinking water, etc.)
I
Actual average metal exposure to a
population.
I
Individual metal exposure
I
Individual metal retention
4
Metal concentration in transport tissue
or intermediate tissue
Metal concentra-
tion in index *"
tissue
Average metal concentration in critical
organ
Metal concentration in critical cell
I
Response
Figure 6. Sequential relationship between dose variables and
a specific response (from Kjellstrom, 1976).
167
-------
700
3600
a
• TOTAL Ha
n METHYL
A INORGANIC
2 4
6 8 10 12 0 2 4 6 8 10 12
Distance from root (cm)
Figure 7. Concentrations of mercury (methylmercury, in-
organic mercury and total) in hair sections from
two patients with methylmercury poisoning in Iraq
(from Bakir et al., 1973).
168
-------
SWEDEN
Prdbits
7
Figure 8. Distribution of cadmium doses measured as urine,
muscle, liver or kidney cortex concentrations
(from Kjellstrom, to be published).
169
-------
Figure 9. Venn-diagram of data from a case reference study,
e.g. P = population, S = smokers, C = cancer
cases, A = smoking cancer cases. Data from the
study = S/P and A/C. The risk ratio for cancer
among smokers as compared to whole population is
A/S = A/C
C/P S/P'
170
-------
GENERAL ASPECTS OF THE PREVENTION OF METAL POISONING
Sven Hernberg
1. Goals: Ideals and practical restrictions
There are many definitions of poisoning, the narrowest
requiring the presence of overt disease, and the widest
including any functional deviation from normality. From the
preventive point of view the concept of poisoning should be
regarded in its widest sense. The ultimate goal should be
an environment in which exposures to toxic metals have been
reduced to levels at which no functional impairment is
caused, either in the present population or in future gene-
rations .
In practice the ultimate goal is far from reached. The main
obstacles are economical and technological, for example, the
removal of metals already dispersed in the environment, the
design of completely safe work processes or the efficient
sanitation of old plants. Hence operative goals must often
differ from the ideal, and the preventive strategy chosen is
not always the most efficient.
In the prophylaxis of metal poisoning, the assumption is
often made that a no-effect level exists for the toxic
action of metals. This assumption implies that some exposure
can be tolerated, and it has resulted in the endorsement of
several different types of health criteria. Some standards
are not determined by health aspects only, but also from the
beginning take into consideration economical and techno-
logical feasibility. Criteria for occupational exposure
sometimes even accept some deterioration from normal health.
Some types of effects, such as the lead-induced inhibition
171
-------
of erythrocyte delta-aminolevulinic acid dehydratase, are
presently thought to be acceptable, even though they occur
in a large proportion of the persons exposed. Others are
accepted if only a small portion exhibit them. For example,
in the documentation of the ACGIH Threshold Limit Values it
is stated that "nearly" all workers may be repeatedly exposed
without adverse effect (ACGIH, 1974; also the chapter on
standards and criteria in this volume).
The acceptability or non-acceptability of functional impair-
ment is a delicate matter indeed, and the criteria vary
according to who defines them and to the circumstances under
which they should be applied. A good illustration is the
well-known dualism that exists between most standards for
the general as opposed to the work environment. Efforts are
being made in several countries, at considerable cost, to
maintain the upper range of lead in blood at 40 or even 30
ug/100 ml for the general population, mostly by banning the
addition of lead to gasoline. Still, it is persistently
claimed that industrial workers and their unborn children
can easily tolerate levels of 70 or even 80 iig/100 ml.
2. General principles of prevention
In all prevention there are some common principles. First,
the hazard must be recognized and the routes and modes of
exposure determined. This knowledge may be readily avail-
able or it may require a vast amount of research. The recog-
nition of the dangers of discharging mercury into the water
system is a good example of a difficult problem. It was in
Japan, where after several years of confusion, the so-called
Minamata disease was recognized as being due to the ingestion
of seafood contaminated by methylmercury (Japanese Ministry
of Health and Welfare, 1967). Since discharges often include
only inorganic mercury, the next step was to demonstrate
172
-------
that metallic mercury is methylated in an aquatic environment
(Jensen and Jernelov, 1969). It was also discovered that
methylmercury is enriched in food chains (Friberg and Vostal,
1972). Because of these findings large surveys of the
content of methylmercury in fish were initiated in Sweden,
and later in other countries. Concomitantly, estimations of
the highest safe daily intake were agreed upon. All of this
activity made it possible to localize the hazard and to take
administrative action. The most important measures initiated
were the strict control of the discharge of mercury into
water systems and a ban on the sale of fish of prey from
contaminated waters, together with recommendations for the
maximal weekly intake of fish.
In many cases prevention is simple as soon as the hazard is
recognized. For example, leaded gasoline is often (mis)used
as a solvent, especially in small industries. Prevention of
lead poisoning is extremely simple: substitution with lead-
free gasoline or some other solvent.
Some general guidelines that can help identify occupational
hazards will be given later.
The second principle of prevention lies, of course, in
the elimination of the hazard, if possible. Generally
speaking, the following procedures should be undertaken:
- cessation of unnecessary use of the toxic metal,
- substitution with less toxic agents,
- prevention of the spread of dust and fumes,
removal of the metal already spread, and
reduction of personal contact with the metal.
As long as exposure levels are potentially hazardous, or
may become so, monitoring the environment and the exposed
population (or some indicator group), the third principle,
is warranted.
173
-------
Most, if not all, preventive efforts fail if the professional
staff is not well trained. Training, the fourth principle,
should not be restricted to health personnel and engineers.
Both the factory workers and the general public should have
basic knowledge of how to protect themselves, and should be
made aware of the importance of control measures. Securing
both is an important task for those responsible for the
health care of a plant or a community. For example, if
lead oxide dust is generated in a restricted area of a factory
and its elimination is based on effective local exhaust
ventilation, the workers should know how to prevent it from
spreading to other parts of the plant. Similarly, in order
for house-to-house campaigns aimed at eliminating child-
hood lead poisoning due to the ingestion of flaking paint
to succeed, health officers must win the cooperation of the
families concerned.
Most programs, at least large-scale ones, should be backed
by some form of authority, this being the fifth principle.
Usually such authority is gained through legislation,
sometimes by agreements between, e.g. labor and management.
Before such measures are taken, there may be a need to
create political pressure. Those engaged in research must not
withdraw into an ivory tower; on the contrary, their task
is to provide politicians with unbiased data and opinions
based on scientific judgement. They should also offer the
public correct facts and fight against any form of exaggera-
tion and misrepresentation.
3. The work environment
3.1 General considerations
In the work environment, the main route of exposure is through
inhalation, and hence preventive measures should be concentra-
ted upon reducing the levels of airborne fumes and dusts.
In this respect the general principles of industrial hygiene
apply. Since there are several good textbooks on this topic
174
-------
(e.g. Patty, 1962; Olishifski and McElroy, 1971), only a
brief recapitulation, with emphasis on the aspects typical
of metals, is given here.
Recognition of hazards. Before any action can be taken,
the hazard must be recognized. Many countries require the
labeling of toxic products and a declaration of their
composition. In some instances this regulation may help.
No simple rule of thumb exists for a general classification
of occupations, workplaces or work phases as "dangerous"
and "less dangerous". On the contrary, the hazard is de-
termined by the many factors that influence the concentra-
tions of respirable metal dust or fumes in the air. First,
the process itself is important. All high processing tempera-
tures, extensive fume, dust or aerosol formation, and pri-
mitive work methods increase the risk; and low temperatures,
low-level dust formation, automation, and encapsulation
promote safety. Secondly, efficient local and general
ventilation and the proper prevention of the spread of
dust reduce the risk while a lack of such measures increases
it. Thirdly, the general level of workplace hygiene is
important. Considerations of this type mean that workplaces
must be looked at individually as far as risk evaluation
is concerned. If any general rule could be given, it would
be that an accumulation of negative factors is more usual
in small, primitive plants than in large, modern ones.
Elimination of unnecessary use. In some instances the pro-
duction of toxic compounds or the use of the toxic metal
itself in a process is unnecessary. Some brass castings
should not be allowed at all in primitive foundries, since
automation eliminates the risk. When planning new chlorine
plants, the diaphragm method should be considered since it
does not call for any mercury at all.
Substitution. Because metals are fundamental to our civili-
zation, substitution of other materials for metals cannot
be widely applied. This has been possible as far as mercury
175
-------
nitrate, which was used for more than two centuries in
making felt for hats, is concerned. This procedure resulted
in outbreaks of severe poisoning and in such expressions as
"mad as a hatter". Mercury nitrate has now been replaced by
less toxic chemicals in most countries, although some exceptions
are still said to occur (Hamilton and Hardy, 1974). The use
of lead arsenate as an insecticide has largely been abandoned.
Red lead or litharge (Pb.,0 .) , once the most important anti-
corrosive paint in shipbuilding, is increasingly being
replaced by less toxic paints (Ziegfeld, 1964), and white
lead (carbonate) has been avoided, sometimes due to laws, in
indoor painting for decades. In Finland, legislation prohibit-
ing such use was passed as early as 1929 (Hernberg, 1973) .
In the pottery industry, the introduction of insoluble
fritted lead polysilicates as glazes has somewhat diminished
the exposure.
On the other hand, substituting the lead sheet of cables
with a plastic covering has increased the risks involved in
their manufacture, since the new material (PVC plastic) is
produced with lead stearate used as a stabilizer. Mixing
this substance into the raw material has proven to be more
hazardous than smelting metallic lead (Tola et al., 1976).
In the vast majority of operations, metals or their compounds
are not easily replaced, and other methods of prevention
must be applied.
Technical control measures. The technical control of metal
exposure follows the principles applied in industrial hygiene
in general. Accordingly success is usually the greatest
when the technical and medical knowledge we have today can
be used in the planning stages. Old plants may be difficult
to remodel satisfactorily. If possible, all hazardous types
of exposure should be concentrated in special areas where
control processes can be effectively utilized. Encapsulation
and other isolation methods prevent toxic dust and fumes
from being spread about while mechanization and automation
reduce the number of workers exposed. Furthermore, special
176
-------
well-ventilated control rooms can be built within the plant.
Surfaces should be plane and smooth in order to prevent
dust from attaching to walls and floors. Nooks and corners
should be avoided. For example, in the case of metallic
mercury, the worst problem is evaporation from spilled drop-
lets. Their removal is facilitated by smooth, unabsorbing
surfaces which slant towards the center of the floor. Secondary
lead smelting is such a hazardous process that priority should
be given to automation of transportation and oven filling
phases. More reduction of the levels of lead in the working
environment can be accomplished if the ovens can be emptied in
the open air. The preparation of all metal pigments, very
hazardous when done by hand, is far less so in closed, automatic
systems.
Local exhaust ventilation. One of the best means for
controlling dust or fumes that cannot otherwise be prevented
from escaping into the air is a local exhaust system. The
air must then be filtered or the metals otherwise removed so
that the general environment will not be polluted. There is
a persistent belief that the exhaust should be from below since
metal dust particles in general are heavy. However, the
respirable fraction is made up of small particles capable of
remaining airborne for long periods and behaving like a gas
cloud. Hence, to be efficient, the exhaust should be placed
very close to the source of pollution. The system for removal
of the polluted air must be constructed so as not to draw
the air into the breathing zone of the worker on its way to
the exhaust.
General ventilation. A good general ventilation is an
adjuvant to a local exhaust system. Since the general
systems involve a much higher cost, e.g., in energy consumption,
the emphasis should be put on the local ones. One important
exception is the elimination of metallic mercury vapors.
Because the droplets are usually spread over the floor and
because evaporation will be reduced by lowering the work
room temperature, general ventilation should be well developed.
It should be stressed that open windows and doors as well as
177
-------
changes in the work process may thwart an otherwise ef-
fective ventilation system. Self-made "improvements" and
extensions may be deleterious, and the ventilation, filtra-
tion and precipitation equipment may deteriorate.
Other methods of elimination. Since most metals occur in
dust form or precipitate as small particles, exposure can be
reduced by keeping the floor permanently moistened, and by
frequent vacuum-cleaning. Sweeping should be avoided, as
this operation returns settled dust to the workroom air. If
sweeping cannot be avoided, oiled sawdust should first be
spread on the floor to bind the dust. The use of compressed
air to blow dust from floors, walls and ledges should be
prohibited. In principle, all cleaning or other operations
that may cause settled dust to whirl about should be per-
formed outside the work shift by well-protected, special
workers.
When moistening the floor, the water should not come in
contact with molten metal, since violent explosions would
occur. Operations involving the welding, flame-cutting or
burning of metals coated with paint of unknown composition
should not be allowed before samples have been analyzed for
at least lead, cadmium and chromates, or the composition
otherwise elucidated. If such metal pigments are present,
the burning is extremely dangerous and should be performed
under powerful local exhaust and general ventilation, with
safety equipment, and, if possible, outside the normal work
shift. Metals coated with metallic cadmium should be
treated in the same way. Even less toxic metals, such as
zinc and copper, cause metal fume fever, and welding and
burning such coatings demand the same protective measures.
Housekeeping. Poor housekeeping can render even the best
1echnical control system ineffective. Education and super-
vision should be detailed, efficient and followed up and
motivation should be kept at a high level. Smoking can
never be allowed in the presence of metal dust. Clothing
178
-------
made of synthetic fibers retains less dust than do cotton
overalls. Pockets should be as few as possible, especially
when metallic mercury is present in the work room. Work
clothing should be vacuumed before removal and changed at
least twice a week. Work and street clothing should be
stored in different lockers to reduce the exposure of the
worker himself and to prevent him from exposing his family.
Because of the danger of food becoming contaminated from the
environment or from dusty hands or clothing, it should not
be prepared, dispensed or eaten in areas where metal dust
occurs. Clean dining rooms should obviously be available.
Personal hygiene should be facilitated by providing suffi-
cient washrooms, and requiring that they be used. This is
important also from the point of view of reducing family
exposure.
The worker should spend as little time as possible in areas
with unusually high concentrations of metal dust or fumes.
These areas include the vicinity of smelting furnaces,
welding places, pasting machines in storage battery fac-
tories, etc. When engaged in welding or burning outdoors,
the worker should always place himself upwind, and other
workers should not be allowed to work downwind of the welding
or burning operation.
Reduction of worker contact with toxic metals. If the
large-scale engineering methods such as encapsulation fail
to ensure enough safety, personal protective equipment,
mostly respirators, must be used, but only as a last resort.
The only situations where respirators can be considered a
primary choice are in emergencies or in occasional, brief
exposures above the threshold limit value. Many types of
filter respirators are on the market, but not all of them
are adequate as far as meeting requirements outlined in for
example the American National Standard for Respiratory
Protection Z 88.2-1969 is concerned. Metal fumes require
more efficient respirators than dust; if both are present,
179
-------
those for fumes are recommended. In certain operations
where exposure is very high, for example, when ships covered
with thick layers of red lead are scrapped by flame-cuttinq,
the only acceptable type of protection is the air-purifying pos-
itive-pressure respirator (NIOSH, 1972). In addition to other prev-
entive measures, reduction of exposure time either by shortened
daily exposure time or by worker rotation every few weeks or
months is necessary e.g. often in shipbreaking and secondary
lead smelting, and sometimes in chlorine alkali plants
(where the problem is exposure to metallic mercury vapors).
For metals, such as cadmium, which accumulate to a high
degree in the body a shortened lifetime exposure may even be
considered, whereby workers who have accumulated large
amounts of cadmium should be taken from exposure for good,
regardless of whether or not they exhibit symptoms.
3.2 Monitoring of the work environment
Measurements of concentrations of metals in the work
environment are mandatory for a preventive program in
order to:
localize high exposure areas
estimate individual exposure (especially when no
satisfactory biological monitoring method exists)
detect engineering failures (leaks, insufficient
ventilation, etc.)
detect the effects of changes in processes or work
methods
document the impact of preventive measures
- ensure that safety standards are being met.
Sampling strategy. Concentrations close to a machine or
process are maximal. They may vary considerably over a
distance of a few meters or even centimeters, especially
when the sources of contamination are few such as in the
case of a smelting furnace. For most purposes sampling
should be performed in regular work areas and from the
worker's breathing zone.
180
-------
The number of air samples needed depends upon the purpose
of the measurement. An overall picture is obtained from
sampling carried out on a long-term basis and from
different sites. When individual exposure is to be studied,
workers who are representative for each area, operation and
other special conditions should be chosen. Especially those
with predicted high exposure should be considered. The
alternative is to use some random sampling procedure, especially
when there is little advance knowledge about the exposure.
Air sampling may be performed at regular intervals, or
randomly over time. Regular intervals must not coincide with
other regular cycles of events that may cause systematic
variations in the measured concentrations.
Short-term samples are preferred for intermittent exposure,
as for example in welding operations. Long-term samples
(7 to 8 hours) are the most relevant when one primarily
wants to ensure that the metal concentrations on the average
are below the threshold limit value. Because long-term
particulate sampling often has to be restricted to about 10
samples or less for practical reasons (e.g. 8 hours/one
sample), their average should be well below the threshold
limit value if a safety margin is to be ensured for variations.
Personal monitoring is of course the method of choice when
individual exposure is to be estimated as far as air sampling
is concerned. The samples should be taken from the breathing
zone, no farther than 40 cm from the nose. It may even be
taken from within a welding helmet. The sampler is usually
attached to the worker's shoulder. No obstruction can be
allowed between the air sampler inlet and the worker's
face, and the equipment must not disturb his work, e.g.
by blocking his sight. Compared to biological samples,
personal monitoring may give a much less correct picture of
exposure unless strict precautions and standardization are
enforced (Williams et al., 1968; Chatterjee et al., 1969;
Williams et al., 1969; Williams, 1975). For example, the
181
-------
angle of the inlet is important, as well as its distance
from the nose. Furthermore, personal sampling does not allow
for differences in physical activity during exposure,
personal hygiene, use of respirators, or in other exposure
(from food, water, etc.). When the metal is cumulative,
exposure during one day is not very informative from the
point of view of assessing the health risk. Hence, biological
monitoring is always preferable whenever reliable methods
exist. Such is definitely the case for lead, partly the case
for mercury and perhaps for cadmium and arsenic. For other
metals, knowledge on the relationship between the
concentrations in ambient air and those ir. biological media
is still insufficient.
Sampling technique. Metals usually appear in the workroom
air as fumes and solid particles, sometimes as vapor
(metallic mercury) or liquid aerosols (chromates above the
electroplating bath). Fumes are formed when molten metal
evaporates or when metal or metal compounds are burnt. The
dust particles are either derived from condensed fumes or
from metal compound dust that has whirled up. Although the
particle size distribution varies, the respirable dust is
composed of small particles that can stay in the air for
many hours. Larger particles settle but may be whirled up
again.
During sampling a measured volume of air is drawn through a
filter or impinger device, with the use of an electrostatic or
thermal precipitator, a cyclone, or some other instrument which
collects particles. Meters for direct-reading of total dust
or mercury vapor concentrations are also available.
The volume of the air sample may vary from a few liters to
several cubic meters, depending on the concentration and
the sampling equipment. It must contain enough particles to
be accurately weighed or analyzed. Often the concentrations,
especially those around or below the threshold limit value,
are so low that a large air volume and high flow rate are
182
-------
needed before enough metal particles can be collected
during a reasonable period of time. The sample size
should be large enough so that accurate estimations of
concentrations within the magnitude of about a tenth of
the threshold limit value can be made.
Analysis. The concentration of the particulates in the air
is calculated from the weight of the particulates collected
per unit volume of the air sample, determined by direct
weighing and/or chemical analysis. The particulates in
a known portion of the sample are counted for the deter-
mination of the number of particles collected. This pro-
cedure is accomplished by microscopy or by automatic
particle counters. Since particles in the respirable
range are the most important from the toxicological point
of view, the size distribution should also be determined
and taken into consideration. The size distribution is
determined with microscopy, liquid sedimentation or,
during the sampling, with cyclones.
Interpretation. The most important sampling problem is
to ensure the representativeness of the sample. Results
obtained during one day, or from a few sites only, cannot
be given too much weight. Monitoring must be regular
if reliability is to be ensured, and plotting the results
of successive measurements on a control chart is recommended.
Needless to say, the analytical methods used should be strictly
standardized and quality control programs should be employed
both within the laboratory and between laboratories.
Only a broad outline of monitoring has been presented here.
For more detailed data the reader is referred to specific
textbooks or monographs on the topic (e.g. Patty, 1962;
Fairhall, 1969; Rossano, 1969; Olishifski and McElroy,
1971; Roach, 1973; Linch, 1974) as well as to the chapter on
analytical methods and sampling in the present volume.
183
-------
4. General environment
Many metals are essential to life in small amounts, but
toxic in higher doses. Others, such as lead, mercury,
and cadmium, probably lack physiological function in man.
Since all metals occur naturally in the earth's crust,
an environment completely without exposure has never
existed. However, civilization, especially during
this century, has caused a sharp increase in the concentration
of many metals in the air, soil and water (e.g., Patterson,
1970). The main sources are industrial emissions, traffic,
burning of fossil fuels, and the weathering of metallic
constructions and paint. Moreover, man-made products
containing toxic metals as such, or in the form of alloys
and compounds, are coming into the market in ever increasing
numbers. Because of the wide use of such products, the
likelihood that the general public will be exposed to toxic
metals is strong.
Exposure levels vary greatly, depending on geographical,
cultural and other circumstances (e.g. WHO, 1973). For
example, urban environments are more polluted than rural
ones, and industrialization increases exposure directly
through emissions and indirectly through its products.
Local habits, for instance, the use of lead-glazed ceramic
containers, may further cause exceptionally high exposure
(NAS, 1972). Finally, enrichment may take place in bio-
logical chains, such as in the case oE mercury (Friberg and
Vostal, 1972).
Preventive measures as far as environmental exposure is
concerned have two goals, namely the prevention of poison-
ing in the presently living population and the reduction
of metal accumulation in the environment so that future
generations are also protected. The risk for poisoning in
the presently living population is obviously highest in
184
-------
connection with massive exposure from air, food or
drinking water. The Minamata and Niigata episodes are
perhaps the best known, but even worse mass poisonings
have occurred. In 1972 there was a severe outbreak of
methylmercury poisoning in Iraq. According to official
sources, 459 patients died, the total number of hospitalized
patients being 6530 (Bakir et al., 1973). In the United
States it is estimated that several thousand children each
year contract lead poisoning from ingesting flaking paint
(Moore, 1970; Guinee, 1971). Old water pipes made of lead
and tanks painted with lead oxide may mean a risk for
poisoning from drinking water. This has been reported in
a district of Glasgow (Beattie, 1972), and probably occurs
elsewhere as well.
There is so far no evidence that leaded gasoline causes
poisoning in the general population (NAS, 1972). However,
because the practice results in continuously increasing
dispersion and accumulation in the entire environment,
authorities in several countries have made moves to reduce
or ban the use of organic lead as an additive in order to
protect future generations.
Generally speaking, it is extremely difficult or even
impossible by man-made means to eliminate metals that
have already become spread. On the other hand there are
many ways to treat current pollution problems. Because of
the manifoldness of both exposure and the methods for
reducing exposure outlets, only broad guidelines for
preventing metal poisoning from the general environment can
be given.
Metals already spread. In some very exceptional cases it is
possible to remove metals already dispersed in the environment.
There are large campaigns in the United States, in some states
backed by law, aimed at removing lead paint from dwellings,
185
-------
either through the reconstruction of whole areas, or through
urban renewal (NAS, 1972). Of course, the lead is not
completely removed from the environment, since it is diffi-
cult to avoid transportation to rivers and lakes through
sewage systems. However, the most dangerous hazard, namely,
that of children ingesting flaking paint, can be eliminated.
Old lead plumbing systems can be reconstructed with safer
materials. Though, aside from a few cases, it is generally
not possible for man to influence significantly the distri-
bution pattern of metals already dispersed, natural processes
may do so. Rainfall removes metals from the air, and sedi-
mentation redistributes metals in water systems (NAS, 1972;
Friberg and Vostal, 1972; Friberg et al., 1974).
Source control. The most effective method of control is of
course the discontinuation of dangerous and dispensable use.
Such a step also reduces the need for production and hence
eliminates the exposure of the worker. Generally, such a
drastic measure requires a legislative ban, as in the case
of the use of methylmercury compounds for seed dressing in
Sweden in 1966 (Abramson, 1967) . Sometimes a practice is
abandoned voluntarily, as in Finland in 1969 when the pulp
industry agreed to discontinue the use of phenylmercury
acetate for slime control. Another measure would be to
recover the metals for reuse as soon as they escaped from
industrial processes. This is seldom economically attractive,
however.
In general, then, control is based on the prevention of
escape at the source of pollution. In principle, there are
two types of escape, namely, "economical," i.e., metals are
emitted during refining, smelting or industrial use, and
"noneconomical " i.e., impurities are released during a
process. "Economical" escape into the air occurs from smoke
stacks, exhaust ventilation systems and, to a lesser extent,
186
-------
from whirled-up dust. Hence the most important method of
preventing such emissions is the purification of smoke and
exhaust. The methods vary immensely, but the following four
principles are fundamental to most methods of particulate
control (Elvingson, 1972) :
separation by gravity forces (e.g., cyclones),
separation by dissolution (e.g., water scrubbers),
separation by filters, and
separation by electrostatic powers (electrofilters).
In addition cooling traps and absorption into organic
polysulfides are used for capturing mercury vapor.
The aquatic system becomes contaminated when industries
discharge waste directly into it (especially small factories
often dispose their sewage water into community sewers
without any purification), when rainwater washes away the
metal-containing dust that has settled around industries,
and when metal trapped in air purification systems is carelessly
handled. Water scrubbers are a particular problem in the
last respect. The purification of sewage water is based on
recirculation and various precipitation processes. For
example, mercury can be precipitated into sulfide and absorbed
into aluminum hydroxide, which is then removed as a foam.
Iron oxide can also be used. Unfortunately, the control of
the spread of dust from the surroundings of industries
sometimes entails a liberal use of water, thereby placing an
extra load on water cleaning systems and in some cases,
polluting rivers and lakes.
The most important "noneconomical" escape of metals into the
air takes place during the burning of fossil fuels, e.g.
contributing twice as much mercury to the atmosphere as do
emissions from industry (MacGregor, 1975). Large conven-
tional power plants are the main polluters. The largest
power station in Czechoslovakia uses coal containing 1,530 mg
187
-------
of arsenic compounds per kilogram. About 138 tons of
arsenic enter the atmosphere each year from this single
source (Cmarko and Rosival, 1970). Home heating also causes
metal emissions.
Small amounts of some toxic metal occurring together with a
less toxic one may undergo a "noneconomical" escape into the
environment. For example, cadmium is an Lmpurity in materials
containing zinc. Since the demands for controlling zinc
emissions are more liberal than those for cadmium, excessive
amounts of cadmium easily escape.
Even when the air is the first environmental medium to
become polluted, both soil and water become polluted secon-
darily through dustfall or washout by rain. Around the
Czech power plant mentioned above, high concentrations of
arsenic (up to 0.21 mg/1) have been measured in surface
water (Cmarko and Rosival, 1970).
Pollution is usually the greatest around industries and
other major polluters. Hence the risk for adverse effects
is the highest among those living in their vicinities.
Taller chimney stacks would distribute the emissions over
larger areas and thereby dilute the concentrations, but this
practice is no substitute for efficient exhaust purification
in the case of toxic metals and should therefore be condemned.
The same can be said of the practice of leading unpurified
sewage water containing e.g. mercury or cadmium far out into
the sea.
While it is possible practically to control emissions from
a large source of pollution, such as a power plant, puri-
fying the smoke from single homes is not economically feasible.
Massive changes such as the development of electrical heating
or district heating systems may help to concentrate the
pollution, thus rendering purification procedures more
feasible.
188
-------
Waste and scrap. Pollution of the soil and water systems
may arise from the weathering of disposed metal products.
The reuse of scrap decreases both such pollution and the
need tor exploiting ores. However, most scrap for reuse
goes to iron and steel works where the melting process
releases impurities (e.g., cadmium and lead) stemming from
alloys, paints, etc. The impurities are seldom regained
but are emitted into the environment with the smoke. Un-
less effective exhaust purification is applied, such
recycling in fact increases rather than decreases exposure
to toxic metals.
Metal objects such as cans and capsules often end up in
community incinerators together with other waste. Burning
such waste may cause significant emissions of the most
common toxic metals, especially lead.
Careless dumping of metal waste can severely pollute water
basins. For example, old gravel quarries are definitely
unsuitable as dumps since dissolved metal compounds have
easy access to the ground water. All dumps must be without
any contact with water supplies. A specific problem has
arisen as sewage sludge has become used as fertilizer.
Such sludge may contain fairly high concentrations of many
metals, e.g. cadmium, which may then enter the food chain
(Friberg et al., 1974).
Contaminated food and beverages. The soil, water and
vegetation near large smelters or other major emitters are
often heavily contaminated with metals (WHO, 1973) .
Vegetables and grain grown in such environments usually
contain undesirably high concentrations of metal, just as
the vegetation near busy highways has shown high lead levels
(NAS, 1972). Community and well water often becomes
contaminated (WHO, 1973). Preventing undue exposure from
such sources requires, in addition to reducing the emissions,
the monitoring of food and water, and a ban of the sale of
189
-------
foodstuffs containing excessive amounts of metals. Food and
wine may also be contaminated from lead-glazed ceramic ware,
which releases large amounts of lead in an acid milieu. In
addition, canned food may include lead that has been released
from the containers. Several cases of cadmium poisoning
have occurred due to the ingestion of food and drink into
which this metal has dissolved from cadmium-plated articles,
e.g., ice trays, juice makers, roasting pans, etc. The
problem of fish contaminated by methylmercury has already
been mentioned. Other toxic metals, e.g. cadmium and arsenic,
have also been found in high amounts in seafood in some
areas. Preventing poisoning from such foods requires system-
atically organized analyses of their metal contents. Ceramic
ware should be tested for lead release during acid treatment
before being allowed on the market.
The consumption of illicit whiskey, often distilled with
discarded car radiators or lead-soldered tubings, causes
many cases of severe poisoning each year in the United
States (Hernberg, 1975) and perhaps elsewhere. Effective
consumer information is probably the only method for prevent-
ing such poisoning, since the trade seems to be flourishing
in spite of legal efforts to eliminate it.
The effect of lead plumbing on tap water has already been
mentioned. It should be added that artificial softening of
water will increase its lead content. According to WHO
(1973), most tap water does not contain excessive amounts of
metals, with the possible exception of cadmium.
Accidents. There are almost unlimited possibilities for
the accidental intake of toxic metals. For example, children
who have sucked lead toys or swallowed small lead items,
such as curtain weights, have even been fatally poisoned
(Bacon, 1967; Fristedt, 1968). The preventive action
lies in banning the use of toxic metals both as components
190
-------
and as paints, in such consumer goods as toys. When
thermometers are broken indoors, especially in bedrooms,
hazardous mercury concentrations may arise unless the spilled
droplets are meticulously removed, e.g., with the aid of
sulfur powder. As already mentioned, what can happen when
mercury-treated grain is consumed by mistake has been
experienced in Iraq.
Environmental monitoring. The need for monitoring metal
concentrations in food and drinking water, whenever there
is reason to suspect excessive amounts, has been stressed.
In addition, air and water monitoring help ensure that
regulations are being met. In principle, two types of
sampling procedures should be used. First, metal concen-
trations should be measured at representative sites so that
environmental standards can be controlled. Second, emissions
from particular pollution sources should be checked so that
individual polluters will comply with emission standards.
5. Medical aspects of prevention
Though the practice and the ambition level differ, the
principles of medical prevention of effects of toxic metals
are the same in public and occupational health, namely,
identification of risk groups,
biological monitoring,
health examinations,
early removal of the subject from dangerous exposure, and
early treatment, if possible.
Identification of risk groups. Advance knowledge of where
a risk is apt to occur is of paramount significance to the
identification of risk groups. Some of these groups have
already been pointed out, i.e., specific groups of industrial
workers, populations living around smelters and other emitters
(true for all toxic metals), slum district children (lead),
191
-------
consumers of illicit whiskey (lead), heavy fish consumers
(mercury,) dwellers in districts with lead plumbing systems
(lead), and users of primitive ceramic ware (lead). In
these population groups health screening procedures can be
recommended; however, such procedures are not warranted or
possible among the population at large.
Biological monitoring. In industry biological monitoring
has been carried out for decades, but in public health no
regular programs exist. Hence most experience from the
latter sphere is derived from ad hoc campaigns, such as the
case finding projects in the USA (lead) and the studies of
fish consumers in Sweden and Japan (mercury).
The choice of indicators and biological samples varies from
one metal to another. In general, the parameter of exposure,
i.e., the concentration of the metal itse;lf in some biolo-
gical medium, is the only one available because special
tests measuring response are lacking. Lead is an exception.
Because of its specific effects on blood formation, several
good parameters of response can be utilized. There are
several metals, such as arsenic, chromium, nickel and
manganese, for which sufficient information regarding the
significance of some given level in biological media is also
lacking. Consequently, while determinations of these
metals in biological media are technically possible, the
results are difficult to interpret. Hence, recommendations
for the routine biological monitoring for these metals
cannot yet be given.
At present lead is more suitable for biological monitoring
than other metals since the dose-effect and dose-response
relationships are fairly well known. The concentration of
lead in blood describes current exposure better than any
other test. Measurement of lead in blood should be preferred
to that of e.g. lead in urine. Exposure to organic lead is
an exception; lead in urine is the measurement of choice in
this case (Hernberg, 1975). When a low enough action level
192
-------
is selected (e.g. 50 or 60 jug/100 ml), enough safety can be
ensured if blood lead alone is used for health monitoring.
If the chosen action level is exceeded, the worker should be
removed from exposure and submitted to additional tests
measuring response. The most widely used ones concern
the concentrations of delta-aminolevulinic acid and copro-
porphyrln in urine. Quite recently the determination of
erythrocyte zinc protoporphyrin has become available as a
routine method (Fischbein et al., 1976). This may become
the method of choice in the future, even for indirectly
monitoring exposure, because of its simplicity and good
reflection of "active" lead. Its major weakness is the time
lag of some weeks that occurs before the concentration rises
due to the relatively slow turnover of erythrocytes. Dangerous,
new exposure may thus pass undetected if monitoring relies
only on this test. For screening the general population/
such as persons living around lead smelters, the determination
of erythrocyte delta-aminolevulinic acid dehydratase activity
is more suitable because of its high sensitivity. It is
also cheaper and has higher analytical accuracy than blood
lead determination (Berlin et al., 1973). Because of the
poor stability of the enzyme activity, the analysis must be
performed within a few hours.
In the case of mercury, exposure tests are less reliable.
Mercury in urine has been the most widely used exposure test
in industry when inorganic mercury has been the main type of
exposure. Recently, mercury in the blood has also come into
some usage. Methylmercury is excreted very slowly in the
urine, for which reason this medium is not suitable for
monitoring exposure to this compound. Instead, blood and
hair samples are good indicators (see Berlin, 1977).
Cadmium can be measured in both blood and urine but the
physiological significance of some given concentrations is
193
-------
not yet well known. It is believed that cadmium in blood
reflects current exposure while cadmium in urine reflects
body burden. Hence both blood and urine tests should be
used concurrently (Task Group on Metal Toxicity, 1976).
Regulations requiring biological monitoring concern
occupational exposure only. Many countries have such
regulations for lead. In principle, the frequency of
monitoring is determined by the degree of risk. For
example, according to Finnish regulations, blood lead deter-
minations should be made from one to six times a year, the
frequency depending on the type of industry. Workplaces
such as printing shops or garages need only one monitoring
annually while high-risk occupations such as lead smelting
and battery manufacturing require six. These are minimumsj
the physician responsible has the right to require more
frequent monitoring, if he considers it necessary (Tola
et al., 1976) .
Health examinations. Health examinations are clearly
warranted in occupational health and can be recommended
as a tool for preventing lead poisoning in
children living in slums. Obviously, they are meaningful
only when exposure is regular; accidental poisoning can
never be prevented by health examination programs. In
occupational health there are two types of examinations.
Preemployment examinations aim at excluding from exposure
workers who for some reason would be especially sensitive
to toxic metals. Periodical checks are programmed to
detect toxic effects of metals, at an early stage, as well
as to discover other diseases, since the latter may weaken
the worker's resistance. Special attention should be paid
to the critical organs involved, i.e., the hematopoietic
and nervous systems in lead exposure, the nervous system
and the kidneys in mercury exposure, and the kidneys in
cadmium exposure. In addition the information gained from
biological and environmental monitoring should be utilized.
194
-------
The frequency of periodical examinations may vary
depending on the degree of risk - usually one examination
a year is enough - provided prompt action is taken whenever
the action level is exceeded.
Cessation of exposure. The most important early "therapy"
for metal poisoning is the cessation of exposure. Usually
no further treatment is needed if exposure is interrupted
early enough. Although this is scientifically hard to defend,
a dualism generally exists between the action levels in
occupational and public health, the action levels forthe former
being more lenient. Removal of the worker from exposure should
be considered when the concentrations of the metal in the
blood or urine exceed the action level, when early symptoms
or signs are present or when reports of exceptional exposure
peaks are made. Ideally, the worker should be transferred
to a job that is free from exposure. Otherwise, sick leave
must be considered. The worker should not be allowed to
return to exposure until the blood or urine levels of the
metal are well below recommended upper limits. In the
general population, which is not under regular supervision,
the policy is to interrupt potentially hazardous exposure under
less urgent circumstances than the ones mentioned above.
Because of the potentially severe outcome, cessation of
exposure is even more important in such cases as when children
have ingested flaking lead paint. It is best to hospitalize
the child while the home is being sanitized. If possible,
the emotional disturbances underlying the pica syndrome
should also be treated (Chisolm and Kaplan, 1968).
Drugs, vitamins and special diets. The prevention of metal
poisoning calls primarily for technical solutions but under
special circumstances, certain adjuvants to technical pre-
vention may be considered. In developing countries where the
nutritional state of workers is poor, it is important to
ensure a diet adequate in vitamins, minerals and protein.
It is urgent that the persistent myth that milk protects
against lead poisoning be abandoned, as there is no
rationale for it. The tendency for pica that exists among
195
-------
slum district children is often due to iron deficiency;
correcting this nutritional deficiency may reduce the
risk of lead poisoning.
Interactions and competition between metals alter their
toxicity. For example, zinc decreases many of the toxic
effects of cadmium, and also of lead (Border et al., 1976;
Abdulla and Haeger-Aronsen, 1973).and calcium deficiency
inreases the absorption of lead from the gut (Task Group on
Metal Accumulation, 1973). While dietary and interactional
factors must be kept in mind and improvements made in
developing countries whenever minimum daily requirements are
not reached, attention should not be diverted from the
only correct preventive strategy, namely, that of improving
industrial hygiene.
Finally, the prophylactic use of chelating agents deserves
some comments. This malpractice is astonishingly widespread
in several European countries and in the United States
(Lilis and Fischbein, 1976). There are several reasons to
condemn this form of "prophylaxis". First, it provides
the employer with too easy a way of continuing to have
poor workplace hygiene. Second, all chelating agents also
mobilize essential metals, such as zinc and copper (Lilis
and Fischbein, 1976). Third, all such agents have adverse
side effects, such as nephrotoxicity (pencillamine and
EDTA), myelotoxicity (penicillamine), and they may be
allergenic (penicillamine) (Lilis and Fischbein, 1976) .
Fourth, oral administration increases absorption from
the bowels. For these reasons no serious physician should
consider the prophylactic use of chelating agents. The
treatment of metal poisoning is quite another matter.
ACKNOWLEDGMENTS
The author wishes to thank Mr. Antti Tossavainen, L.Sc. (Tech.),
Mrs Pirkko Pfaffli, M.Sc., and Dr. Sakari Tola, M.D., for
their constructive criticism of the manuscript.
196
-------
REFERENCES
Abdulla, M. and Haeger-Aronsen, B. (1973). International
Research Communication System - General Pharmacology 73-8,
8-14-1.
Abramson, E. (1967). Lakartidningen 64, 3638-3640.
ACGIH (1974). "Threshold Limit Values for Substances in
Workroom Air." American Conference of Governmental Industrial
Hygienists, Cincinnati.
American National Standard for Respiratory Protection Z88.2-
1969 (1969). U.S. Standards Institute, New York.
Bacon, A.P.C. (1967). Lancet I, 264-266.
Bakir, F., Damluji, S.F., Amin-Zaki, L., Murtadha, M.,
Khalidi, A., Al-Rawi, N.Y., Tikriti, S., Dhahir, H.I.,
Clarkson, T.W., Smith, J.C. and Doherty, R.A. (1973). Science
181, 230-241.
Beattie, A.D., Moore, M.R., Devenay, W.T., Miller, A.R. and
Goldberg, A. (1972). Brit. Med. J. 2, 491-492.
Berlin, M. (1977). In: "Toxicology of Metals." Volume II, pp
301-344. Environmental Health Research Series, 600/1-77-022,
Environmental Protection Agency, Research Triangle Park.
Berlin, A., Del Castilho, P. and Smeets, J. (1973). In:
"Environmental Health Aspects of Lead." pp 1033-1049. Commission
of the European Communities Directorate General for Dissemination
of Knowledge, Center for Information and Documentation CID,
Luxembourg.
Border, E.A., Cantrell, A.C. and Kilroe-Smith, T.A. (1976).
Brit. J. Ind. Med. 3_3, 85-87.
Chatterjee, B.B., Williams, M.K., Walford, J. and King, E.
(1969). Amer. Ind. Hyg. Assoc. J. 30, 643.
Chisolm, J.J., Jr. and Kaplan, E. (1968). J. Pediatr. 73,
942-950.
Cmarko, V. and Rosival, L. (1970). Preprint to Conference on
Chemical Pollution and Human Ecology, Prague, 1970.
Elvingson, C.O. (1972). In: "Environmental Science in Practice."
(P. Brink, T., Cewe, A. Horn af Ranzien, E. Olerud, C. Ramel
and H. Sjors, eds) pp 172-205. Centraltryckeriet, Boras,
Sweden (In Swedish).
Fairhall, L.T. (1969). "Industrial Toxicology." 2nd edition.
Hafner Publishing Company, New York.
197
-------
Fischbein, A., Eisinger, J. and Blumberg, W.E. (1976). Mount
Sinai J. Med. 4^3, 294-299.
Friberg, L., Piscator, M., Nordberg, G.F. and Kjellstrom, T.
(1974). "Cadmium in the Environment." CRC Press, Cleveland.
Friberg, L. and Vostal, J. (1972). "Mercury in the Environment."
CRC Press, Cleveland.
Fristedt, B. (1968). Lakartidningen 6j>, 3028-3034.
Guinee, V.F. (1971). Trans. N.Y. Acad. Sci. _3_3, 539.
Hamilton, A. and Hardy, H.H. (1974). "Industrial Toxicology."
3rd edition, pp 134. Publishing Sciences Group, Acton, Mass.
Hernberg, S. (1973). Work Environ. Health 10_, 53-61.
Hernberg, S. (1975). In: "Occupational Medicine: Principles
and Practical Applications." (C. Zenz, ed) pp 715-770. Year
Book Medical Publishers, Chicago.
Japanese Ministry of Health and Welfare (1967) . Report on
the cases of mercury poisoning in Niigata, Tokyo (mimeographed
document).
Jensen, S. and Jernelov, A. (1969). Nature 223, 753-754.
Lilis, R. and Fischbein, A. (1976). J. Amer. Med. Assoc.
235, 2823-2824.
Linch, A.L. (1974). "Evaluation of Ambient Air Quality by
Personnel Monitoring." pp 121-170. CRC Press, Cleveland.
MacGregor, A. (1975). Environ. Health Perspect. 12, 137-148.
Moore, J.E. (1970) . Amer. J. Public Health 6_0, 1430.
NAS (1972). "Airborne Lead in Perspective." Committee on
Medical and Biological Effects of Environmental Pollution,
National Academy of Sciences, National Research Council,
Washington, D.C.
NIOSH (1972). "Criteria for a Recommended Standard ...
Occupational Exposure to Inorganic Lead." U.S. Department of
Health, Education, and Welfare, Public Health Service,
Health Services and Mental Health Administration, National
Institute for Occupational Safety and Health.
Nordberg, G.F., ed (1976). "Effects and Dose-Response Relationships
of Toxic Metals." Elsevier, Amsterdam.
Olishifski, J.P.E. andMcElroy, F.E., eds (1971). "Fundamentals
of Industrial Hygiene." pp 439-456. National Safety Council,
Chicago.
Patterson, C.C. (1970). Proc. Amer. Phil. Soc. 114, 9-12.
198
-------
Patty, E.A. (1962). "Industrial Hygiene and Toxicology."
2nd revised edition, Vol. I: "General Principles." Interscience
Publishers, New York.
Roach, S.A. (1973). In: "The Industrial Environment - its
Evaluation and Control." pp 139-155. U.S. Department of
Health, Education, and Welfare, NIOSH.
Rossano, A.T., ed (1969). "Air Pollution Control Guidebook
for Management." Environmental Sciences Service Division,
E.R.A., Stamford, Conn.
Task Group on Metal Accumulation (1973). Environ. Physiol.
Biochem. 3, 65-107.
Task Group on Metal Toxicity (1976). In: "Effects and Dose-
Response Relationships of Toxic Metals." (G.F. Nordberg, ed)
pp 1-111. Elsevier, Amsterdam.
Tola, S., Hernberg, S. and Vesanto, R. (1976). Scand. J.
Work Environ. Health 2_, 115-127.
WHO (1973). Long-term programme in environmental pollution
control in Europe. The hazards to health of persistent
substances in water. Annex to a report of a working group.
Technical documents on Arsenic, Cadmium, Lead, Manganese and
Mercury, Helsinki, 10-14 April, 1972. WHO Regional Office
for Europe, Copenhagen.
Williams, M.K. (1975). J. Occup. Med. 17, 93.
Williams, M.K., Walford, J. and King, E. (1968). Brit. J.
Ind. Med. 2J5, 144.
Williams, M.K., King, E. and Walford, J. (1969). Brit. J.
Ind. Med. 2£, 202.
Ziegfeld, R. (1964). Arch. Environ. Health 8, 14.
199
-------
GENERAL ASPECTS AND SPECIFIC DATA ON
ECOLOGICAL EFFECTS OF METALS
Karin Beijer and Arne Jernelov
1. General aspects of ecological effects
1.1 Equilibrium and balance in the ecosystem
Ecology is the study of the reciprocal relations between
living organisms and their environment. Each living organism
affects its environment by consuming some substances such as
water, solid foodstuffs and oxygen and by secreting others
such as urine, feces, carbon dioxide etc. A specific
organism, when confined on its own in a closed environment,
tends to destroy the prerequisites for its own existence. The
substances needed tend to diminish, while useless and sometimes
poisonous waste heaps up in the surroundings. Different
organisms have different metabolisms, however. That which is
useless or harmful to one organism may be useful to another.
One of the most obvious examples is that plants use carbon
dioxide and water and give off oxygen while animals use
oxygen and give off carbon dioxide.
In order to subsist, each ecosystem must recirculate
all substances that are not regularly supplied from outside.
This leads to a balance between the organisms which are part
of the system and the relevant physical-chemical parameters.
The schematic figure below illustrates the above mentioned rule.
A ->->->->-»
B -»-»-»-»-»
a -»-»->-»->
b -»-»-»-»-»
c -»-»-»-»-»
a -»-»->-»-»
3 ->->->->-»
y -»-»-»-»-»
species
1
species
2
species
3
->-»->->-> a
->-»-»-»-» b
-»->-»-»-» a
-»-»-»-»-» 3
-»->->-»-» y
->-»-»-»-» A
->->-»->-> B
-»->-»-*-> C
201
-------
Since each species (or individual organism) shows a tendency
to destroy the prerequisites for its own existence it is
evident that the species (or individual) cannot live alone
in a secluded system. An experiment with the growth of
microorganisms in a closed cultivating duct with a given
substrate or a given quantity of oxygen illustrates this.
At first the growth of the population is exponential, where-
after it slows down; finally, the population breaks down and
dies out, see Figure a. If an additional species is introduced
into the system and this species is a predator of the first,
the growth curve in Figure b or the one in Figure c will be
obtained.
a)
b)
4-i in
0 g
w
•H
C
. . (0
e tn
3 M
2 O
0)
-fr
•" V
Time
If the number of species in a food-chain increases or the
number of the environmental parameters regulating the number
of individuals increases, the probability for a balanced system
will increase as well, i.e. the number of individuals of a
species will never be so low that it will be eliminated.
Obviously the different species can interact in many ways,
for example by using and transforming each other's excretion
products, through predation or through competition for
scarce factors. Also, environmental factors interact with
organisms and with each other. The number of simple inter-
action possibilities increases faster than the number of both
species and environmental factors in the system. The inter-
202
-------
actions will equal n(n-l)/2 where n is the number of species
and environmental factors. The interactions can be said to
stabilize the system - e.g. by tending towards an even
distribution of individuals among the species.
An ecosystem with a high degree of constancy in physical
parameters and a large number of species is referred to as
having high stability. Conversely, an ecosystem with varying
(and unpredictable) physical parameters and few species is
characterized by a large variation in the number of individuals
within the species and is referred to as having low stability.
This does not necessarily imply that the stable high diversity eco-
system is more tolerant to additional stress than one of low
diversity already subject to stress.
The relationship between species and the number of individual
organisms on the one hand, and the degree of diversity and
instability on the other, have been used to characterize an
ecological system. Many mathematical formulae have been used
to express this relationship in the form of diversity indices,
e.g. the Shannon-Wiever function, Simpson's index, Sander's
rarefaction technique (Pielou, 1969; Fager, 1972).
Reduction of species diversity (changes in indices) when passing
from moderately to strongly polluted parts of an area is
frequently used in pollution ecology to measure the stress to
which the ecosystem is being subjected.
Because of the interdependence of all included species and
environmental factors a change in any parameter will have
consequences for all the other parameters. The consequential
changes as a rule tend to neutralize the effects of the primary
change so that the system may return gradually to its earlier
point of balance. Sometimes, the system will instead come to
an altogether new point of balance.
The following lines illustrate different types of ecosystems with
different capacities for "balance" and "stability":
203
-------
A) B)
v
The lowest points of the curves are the balance points. The
horizontal curve sections with slight slopes constitute the
amplitude of oscillation, and the vertical difference in
altitude. The ecosystem may be regarded as a ball which is
removed from the balance point and which later on falls back
to the same point.
System A is obviously very stable. It has a given point of
balance to which the system is rapidly brought back if
disturbed. Much energy is required to disturb such a system.
System B has a less marked point of balance and can be disturbed
quite easily. The system will thus oscillate around the point
of balance before being stabilized.
System C, finally, has no specific point of balance - it
has a "balance area" within which no intrinsic force will
move or maintain the system. When moved outside the "balance
area" it will be brought back to it.
It should be pointed out, already at this stage, that knowledge
about a system's point of balance, amplitude and stability -
e.g. the shape of the curve within a certain interval - does
not tell us about its reaction to strong interferences. That
is, it is possible to interpolate earlier observations but
never to extrapolate them. Moreover, the distinction between
one system and another where a new species or environmental factor
has disappeared or been added is often subtle or vague, but
204
-------
according to the above definition the new system has other
properties. Many of the interactions between organisms are of
a regulating feed-back type, which is one of the reasons for
their stabilizing effect on the system.
1.1.1 Examples of stabilizing interactions
An ecological system consists of five species of which two are
predators. Each predator lives on two of the other three species,
A
Predator soecies
Food species
The number of predators is regulated by the food supply and
the predators usually show a preference for eating the most
commonly available food species. The total number of the food
species individuals is regulated by the supply of nutriment
and the relative number of the relative predation pressure.
Now, let us assume that all three food species and the two
predators are equally numerous to start with, and as a result
of an external disturbance imposed on the system, the
number of individuals of food species 1 is strongly reduced.
Predator species A will then turn to eat principally 2. At
the same time, the decreased total food supply causes the number
of A to diminish relative to B. Owing to the fact that both
A and B (A principally and B partly) now eat 2, the number
of 2 is reduced. B then starts to eat mainly 3 while A
continues to eat 2 since it is still the most common of the
food factors. When A eats 2 and B eats 3, the number of 1 can
increase again while the number of 2 and 3 decreases. There
are now more B than A, because of the earlier adaptation to the
food supply, and the predation pressure is found to be strongest
on 3 which accordingly diminishes most rapidly. When 1 becomes
more numerous than 2, A will proceed to eat mainly this species,
and when 3 becomes rarer than 2, B will proceed to eat 2. Now
205
-------
that the food supply for A is more abundant than that for B,
A will increase in number while B decreases.
In this way the number of the different orcanisms oscillates
with a consequential successive reduction of amplitude until
the system gradually recovers its earlier equilibrium. The
interaction between the organisms contributes to the stability
of the system. Thus, the mutual influence of the organisms
in this kind of system tends to reduce the amplitude of the
variations in the number of individuals.
1.1.2 Example of spiral to a new point of balance
Take a fish species, which lives the greater part of its life
in the ocean but spawns and subsequently dies in running
fresh water - in the same river from which it migrated after
its first year. Provided the time period the fish needs to
reach sexual maturity is relatively constant and spawning occurs
only once, the fish will gradually tend to split up into
subponulations separated by different spawning years.
We assume that the number of individuals in a population is
primarily limited by the number of fish surviving the first
year in the river. For the sake of simplicity we further
assume that the limiting factors are predators and supply of
food organisms. In every natural population some variations
in the number of individuals will occur.
A numerous offspring of fish in the river one year tends to
cause the food organisms to decrease and the number of pre-
dators to increase because of the abundance of prey. As a
consequence, the following year, the fry smolt will encounter
a scarcity of food and a larger number of predators and thus
fewer fish will survive in the river that year. The year after
that the food will be abundant and the predators few because of
starvation or migration which means a good year for the fish
and so on. Thus the annual subpopulations will have different
sizes.
206
-------
This type of increased cyclic variation need not express
unstability in the ecosystem. It merely emphasizes the
complexity of the stability concept and shows that different
development stages separated by some spatial dimension as
well as regular migration should in some way be included in
the concept "stability".
1.1.3 Ecological subsystems
Most ecological concepts are ambiguously defined. This applies
to the word "system" as well. By way of introduction, the
word "system" was applied in its theoretical meaning: a
closed ecosystem in an area in which only the supply of
energy in the form of light recirculates all substances and
maintains all included types of organisms. Through the examples
given above, the concept "system" has received a more practical-
operational and thus a less strict meaning. All known ecosystems
are more or less open and the majority consists of various sub-
systems. For the purpose of more exhaustive studies a lake
as an ecosystem is often divided into a number of subsystems
as for instance: two different littoral zones; the Phragmites
and Nymphea zones, the lake bottom without vegetation, sometimes
separated into soft and hard bottom, and the pelagial and
profundale water bodies outside the littoral zones. Certain
groups of animals, such as fish, migrate freely between these
subsystems.
The subsystem can be internally unbalanced, but the develop-
ment towards a breakdown in one of them will be checked
at some given point by the surrounding subsystems - by
external forces - and the balance thus restored.
In the discussion above it was assumed that the subsystems
constitute different systems both physically and by appearance
e.g. the littoral zone or the free water mass. The same
argument may be held for a physically uniform system like
e.g. the surface water - the pelagial.
In a fairly big lake or the oceans the pelagial constitutes, in
one respect, a unit. Yet for an individual organism of the
207
-------
phyto- or zooplankton types, large parts of the space within
this unit are beyond its reach. This would not be of significance
if the organisms were distributed uniformly within the unit.
However, this is not the case. Chance, if nothing else, will
lead one to expect that some water subunits contain more
plankton than others. Due to the limited mobility of these
organisms one subunit, during a certain period of time, will
constitute an ecological subsystem. During this period the
migration in and out of the subunits will be of importance for
the balance between the organisms. It is evident that the
extension of such a closed unit both in time and space is
rather vague.
Within these subunits with their different contents of organisms
the development might proceed very differently. In some, the
consumer organisms may eliminate the producers. In others, the
consumers may be very scarce or totally absent and thus leave
the producers to increase exponentially. This need not
lead to a breakdown of the system as a whole since the
durability in time of the subsystems is limited, and their
borderlines shifting. This is an example of internal balancing.
In reality not only chance will cause an uneven distribution of
organisms. The wind and currents, for instance, will affect
groups of organisms differently according to their mobility,
size, density and shape etc, leading to a much greater
spatial separation than could be accounted for by chance
only. This "patchiness" is a well-known phenomenon among
plankton and also among benthic organisms.
From the above discussion it is clear that the concept of
"balance" must be further extended and diversified. Balance
in the total system does not presuppose a mathematical sum
of balances in the subsystems. Spatial separation and the
presence of non-discrete time and space dimensions make con-
tinued existence of the species possible.
208
-------
1.2 Disturbance of ecosystem stability
The concept of biological effects of a pollutant includes
changes in the individual, the whole population and the
ecosystem. Most of the research in this field has so far
been directed towards the individual. This is the obvious
approach when studying effects in humans but it is impossible
to predict the effects on an ecosystem from knowledge about
the effects on individuals from an isolated species.
The effect produced in the individual by a harmful substance,
may or may not affect the whole population (cf. the discussion
on "litter mate-correlation" below). If the whole population
is affected this in turn may or may not -depending on in what
way the population is hit - affect the concentration in the
organism of the substance in question. Thus one possible effect
is a negative feed-back where the final result may be a
steady state as for instance in the hypothetical example:
high concentrations in fish -> reduced reproduction -> less
competition for food -> a more rapid growth -» lower concentra-
tions in the fish -> improved reproduction ....
Another conceivable effect is a kind of vicious circle or spiral
with the total elimination as the final result: high con-
centration in the fish -» impaired growth -» higher
concentrations -» further impairment of the growth -» even
higher concentrations ....
A number of examples can be given which show that the lack
of ecological understanding leaves us without a guide as to
the consequences of a pollutant.
We might for instance want to find out the effects of a
temperature increase in a lake on a certain fish species.
In laboratory tests it is found that, given enough food, the
fish will grow faster and give more offspring when the temperature
is raised. Thus it can be predicted that the result would be
more and larger fish in the lake. But in reality this might
not be the case at all. If there are other species even more
209
-------
favored bv tne ciKinqed coridi niona, these will take over
•_.irouqh conpe*- i t.ion L'oi rooc and os a consequence our species
might even become total lv elinrj netted.
An attempt to anwer the question about the effects on a fish
population if 25? of the brood is eliminated will give rise
to a new question as; to whether '-he "litter mate-correlation" -
that is the probability for survival of one individual when
affected by the death of another - is positive or negative.
For decades attempts have been made to increase the population
of pike in Swedish lakes through the introduction of fry.
These have shown that the survival of the individual mainly is
dependent on the competition foi food with other fry of th --
same age. Thus if the number of fry is increased the
probability for survival of the individual decreases. It
therefore seems likely that the probability for survival of the
individual would increase with a decreasing total number. The
conclusion would be that the effect on the adult population of
this kind of fish species through a 25% kill of the fry would
not be drastic.
However, in the case of grouper species like e.g. herring the
effects would most likely be severe on the adult population.
The predators are always superior and take the number of
fish they need. This may be an explanation for the develop-
ment of shoals, since each individual fish will stand a better
chance of survival when surrounded by many others of the same
size.
Thus, if the number of fry is decreased due to increased
competition for food, each fish will be in a statistically
worse situation and the probability for it to fall prey to
prc'V'ion -?i]l increase. (Provided 'chat fcod is abundant enough
so thut .ihe forming cf shoals will not have a negative effect
on the growth. Whe.:e there is a scarcity of food a decrease
in the number of fry would leave more food per individual and
thus promote a higher growch rate. This would give rise to yet
another unpredictable situation.)
210
-------
From these two examples it can be seen that knowledge on the
litter mate-correlation is necessary for drawing conclusions
when testing toxic effects on fish fry or disturbances on re-
production. However, this relationship is very rarely known
or possible to guess.
If the balance between zoo- and phytoplankton is based on a
feed-back system where a decrease of phytoplankton - food -
leads to fewer eggs for the zooplankton which means less grazing,
th.'.s in turn leads to an increase of phytoplankton which means
more eggs etc. The system is very sensitive to disturbances
of reproduction. If, on the other hand, the total balance is kept
up through the interaction of unbalanced subsystems separated
in space, the effects would not be so serious.
Usually it is not known through what principles the ecological
balance is maintained. Thus the impact on the ecosystem as
a whole of decreased reproduction of a certain species caused by a
pollutant cannot be predicted. Like in the examples above,
different hypotheses concerning the premises for an ecological
balance lead to diametrically opposed conclusions about the
ecological consequences of a certain effect on the individual.
1.3 Accumulation of trace substances in biological systems
The concentration of a substance in a biological system - be it
an organ, an individual, a population, a trophic level etc. -
is the result of input, dilution and output. Bioaccumulation of
a substance will occur when the rate of uptake exceeds that
of transformation, dilution and/or elimination. Any substance
with such physico-chemical properties that it may be transported
along with the flow of material may potentially be accumulated
in a biological system. Biomagnification is the increase of
the concentration of a trace substance on a body weight basis in
successively higher trophic levels - e.g. plankton -» fish -»
seal. A prerequisite for this is usually that a major uptake
route is via the food and that the substance is accumulated. Thus,
in the aquatic environment, where the uptake of many trace
substances from the ambient water is of importance, the lifespan
211
-------
and growth rate of an organism may be of greater significance
for the concentration than _i ts position ir: the food ohctJii.
Fagerstrom „.:£ ~ , . L '_v, 3) have discussed a general
methodology f c r .:. . .« j.l '.ii.g !:i iac . umul a ~c ion onenomena . It is
suggested that any orooess in a biological system that will
result in a transport of a crace substance should be considered
in terms of a earner- trace substance complex. The following
are some points u.ade in their paper :
Through a biological system there Is a "flow (continuous or
intermittent) of a ^PE/ii^E- Tlli-s may ije a well defined chemica !.
species, such ;i s : ?. 1 ^ .'. ;. 11 or o::ygen or a chemically less specify
concept, such ;.o :'."_ip:ds'i or "energy" L+- is essential that the
c~rrier flow ij --. pr jcees \,hich is a normal and necessary part of
the metabolism o >: ;:aa ,?v rcern, having its own regulatory mechanisms
in order ;"o m. "•:'.' TVI r -efficient degree of homeostasis.
Through the ~;,*
with such chap." '.
an analogue to '
ca"-"ri=r flow i'_
t r a n .5 _;< o r!: j d a 11
may b? a _aui^ •
a me tal such e =s
"':-" ~: is also a flow of a trace substance
•* .~-c .les that it acts more or less as
a-'i'l-r. Phis ".cans that it traces the
sen^e t>.ac it is being passively
90
.'"ier. The trace substance
Sr, a pesticide such as DDT,
The rate of fj.cw of the
by a variety cf ^=>ri^^]f
: -. r example, if
ier '-hro-gh the system is affected
,:-dege:ious as well as exogenous.
'• \" '.-• ler^d is a fish specimen and
LJ el'.ar that the flow of energy
- = es i: ,:icerat are and body size.
:• -. • o a •? the er "i, '. ;ii
ir:i--1;e ':o t'le '- •, "t."1.
taken in bv a Tish ee
- ? f O'- of the trace substance.
- . ':. - • a i o^ue to ";ie carrier, the
'_•; e ': carr-er would always be the
•-_•'_•: of carrier in the source of
" . Lhs -""ur. vnt of methylmercury
\ ^ l -,'o' ]d be the same as the
212
-------
concentration of methylmercury in the food. This is not the
case, however, since the analogy between food and methyl-
mercury is far from perfect.
The summed effect of all flows into and out of a system,
integrated over time, will determine the pool sizes of
carrier and trace substance. Thus, it follows that if the
trace substances were to act as an analogue to the carrier,
the ultimate steady state situation in a food chain would
be that the concentration of a trace substance relative to the
carrier in the organisms would equal that in the source of
contamination, that is, no accumulation in the food chain. This
is the case with for instance C where the isotopic effect is
negligible and a constant input to the biosphere has been
going on for a long time. In other cases where there is an
analogy but where the contamination has started recently
or is changing with time, the ecosystem may be in a transient
state. In such systems the ultimate steady state situation is
predictable when the ratio between the trace substance and the
carrier in the source of intake is known. From this follows
that the concentration calculated on a body weight basis of
90
a trace substance - say Sr - may be higher in one type
of organism (e.g. a snail) than in another (e.g. plankton) although
both have the same concentration with regard to the carrier
system, in this case calcium.
Trace substances which will be accumulated (e.g. methylmercury)
do not act as analogues to their carrier in most processes,
such as transport over surface membranes and metabolism. This
is basically the reason that they will be biomagnified in food
chains. In other words: the fact that a trace substance may
act as an analogue to a carrier in a biological system does not
imply that it will be biomagnified in food chains. On the
contrary, its very deviation from analogy, for instance by
being more efficiently retained in a tissue than the carrier is,
is what causes its accumulation. Therefore, the commonly used
argument that a trace substance is magnified in food chains
because of a ~ 90% loss of energy in each trophic level is
213
-------
misleading. It is not the 9:1 relation between metabolized
and stored carrier as such that is important; it is the
fact that the relation between metabolized and stored trace
substance is different from (and less than) 9:1.
2. Effects of metals on ecosystems
Little is known about the gradual ecological changes which
may result from metal contamination of the environment. It is
generally true that e;ffects will go unnoticed unless they
are extreme. Also, there is very little knowledge about at what
level in a certain organ a metal or a combination of metals
start to be harmful to the individual and thereby to the
ecosystem. In most cases, observed elevations in metal levels
in the biota are our only indication of metal disturbances.
2 .1 Terrestrial ecosystems
Metals reach the terrestrial environment through atmospheric
fallout (wet or dry deposition), through iirigation with
contaminated water, with pesticides and different types of
fertilizers and also through the disposal of solid waste.
2.1.1 Soil
The influence of air-deposited metals on biological and bio-
chemical processes in soil has been studied. It has been
found that some rnetals, particularly copper, limit the de-
composition rate of forest litter at concentrations of a few
times the background level. Microbial decomposition of cellulose
and starch, mineralization of nitrogen and phosphorus, and
activities of a number of soil enzymes have also been shown
to be disturbed by an increase of copper and zinc levels.
The ultimate consequence of this may be a productivity decrease
of the ecosystem (Tyler, 1975).
2.1.2 Vegetation
Metals are taken up by the plant roots and subsequently
distributed to stems, leaves and ceeds. The uptake of a
metal is affected by soil organic matter content, pH, the
presence of other metals etc. The extent oE foliar absorption
214
-------
is difficult to determine but estimates indicate that it may
not be very significant percentagewise for a number of
metals e.g. Pb. Plant metabolism may be disturbed in many
ways by toxic metals, resulting in reduced standing crops
and fertility.
Some plants tolerate high soil levels of specific metals which
they tend to take up and enrich, a fact that has been used
by geologists for base metal prospecting. "Accumulator species"
have been found for F, As, Si, V, Cr, Co, Ni, Cu, Zn, Se,
Sr, Y, Sn, Ba, W, Re, Bi and U. Both Hg and Cd are highly
toxic to most organisms and very few natural accumulator
species have been found for these two elements (Peterson, 1971) .
However, experiments indicate that plants may readily absorb and
translocate Cd to above-ground tissue when cadmium levels in soil
are elevated (John, 1975). Cadmium can be enriched to a very
high degree in crop species such as wheat and rice. Plutonium
may be solubilized by soil microorganisms and taken up by
plants (Drucker, 1976).
The enrichment of trace elements from sewage sludge fertilizer
in soils and plants has been studied (Andersson and Nilsson,
1972) . It was found that larger amounts of trace elements
are added than removed when sewage sludge is repeatedly used
on cultivated soils, leading to a gradual enrichment in the
soil and for some elements an increased uptake by plants and
transference to the food chains. The extractable soil content
of Hg, Zn, Cu and Se was found to be increased more than 100
per cent and that of Ni, Cr and Pb more than 50 per cent.
In vegetation the levels of Hg, Zn, Cu, Ni, Cr, Pb, As and
Mg were increased by about 50 per cent or more. Similar experi-
ments with corn plants have shown that Cr, Mn, Ni, Pb and Cu
were not significantly affected by waste additions, whereas
Fe, Zn and Cd were (Webber and Beauchamp, 1975).
The enrichment of toxic metals in the soils around smelters
has been shown to result in denudation of forests and in
the inability of plants to recolonize the exposed soils
(e.g. Whitby, 1975; Hutchinson, 1975). In the environs of
215
-------
a zinc smelter also emitting Cd, Pb and As altered floristic
composition and lowered species diversity could be observed.
The degree of effects was found to be correlated with increased
smelter proximity and with levels of the metals in both soil
and vegetation (Benenati and Risser, 1975).
2.1.3 Fauna
Aerial emission of metals - especially from smelters - is known
to have a damaging effect on livestock and other grazing animals
In connection with a study of atmospheric metal fallout, horses
that probably died from metal poisoning were analyzed and
found to have very high levels of some iretals, notably lead
and cadmium (Goodnan and Roberts, 1971). As was shown by
Schroeder and Balassa (1961) cadmium is accumulated in grazing
animals, especially in kidney and liver. Westermark et al.
(1973) found indications of a biomagnification of cadmium
but not of arsenic in the terrestrial food chain.
Automobile exhausts are a major source of lead emission -
both organic and inorganic. Kidney and liver levels in sheep
grazing near a major road were found to be four to five
times higher than in a control group (Landstrom, 1971) , and
the level in milk produced by cattle fed on hay harvested
along a road was found to be four times higher than normal
(Bovay, 1971) . The transport of lead in an arctic food chain
was studied by Holtzman (1968). No biomagnification was
demonstrated, probably because inorganic lead is stored
mainly in bone and thus made less and lass available to each
successive trophic level.
It seems that the terrestrial fauna is to a large extent
contaminated with mercurials through the ingestion of seeds
treated with mercury-containing fungicides. Analysis of
feathers from goshawk (Accipiter gentilis) shows a strong
correlation between the introduction and then the withdrawal
of alkyl mercurials on the one hand and the mercury levels
in the feathers on the other. Levels were below 3 mg/kg
before the introduction of mercurials, 20 mg/kg during their
216
-------
usage and 3 mg/kg after their withdrawal (Westermark and Sjo-
strand, 1972).
In connection with the outbreak of mercury poisoning in Iraq
in 1974, where people had been eating methylmercury-treated
seed, wild life samples from different areas were analyzed for
mercury content (Jernelov, 1974). Tail feathers from seed-
eating birds were found to contain 13.5-21 mg/kg of mercury
which is about ten times higher than values reported from
Ethiopia and within the range found in Sweden and Canada. Extremely
high concentrations of mercury were found in muscle tissue (up to
40 mg/kg) and feathers (up to 52 mg/kg) of dead seed-eating birds.
These very high levels were only found in the vicinity of
storage houses where returned seed was kept. No predatory
birds could be caught and analyzed.
2.2 Aquatic ecosystems
Metals enter the aquatic environment vid industrial and
domestic sewage discharges, dumping, surface run-off, atmos-
pheric fallout etc.
2.2.1 Effects of metals
Metals - normal constituents of the aquatic environment as
opposed to many other contaminants -are always present in at
least trace amounts in the organisms which are adapted to
natural variations in metal concentrations in food and water.
Usually the natural background levels are extremely low and
every increase through anthropogenic input can be expected
to be harmful to the biota.
The aquatic environment is extremely variable and there are
many factors which will modify the effects of a metal on the
biota. Ionic strength is one variable - marine, brackish and
freshwaters are obviously very different - but also pH, redox
potential, hardness, the presence of organic and particulate
matter etc, greatly influence both the chemical speciation
of the metal, the type and stability of the ecosystem and
thereby the impact of a metal. Generally, only the free metal
217
-------
ion is toxic whereas the portion bound in chelates and
complexes or adsorbed onto particles is comparably inactive.
Thus, many metals show a reduced toxicity in hard water and in
water with a high content of organic or particulate matter.
In acidified lakes, a larger portion of most metals is in
free ionic form, rendering them more toxic than in a lake
with neutral or basic water.
In the aquatic environment the most obvious ecological effects
of metal pollution can be found in rivers and lakes receiving
waste water and tailings from mines and from the processing
of ores. Many rivers and lakes have parts where invertebrate
fauna is lacking or markedly poor in variety and abundance and
where the water is too toxic for survival of many fish species.
2.2.2 Uptake and accumulation
Environmental variables and chemical speciation also greatly
influence the uptake and accumulation of a metal by an organism.
For example, the arsenic levels in fresh water fish are lower than
in marine fish. The concentrations have been found to be
well correlated with salinity. Even the extremely high local
contamination in the northern low-salinity part of the Baltic
does not result in arsenic levels much higher than those
commonly found in the oceans. Another example is the great
difference between methylmercury - which is accumulated -
and inorganic mercury - which is not.
Plankton, macroalgae and many species of molluscs do not
seem to be able to regulate their uptake and/or output of
metals. They can concentrate metals from the ambient water
up to several thousand times. The passive nature of the uptake
of cadmium, copper and zinc has been demonstrated for a brown
algae species for which consistent concentration factors
were obtained over a wide range of dissolved concentrations
(Morris and Bale, 1975) .
Cadmium and mercury (methylmercury) are accumulated in aquatic
organisms to a very great extent and many other metals, e.g.
218
-------
Pb and As, may be accumulated to a varying degree. It generally
seems that the uptake from the ambient water makes a more
significant contribution to the metal contents in aquatic
organisms than a biomagnification via the food chain.
Bioaccumulation of lead has been studied in a river ecosystem
(Leland and McNurney, 1974). Accumulation by benthic invertebrates
and fishas was shown to be a function of niche and habitat.
Lead did not seem to biomagnify in the food chains, detritus and
herbivores having higher lead levels than the carnivores.
Fagerstrom and Asell (1973) employed a mathematical model to
investigate the bioaccumulation of methylmercury in a three
step fresh water food chain (bottom fauna - roach - northern
pike). It was concluded that: 1: Bottom fauna, even when
containing low amounts of methylmercury, may account for a
significant fraction of the gross intake of methylmercury into
the food chain; 2: The question of whether dietary or gill
uptake is the more important can only be answered related to a
specific system and trophic level. For the overall food chain,
gill uptake is probably the most important route; 3: The top
predator, northern pike, does not reach a steady state with
respect to mercury in flesh during its lifetime; 4: If the
rate of methylmercury supply to the food chain is instantaneously
changed, the time to reach a new equilibrium with the
environment may be long as 10 years in the top predator.
219
-------
REFERENCES
Andersson, A. and Nilsson, K.O. (1972). Ambio !L, 176-179.
Benenati, F.E. and Risser, P.G. (1975). In: "International
Conference on Heavy Metals in the Environment." p. C-154
(abstract only) Oct. 27-31, 1975, Toronto.
Bovay, E. (1971). In: "Bulletin des Eidqenossische Gesund-
heitsamtes." Beilage B. No. 3. Zurich (In French).
Ducker, H. (1976). Personal communication.
Fager, E.W. (1972). Amer Nat. 106, 293-310.
Fagerstrom, T. and Asell, B. (1973) . Ambio 2_, 164-171.
Fagerstrom, T. and Asell, B. (1974). In: "Proceedings of the
International Conference on Transport of Persistent Chemicals
in Aquatic Ecosystems." pp IV:11-16, May 1-3, 1974, Ottawa.
Goodman, G.T. and Roberts, T.M. (1971). Nature 231, 287-292.
Holzman, R.B.C. (1968). In: "Radiation Protection Pt 2."
(W.S. Snyder, ed) pp 1087-1096. Pergamon Press, London.
Hutchinson, T. (1975). In: "International Conference on Heavy
Metals in the Environment." p. C-316 (abstract only) Oct. 27-31,
1975, Toronto.
Jernelov, A. (1974). In: "Conference on Intoxication due to
Alkyl Mercury Treated Seed, Baghdad, Iraq." Nov. (Abstracts
published by Swed. Wat. Air Pollut. Res. Lab., Stockholm).
John, M.K. (1975). In: "International Conference on Heavy
Metals in the Environment." p. C-233 (abstract only) Oct. 27-31,
1975, Toronto.
Leland, H.V. and McNurney, J. (1974). In: "Proceedings of
the International Conference on Transport of Persistent Chemicals
in Aquatic Ecosystems." pp 111-17-23, May 1-3, 1974, Ottawa.
Lundstrom, H. (1971). "Environmental Research." National
Swedish Environment Protection Board. (In Swedish).
Morris, A.W. and Bale, A.J. (1975). Estuarine Coastal Marine
Sci. 3^ 153.
Peterson, P.J. (1971). Sci. Prog. Oxf. 59, 505-526.
Pielou, E.G. (1969). "An Introduction bo Mathematical
Ecology." Wiley-Inter Science, New York, VIII + 286 pp.
Schroeder, H.A. and Balassa, J.J. (1961) . J. Chron. Dis. 14,
236-258.
220
-------
Tyler, G. (1975). In: "International Conference on Heavy
Metals in the Environment." p. C-79 (abstract only) Oct. 27-31,
1975, Toronto.
Webber, L.R. and Beauchamp, E.G. (1975). In: "International
Conference on Heavy Metals in the Environment." p. C-60
(abstract only) Oct. 27-31, 1975, Toronto.
Westermark, T., Forberg, S. and Odsjo, T. (1973). "Preliminary
Studies on the Occurrence of Mercury, Indium, Cadmium and
Arsenic in Plants and Animals in the Vicinity of a Smelter."
Report from the Museum of Natural History, Stockholm. (In Swedish)
Westermark, T. and Sjostrand, B. (1972). Adv. Activat. Anal.
2_, 57-88.
Whitby, L.M. (1975). In: "International Conference on
Heavy Metals in the Environment." p. C-121 (abstract only)
Oct. 27-31, 1975, Toronto.
221
-------
STANDARDS AND CRITERIA
Lars Friberg and Velimir Vouk
There is a need for administrative regulations to prevent
adverse health effects among people exposed to chemical or
physical agents in the general as well as the industrial
environment. Such regulations may have different legal
implications in different countries. For the time being they
will be referred to as standards.
The aim of environmental health standards is to protect
individuals, human populations and their progeny from the
adverse effects of hazardous environmental factors, including
chemicals.
A sound principle of health protection is to keep all exposures
as low as reasonably achievable, subject to the conditions
that the appropriate exposure limits, defined by the standard,
are not exceeded. The relationships between human disease
and exposure to chemicals are complex and poorly understood.
Death and disease are only the extreme end of a spectrum of
biological changes which include a variety of "adverse" and
"non-adverse" effects.
Environmental health standards for chemicals may be formulated
either in terms of concentrations in environmental components
(examples are air quality standards, "threshold limit values"
and drinking water standards) or in terms of amounts of
substances that may be taken into the body within an appropriate
unit of time (e.g. acceptable daily intake). These con-
centrations and amounts should be sufficiently low so that the
threshold dose (if it exists and can be determined) will not
be reached, or so that the population of concern will not be
subjected to "unacceptable risk" even following life-time or
working life-time exposure. In some cases, as for irritant
air pollutants, the distribution of exposure concentrations
in time should also be considered.
223
-------
Standards may be also emission standards or effluent standards,
defining permissible concentrations or amounts of a substance
in stack gases or industrial or municipal waste waters. Setting
up emission standards may be one way to arrive at permissible
concentrations of toxic metals in fuels, for example, coal and
oil. Standards may limit the concentration of toxic substances
in products used on a large scale, for example lead compounds
in paints, or they may prescribe the quantity of a substance to
be used, handled or transported at any time, and the manner of
its use, handling or disposal.
It is obvious that the standard setting process will necessarily
involve many considerations besides toxicology. Social,
cultural and economic considerations should be taken into account
in setting standards, but never to the detriment of health pro-
tection which should be of primary concern. The standard setting
process is often very different in different countries and
different types of society, and with few exceptions, such as
food standards, there are no international agreements on
standards for chemical substances. In general, however, it
involves appraisal of toxicological data, particularly of
dose-response relationships, and of information on the effects
on non-human targets (plants, animals, materials); social and
economic analysis, policy analysis and review of experience
elsewhere, leading eventually to an administrative or policy
decision concerning the standard. Other relevant questions include
the technological feasibility of achieving a standard, cost
and benefit of implementing it, means of enforcement, other
public health priorities, etc. Many of these topics are outside
the scope of the present discussion.
Standards are necessary for toxic metals, often more so than
for other chemical substances. Metals are not disintegrated
in nature and may be transported over great distances .
Furthermore, they have a tendency towards accumulation in
nature as well as in the human organism.
224
-------
Theoretically an ultimate goal might be to ban the use of
highly toxic metals. For several reasons this is not possible.
Metals are indispensable in a technological society. Furthermore,
highly toxic metals often occur together with nontoxic metals
in nature. Cadmium, for example, is found together with zinc
in a certain proportion in ores and it would be impossible
to ban cadmium without taking severe measures concerning the
use of zinc as well. Some metal compounds have been success-
fully prohibited under certain circumstances, however.
Several governments have, for example, banned the use of
methylmercury as a fungicide. In some countries, the use of
tetra ethyl lead as an additive to motor gasoline, previously
widespread, has now been forbidden. Generally, there is a
tendency to reduce the concentrations of lead in gasoline by
means of legislative measures. Cadmium and lead have been
widely used as color pigments. In several countries severe
restrictions on this practice are now being enforced.
Thus, with a few exceptions we have to live with some exposure to
metals both industrially and in the general environment. A
major problem will be to define which exposure can be accepted
with negligible risks or with only a minor, defined risk.
Such risk evaluations may refer to effects on the environ-
ment as well as on humans. The present discussion will focus
on the effects on humans only. Basically, however, the same
principles apply when evaluating effects on the environment.
For certain metals such risks may be the critical ones,
meaning that they should receive priority from the standpoint
of prevention and control.
There are different ways to control exposure. One approach
is called the best practical means approach while another
focusses on limiting concentrations known or suspected to be
associated with health effects.
The best practical means approach implies that an industry
installs the best available technique to reduce emission of
hazardous substances, taking into consideration, however,
225
-------
technical and economical aspects. What is meant by "best
practical means" is often a question of debate. This has to
do with economic considerations and the priorities given to
different forms of health protection. It also has to do with
differences in philosophy of risk acceptance.
If the best practical means approach is used in a proper way
it brings the advantage that resulting concentrations in
different media, e.g. in ambient air and soil, around a
point source will be as low as possible with the available
cleaning technique. Under certain circumstances this might
mean that e.g. ambient air concentrations will be considerab-
ly lower than would be necessary from the toxicological
point of view. Large safety margins may thus sometimes be
achieved. The approach has obvious drawbacks as it does not
take into consideration, at least not explicitly, information
on dose-effect and dose-response relations for hazardous
substances. There is thus no assurance that the best practical
means approach will bring hazardous substances in the different
media down to levels which can be considered acceptable
from the toxicological and epidemiological viewpoint.
Standards based on health effect evaluations have in prin-
ciple existed for many years due to the efforts of several
organizations and regulatory agencies. Well-known is the
work carried out in the USA by a private group belonging to
the Association of Governmental Industrial Hygienists.
Threshold limit values recommended by this group certainly
have been of value both nationally and internationally in
the absence of more detailed evaluations. During recent
years agencies like US Environmental Protection Agency and
US National Institute of Occupational Safety and Health have
enbarked on large programs for working out criteria documents.
These are basically an evaluation of dose-effect and dose-
response data for different substances, including metals.
The World Health Organization has given considerable atten-
tion to standards and criteria for the quality of the environ-
226
-------
ment during the last two decades. The work has involved
drinking water, food additives and contaminants, pesticide
residues, carcinogenic risks of chemicals, and air quality.
In 1973 the World Health Organization embarked on an inter-
nationally coordinated program on the assessment of health
effects of environmental conditions. This program, in which
actively participate more than 20 WHO member states, is
called the WHO Environmental Health Criteria Programme and
was initiated in response to a number of World Health Assembly
resolutions. It also took into consideration the relevant
recommendations of the UN Conference on the Human Environ-
ment in Stockholm, 1972, and of the Governing Council of the
United Nations Environment Programme (WHO, 1976).
When selecting priorities of the program and preparing criteria
documents the following considerations are taken into account:
a) severity of adverse effects on the population, in
particular, whether there are irreversible or chronic
effects, adverse genetic implications, or embryotoxic
or teratogenic effects;
b) ubiquity and abundance of the agent in man's environment,
particularly those occurring naturally or produced
inadvertently;
c) persistence of the agent in the environment, resistance to
environmental degradation and accumulation in man or in
food chains;
d) possibility of metabolic degradation or synthesis in
biological systems which may produce metabolites either
more or less toxic than the parent compound;
e) size, type and demographic characteristics of exposed
populations; the frequency and magnitude of exposure
particularly of highly vulnerable groups of the popu-
lation.
227
-------
The full list of priorities for the criteria documents embraces
some 70 chemicals and/or classes of chemicals as well as
physical hazards. The different criteria documents are now
in different stages of progress. Criteria documents for some
metals, such as mercury and lead have already been published
(WHO, 1976, 1977). For other metals the editing is almost
complete and publication will take place during 1978. This
applies to cadmium, manganese, tin and organotin compounds and
titanium. For certain of the substances, e.g. cadmium, lead,
and mercury, it has been decided to have a reevaluation
within the next three years.
One of the main objectives of the WHO Environmental Health
Criteria Programme is to assess existing information on the
relationship between exposure to environmental pollutants
(or other physical and chemical factors) and human health,
and to provide guidelines for setting exposure limits consistent
with health protection, i.e. to compile environmental health
criteria documents. Of importance is that the assessments
referred to should be made in relation to man's total exposure
to a given agent, from different media or conditions, for
example, food, air, water, work and the home. In the past,
when evaluating risks, the to eel exposure was not generally
considered. The inadequacy of this approach was apparent for
pollutants that may reach man by several pathways as is the
case with lead, cadmium and some other metals and certain
persistent organic compounds.
Cadmium, for example, accumulates in the body over the
entire life span. The most important exposure to cadmium for
those not industrially exposed is through food, which in
turn has been contaminated through deposition of cadmium
from the ambient air or the use of cadmium-containing ferti-
lizers, including sludge. For smokers, however, the smoking
of 20 cigarettes a day will alone give rise to an accumu-
lation of cadmium in the body comparable with the accumu-
lation from cadmium in food. For workers in certain indu-
stries the most important exposure will be from cadmium in
workroom air.
228
-------
The developing of a criteria document for a substance means
evaluation of dose-effect and dose-response relationships.
Dose-effect relationships indicate how an effect may in-
crease in severity with increasing exposure. A dose-response
relationship shows how the number of persons afflicted
varies with increasing exposure (see Chapter 00).
In evaluating dose-response curves for chemical substances,
including metals, it is usually implied that there is a
threshold below which no adverse effects will occur. On
purely statistical grounds it is as a rule not possible to
determine with any degree of certainty a threshold, i.e. a
no-effect level for humans. One reason is the great variability
of susceptibilities (tolerances) of individuals in human
populations, and the size of populations at risk which are
often very large compared with the populations on which the
information on health effects has been obtained. Furthermore,
no-effect level may vary depending on the conditions of
exposure and on the presence or absence of various factors
that may modify the response (e.g. other chemicals, heat,
radiation, etc.). For this reason no-observed-effect levels
are divided by safety factors.
Most regulatory authorities rely on the use of safety factors
but there are no precise guidelines for deciding the appropriate
size for such factors. In general, the size of the safety factor
will depend on (a) the nature of toxic effects, (b) the size
and type of population to be protected, and (c) the quality
of toxicological information available. A factor of 2-5 may
be considered as sufficient if the effect against which pro-
tection is desirable for individuals or a population is, for
example, lung or skin irritation, if only a small number of
workers are likely to be exposed and if the toxicological
information is derived from human data. On the contrary a
safety factor as large as 5,000 may be required if the
possible effect is very serious (e.g. cancer), if the general
population is to be protected, and if the toxicological data
are derived from limited experiments on laboratory animals.
229
-------
In some cases, the safety factor may be a value that has been
used with reasonable success and is, therefore, perpetuated.
In developing rules for protection against ionizing radiation,
the assumption has been made that thresholds do not exist
and that any exposure may involve some degree of risk. This
philosophy has nowadays been adapted in many quarters also
for carcinogenic and mutagenic effects of chemical substances,
including metals. The question of the form of the dose-
response curve for carcinogenic substances was discussed in
detail at an international symposium at the Karolinska
Institute in Stockholm, March 1977 (Task Group on Air Pollution
and Cancer, 1978). The meeting was arranged to evaluate
risks of cancer from carcinogenic substances, particularly
substances in the ambient air derived from a combustion of
fossil fuels. The conclusions reached, however, have implications
for the evaluation of carcinogenic risks more generally, in
the general as well as in the industrial environment.
One important conclusion reached with regard to risk assess-
ment methodology for carcinogenic substances was, "In con-
sidering protection of human populations, and in the absence
of firm evidence to the contrary, it is not justified to
assume a "threshold", i.e. that there is a dose below which
no response is obtained." It was also stated, "In the ab-
sence of relevant dose-response data the most appropriate
way to estimate the risk of lung cancer is to assume that it
will be directly proportional to the increase in dose. For
small added doses, a simple linear dose-response curve, as
used in radiation carcinogenesis, is appropriate." The
conclusions mean that basically the same philosophy was
adopted as the one long used in radiation protection, namely
the assumption that thresholds do not exist and that any
exposure may involve some degree of risk. Similarities with
ionizing radiation were considered as best founded in regard
to direct (ultimate) carcinogens (not requiring metabolic
activation) acting on directly exposed tissues.
230
-------
The conclusions of the Karolinska Institute symposium also
mean that, one could apply the concept of collective dose in
risk assessment relative to carcinogenic substances that
are persistent in the environment and which are ultimate
carcinogens (direct carcinogens), e.g. metals. The collective
dose can be defined as the product of the average dose to
an individual in a defined exposed group, called "per caput"
dose, and the number of individuals in that group (ICRP, 1977).
The use of the collective dose (i.e. the simple summation of
doses) in the evaluation of the total risk to a population is
valid only under the assumption that there within the range of
exposure conditions considered is a linear relationship without
threshold between dose and the probability of an effect. In
radiation protection, the collective dose is expressed in
man.rad units. For chemical carcinogenesis, the unit could
be man.mg/kg or in case of carcinogenic air pollutants man.ng/m
(if it is assumed that dose is proportional to the concentration
of a carcinogen in air).
If the exposure is extended in time, the concept of dose
commitment becomes useful. The dose commitment is defined as
the time integral of the per caput dose rate. The collective
dose commitment is obtained by integration of the collective
dose rate. For chemicals with a tendency to accumulate in the
organism and in the environment, such as cadmium, the dose
commitment approach might be an appropriate way of assessing the
risk from a practice leading to the release of the substance
into the environment.
Complete information concerning dose-effect and dose-response
relationships for humans is not available for any single
metal. It is therefore necessary to rely on scanty data when
assessing health risk. Sometimes there are no data on humans
at all and animal data must suffice. This makes every evalua-
tion all the more complicated. Animals may react quite
differently, partly in terms of their sensitivity to a
particular substance and partly in terms of their metabolism
of the substance. This may refer to deposition as well as
231
-------
absorption, biotransformation, accumulation and excretion of
the substance. As an example, it can be mentioned that
methylmercury is a very toxic form of mercury in humans. It
is highly stable and is accumulated in the body as methyl-
mercury. In rats, this compound is readily broken down into
inorganic mercury and the effects are less severe.
What has been said does not imply that animal experiments
are of no value. For studying mechanisms of action the
opposite is definitely true. When screening for potentially
carcinogenic or mutagenic substances, animal experiments are
of tremendous value. One has to be aware of the limitations
of the methods, however, and also appreciate that it is very
seldom possible for an animal experiment to reproduce a
real-life human exposure situation. Animal experiments will
have to be used as a basis of risk assessment in the case of
new chemicals including, e.g. organometallic compounds, for
which there are no data regarding human exposure.
For metals very few published, long-term studies on humans
are available, in which reliable exposure data are reported
in conjunction with observed effects. There are more studies
where effects and response have been related to concentra-
tions of metals in some biological indicator media, e.g.
blood or urine. Such evaluations have the advantage of
allowing the use of information from other exposure situa-
tions than those related to inhalation. For example, most
information of importance for environmental health criteria
for mercury and cadmium comes from studies where humans have
been exposed to the metal through contaminated food.
A prerequisite for using data on dose-response relationships
based on concentrations in indicator media and resulting
health effects is knowledge of the metabolic model for the
metal. This means the knowledge of parameters related to
absorption, biotransformation, distribution, accumulation
and excretion of the substance. For certain substances the
metabolic model is well defined and simple. Methylmercury is
232
-------
absorbed to almost 100% whether ingested or inhaled and the
accumulation within the body follows a one-compartmental
model with an average biological half-time of about 70 days.
If the daily exposure of the metal is known, it is therefore
possible to use simple equations to estimate the accumula-
tion of mercury. Data from epidemics of mercury poisoning in
Japan and in Iraq have been reasonably good as far as elucidating
associations between concentrations in indicator media (hair
and blood) and effect and response is concerned. Based on
such data and the metabolic model it has been possible to
come up with environmental health criteria for methylmercury
for ambient air, industrial air and food.
For cadmium, data based on concentration of cadmium in
kidneys, together with a metabolic model, have been of value
in calculating acceptable concentrations in air, food and
water. The metabolic model for cadmium is not as well-known
as would be desirable, however, and environmental health
criteria based on the use of such a model are therefore
somewhat more uncertain.
As mentioned in the introduction, standards as a rule refer
to concentrations in different environmental media, e.g.
ambient air, food, water, and industrial air. There is a
need also to come up with standards for metals in biological
media, like blood, hair, and urine. In practice such standards
are often used, e.g. lead in blood, cadmium in urine and
mercury in blood or hair. Before setting such standards it
is imperative to examine in detail what the indicator values
really mean. For certain substances such values may be of no
assistance at all; for others they may reflect recent
exposure, body burden and concentrations in critical organs.
For methylmercury there are definite possibilities of using
concentrations of mercury in blood or hair. As mentioned,
such concentrations and knowledge of the metabolic model
have been the starting point for evaluating risks from
exposure via air as well as food. For cadmium the situation
233
-------
is much more difficult. During a long-term low-level expo-
sure urinary cadmium values are related to concentrations in
kidney and total body burden. On the other hand high urinary
cadmium values may be found when kidney dysfunction has
occurred. High cadmium levels may therefore be a result
of cadmium intoxication. Cadmium values in blood will first
of all reflect recent exposure, perhaps the exposure during
the last few weeks. As such, blood values may be of substantial
value. If exposure ceases, however, cadmium concentrations
in blood will decrease more rapidly than the cadmium level
in the body as a whole. One may then find low blood values
even though the concentrations in the critical organs, the
kidneys, could well be high.
The situation for lead is also complicated. Most of the
absorbed lead, about 90%, is found in the skeleton, where
it is in an inert form. The lead that constitutes the direct
danger to health is in the blood and bone marrow. As the
biological half-time in blood is about 20 days, while the
half-time in the bones may be 10-20 years, it is easily
understood that blood concentrations do not reflect the
accumulated lead. Even so, the monitoring of blood levels is
of great value in preventing clinical effects because of the
known relation between concentrations in blood and effects.
Unfortunately there are not very many data which show how
air lead levels relate to blood lead levels despite the
several years of industrial monitoring of both air and blood.
The work within the WHO Environmental Health Criteria Pro-
gramme has clearly shown that substantial information is
available for several metals. It is also obvious, however,
that available data from toxicological and epidemiological
studies do not necessarily contain information which would
make it possible to come up with a detailed evaluation of
dos.e-ef f ect and dose-response relationships . There is an
obvious need for more intense research directed towards
information needed for a health risk evaluation. Furthermore,
it seems to be extremely important that information on
234
-------
studies carried out be made available to a greater extent
than is now the practice. This has been pointed out explicitly
by the Task Group on Metal Accumulation (1973) and The
Task Group on Metal Toxicity (1976). The Task Group on Metal
Accumulation said, "There is a lack of epidemiological data
on the parameters discussed above in populations with industrial
exposure. This lack is even more striking in relation to the
growing number of such studies on populations with general
environmental exposure, which is usually at a much lower
level. There is a need in industrially exposed populations
for standardized, wherever possible collaborative, epidemio-
logical studies, where cohorts can be followed in time and
where groups can be related to each other. With some occupa-
tional exposures to the less common metals, only small
groups may be available for study in any one country, so
that international collaboration in epidemiological studies
would again be of value."
235
-------
REFERENCES
ICRP (1977). Radiation Protection, Recommendations of the
International Commission on Radiological Protection, ICRP
Publication 26 (In press).
Task Group on Air Pollution and Cancer - Risk Assessment
Methodology and Epidemiological Evidence. (1978). Report
from an International Symposium at the Karolinska Institute,
Stockholm, 1977. Environ. Health Perspect,. (In press, 1978).
Task Group on Metal Accumulation (1973) . Environ. Physiol.
Biochem. JB, 65-107.
Task Group on Metal Toxicity (1976). In: "Effects and Dose-
Response Relationships of Toxic Metals." (G.F. Nordberg, ed)
pp 1-111. Elsevier, Amsterdam.
WHO (1976). Background and purpose of the WHO Environmental
Health Criteria Programme. In: "Environmental Health Criteria
1: Mercury." pp 5-14.
WHO (1977). "Environmental Health Criteria 3: Lead." pp 3-
160.
236
-------
CLINICAL EFFECTS, DIAGNOSIS AND TRE VTMENT OF METAL POISONING
GENERAL CONSIDERATIONS
George Kazantzis
1. Clinical effects
1.1 General considerations
Clinical effects following the absorption of a toxic metal
are manifested by abnormal signs accompanied by symptoms, and
this pattern of signs and symptoms may comprise a syndrome. A
syndrome may be acute, terminating in death or followed by
partial or complete recovery. It may be subacute in onset or
there may be a longer latent interval between absorption and
its development, which may then follow the same course as
above. Such a syndrome is usually referred to as a "poisoning",
it being customary, for example, to refer to "heavy metal
poisoning". However, toxic metals produce other important clinical
effects not usually included under this term. These effects are
(i) hypersensitivity reactions which may involve the skin,
lung, kidney, hemopoietic system and possibly the nervous
system; (ii) the induction of cancer after an appropriate
latent interval; and (iii) teratogenic malformations. Furthermore,
an increasing number of metals are being shown to give rise
to genetic damage in test systems and the question arises as
to whether such damage may also occur in man, with a con-
sequent increase in abnormal individuals in later generations.
These other"effects, because of their importance and magnitude,
are considered elsewhere.
Adverse effects, which may be manifest by biochemical or
physiological tests or even by abnormal clinical signs, may
be present in the absence of symptoms or in the presence of
vague, ill defined complaints of ill health which do not comprise
a recognizable syndrome. Common examples are anemia following
exposure to lead compounds and renal tubular dysfunction following
long-term exposure to cadmium.
237
-------
The clinical effects described above imply that an adverse
biological change has been produced in the organism with some
impairment of cellular function resulting from exposure. The
model on which this adverse effect is based, assumes that a
critical concentration of the toxic metal has been attained in
the critical organ to give rise to a critical effect. This
critical effect nay or may not be of immediate importance to
the health of the organism as a whole (Nordberg, 1976). It may
be argued that such a critical effect cannot be considered to be
synonymous with a clinical effect. Here it is useful to
distinguish between critical and subcritical effects. Effects
may also occur below the critical concentration in the critical
organ which do not impair or appear to impair cellular function
(Nordberg, 1976). While the biological significance of such a
subcritical effect may be unknown, it would be reasonable, given
the existing state of knowledge, to exclude such a subcritical
effect from clinical consideration.
1.2 Exposure pattern and clinical effect
A number of factors act as determinants of clinical effect
following exposure to a toxic metal. Such factors include dose;
route of absorption; the chemical and physical form ot the
metal concerned; genetic variation manifested through racial,
familial and individual susceptibility; dietary pattern and
nutritional status; immunological status; presence of inter-
current disease and interaction with other chemical compounds
in the environment. Some of these determinants are considered
in the chapter on Factors modifying dose-effect and dose-
response relationships in this volume and in Volume II of
this series, on Specific metals. The exposure pattern, in
terms of concentration, time and route of exposure is an
important determinant of clinical effects which required
further consideration here.
A jhort-term exposure may produce a very different clinical
effect from a similar exposure in terms of total dose over a
longer period of time. The effects of ingestion of a toxic
metal are more often seen in the domestic or general environment
than in an industrial setting. Short-term, high-level exposure
by ingestion may follow the accidental, suicidal or homicidal
238
-------
ingestion of a toxic metal compound, giving rise to well
recognized acute syndromes, usually involving the gastro-
intestinal tract, but which may involve secondarily the
renal, cardiovascular, nervous and hemopoietic systems.
Long-term, low-level exposure by ingestion is seen increasingly
in the general environment as a result of the contamination of
food and drink by metals which have cumulative properties in
the organism. Clinical effects may involve any organ system in
the body, but the gastrointestinal tract is not primarily
involved.
By contrast, short-term, high-level inhalation exposure is
most often occupational in origin. It may give rise not only to
acute respiratory effects, but may also involve the cardio-
vascular, central nervous, renal and hemopoietic systems. Again,
long-term, low-level inhalation exposure is usually occupational
in origin and its control forms a large part of industrial
hygiene practice. However, long-term, low-level inhalational
exposure to certain toxic metals may also occur in the general
environment and from cigarette smoking. The effects may
involve any organ system in the body and may spare the respiratory
system.
Mercury forms a good example of the extreme variation in
clinical effect which may be produced depending on the pattern of
exposure and the chemical form of the metal. The lung is the
critical organ following short-term, high-level inhalation
exposure to mercury vapor, resulting in pneumonitis and re-
spiratory failure, while the central nervous system is the
critical organ following long-term exposure to mercury vapor.
Following the ingestion of an inorganic soluble mercuric
compound, the kidney is the critical organ, manifesting with
anuria due to tubular necrosis. As a result of the long-
term ingestion of methylmercury as a food contaminant, nervous
system effects may develop but with a different clinical
syndrome to that seen after the long-term inhalation of in-
organic mercury vapor.
239
-------
The variability in clinical effect produced by the toxic
metals is further illustrated in the following section.
1.3 Acute clinical effects of metals
1.3.1 Gastrointestinal effects
Acute gastroenteritis follows the ingestion of a sufficient
quantity of most metals in the form of soluble salts. A not
uncommon occurrence is the contamination of food or drink,
especially if acidic, by dissolution of metal from food
containers. Symptoms develop a short time after ingestion,
often involve a number of people partaking of the meal and may
be labelled as "food poisoning". Vomiting and diarrhea may be
followed by circulatory collapse and involvement of other
systems depending on the poison absorbed. Poisoning with
soluble compounds of antimony, cadmium, copper, lead, tin and zinc
has occurred in this way. Acute gastroenteritis with collapse
may be the predominant features following the ingestion of
rodenticides containing arsenic, bariurr or thallium, yellow
phosphorus or zinc phosphide. Similar symptoms may follow
the ingestion of soluble compounds of bismuth, chromium, iron,
silver and vanadium. The ingestion of a soluble mercuric salt
gives rise to a gastroenteritis with a bloody diarrhea which may
resemble fulminating ulcerative colitis. Lead colic has on many
occasions simulated an acute surgical emergency resulting in an
unnecessary laparotomy.
1.3.2 Respiratory effects
Acute chemical pneumonitis which may be accompanied by
pulmonary edema follows the inhalation of a number of
freshly formed metal fumes. Particularly toxic in this respect is
the inhalation of freshly formed cadmium oxide fume, acute
symptoms developing some hours after an apparently innocuous
exposure (Beton et al., 1966). The inhalation of antimony
pentachloride, arsine, beryllium fume, iron pentacarbonyl,
lithium hydride, nickel carbonyl, titanium tetrachloride,
selenium dioxide, hydrogen selenide, vanadium pentoxide or zinc
chloride can give rise to a similarly acute picture with
pulmonary edema. The inhalation of high concentrations of
240
-------
mercury vapor or dust or inorganic mercury compounds can also
give rise to pneumonitis before other symptoms of mercurialism
develop (Teng and Brennan, 1959). Respiratory symptoms with
rigors and fever resembling an acute respiratory infection
may follow the inhalation of freshly formed zinc fume, brass
fume or other metallic oxides, giving rise to metal fume fever
(Hunter, 1975). Pneumonic consolidation has followed the
inhalation of manganese dust. The inhalation of complex
salts of platinum can give rise to acute and severe asthma.
Decomposition products or other contaminants of industrial air
may be responsible for some of these effects.
1.3.3 Cardiovascular effects
A number of metallic ions interfere with the normal function
of myocardial cells giving rise to arrhythmias including
ventricular fibrillation. Fibrillation may be responsible for
a fatal termination in poisoning by antimony, barium or
lithium salts. Cobalt can give rise to cardiomyopathy (Alexander,
1972). Some metals have been shown to have a hypotensive effect.
These include antimony, cadmium, cobalt, copper, iron and
vanadium, a state of shock being a common presenting feature
in poisoning with these metals.
1.3.4 Effects on the central nervous system
Metal poisoning may present with an acute illness involving
the central nervous system. Most important, because still
unfortunately not uncommon in children, are convulsive attacks
which may terminate in coma or death as a result of acute
lead poisoning. A successful therapeutic outcome in such a
case is dependent on early diagnosis and treatment. Convulsions
may also follow the absorption of iron, barium, lithium, thallium
and organic tin compounds. An acute psychosis may also be the
presenting feature in metal poisoning. Following heavy exposure
to tetraethyllead a patient may present with delusions,
hallucinations and hyperactivity which may precede coma and death
(Beattie et al., 1972).
241
-------
1.3.5 Renal effects
Renal damage, manifesting as oliguria or anuria due to acute
tubular necrosis is another way in which metal poisoning
may present, although this feature often follows an
initial presentation with acute gastrointestinal, circulatory or
respiratory effects. Oliguria and anuria due to tubular
necrosis are common occurrences especially in children,
following the ingestion of soluble mercuric or iron salts.
The condition may also follow pneumonitis resulting from
cadmium fume inhalation (Beton et al., 1966). Renal failure
has been a sequel to the absorption of a number of soluble
metal compounds including antimony, arsine, bismuth, copper,
uranium and vanadium salts, particularly if there is associated
short or intense renal vasoconstriction.
1.3.6 Hemopoietic effects
An acute hemolytic anemia often accompanied by renal failure
may be the presenting feature following the inhalation of arsine
or of stibine gas (Jenkins et al., 1965). Acute hemolysis has
also followed the ingestion of large doses of soluble copper salts.
1.4 Chronic clinical effects of metal toxicity
The term chronic effect is a relative one, signifying that the
clinical effect may develop gradually and may persist for a longer
interval than an acute effect. Chronic effects are associated with
longer term exposure at lower levels than acute effects. Only
examples of chronic effects will be given here, the great variety
of these being described in the individual metal chapters.
1.4.1 Chronic gastrointestinal effects
Some chronic gastrointestinal effects may in fact be examples
of mild acute poisoning following the repeated ingestion of
the toxic metal over a period of time. Thus prolonged diarrhea
with weight loss occurred in a child whose drinking water was
heavily contaminated with copper. Similarly chronic gastro-
intestinal symptoms have occurred in persons ingesting canned
juice contaminated with high concentrations of tin or of
zinc. However intestinal colic has been observed in industrial
242
-------
workers and in children with relatively low-level lead
exposure, and this can be considered a true example of chronic
lead poisoning (Beritic, 1971). Anorexia, nausea, diarrhea or
constipation with weight loss are seen in chronic arsenic
poisoning but are usually accompanied by neurological and
skin involvement. Anorexia, nausea and vomiting in industrial
workers may be seen in selenium and in tellurium poisoning,
when the characteristic garlic odor of the breath may also
be elicited. Anorexia and often digestive disturbances have
also been reported in people living in seleniferous regions
with a high dietary intake of selenium. Anorexia, nausea,
vomiting and diarrhea followed by constipation accompanied
by a burning sensation of the tongue and stomatitis have
resulted from occupational exposure to thallium compounds.
1.4.2 Chronic hepatic effects
A number of metals are hepatotoxins, giving rise to effects
ranging from abnormalities in hepatic enzyme levels to clinical
jaundice. Such effects have been reported following exposure to
antimony, arsenic, bismuth, copper, chromium, iron, manganese and
selenium.
1.4.3 Chronic respiratory effects
Chronic pulmonary disorders giving rise to dyspnea may result
from absorption of metal dust or fume. An inflammatory response
consisting of noncaseating granuloma formation followed by
pulmonary fibrosis may result from exposure to beryllium after
a latent interval of many years (Stoeckle et al., 1969). Progressive
dyspnea with the clinical, radiological and functional changes
associated with emphysema is seen in workers exposed to cadmium
oxide fume or dust. The respiratory abnormality is usually,
but not invariably, associated with renal tubular dysfunction.
Pulmonary fibrosis has followed occupational exposure to
finely powdered aluminum metal dust. Dyspnea, wheeze and
productive cough giving rise to an illness resembling asthmatic
bronchitis have followed occupational exposure to vanadium
pentoxide dust. Progressive breathlessness associated with
pulmonary fibrosis has occurred in hard metal workers,
243
-------
exposed to the dusts of tungsten and titanium carbides with cobalt
used as a binder. Chronic asthma can occur in workers
sensitized to chromium, following the inhalation of chromate
dust or chromic acid fume and also to organic platinum
compounds. Once sensitization to platinum has occurred, the
reaction can be triggered by exposure to minute quantities of the
salts of chloroplatinic acid.
1.4.4 Chronic effects on the nervous system
Peripheral neuropathy may develop in the recovery stage
of acute arsenic intoxication, about one to three weeks after
exposure. It is a mixed motor and sensory neuropathy, with a
"glove and stocking" distribution. Neuropathy develops in those
who survive the acute gastrointestinal effects of thallium
poisoning and may lead to a later fatal termination. With both
these metals skin changes occur at a later stage, with the former
metal, arsenical pigmentation, and with the latter, hair loss. A
motor neuropathy involving predominantly the upper limbs with
wrist drop and extensive weakness of the fingers is seen in
chronic lead poisoning. By contrast, antimony salts of organic
acids give rise to a sensory neuropathy which may involve the
trigeminal nerve. Bismuth and copper have also given rise
to peripheral neuropathy.
Permanent brain damage with cerebral cortical atrophy or
hydrocephalus may be the sequel to acute lead encephalopathy.
Convulsions may recur over a long period, and idiocy may develop.
The complex relationship between moderately elevated body burdens
of lead in childhood and certain functional neurological disorders
is considered in the chapter on lead (see also Landrigan et al.,
1975) .
Degenerative changes in the nerve cells of the basal ganglia
giving rise to a parkinsonian syndrome result from the absorption
of manganese following long-term occupational exposure (Mena
et al., 1970).
Degenerative changes affecting in particular the granular cells
244
-------
in the cerebellum and neurones in the calcarine, pre- and
post-central cortex, follow the absorption of alkylmercury
compounds and present with a characteristic neurological syndrome
whose principal features are paresthesia of extremities and
face, ataxia, dysarthria and concentric constriction of the
fields of vision. Pyramidal signs may also occur. Another
characteristic neurological disorder consisting principally
of intention tremor of the hands, tremor of the eyelids and
tongue and a combination of behavioral and personality
changes known as erethism develops after chronic exposure
to mercury vapor.
1.4.5 Chronic renal effects
Certain chronic renal disorders may also follow exposure to
toxic metals. Proximal renal tubular dysfunction consisting
principally of tubular proteinuria, glucosuria, aminoaciduria
and phosphaturia may develop following cumulative exposure to
cadmium, lead and uranium compounds. Hypercalciuria has also
occurred in chronic cadmium poisoning as a further manifestation
of renal tubular dysfunction and this has led to renal stone
formation and to osteomalacia in a few cases following industrial
exposure (Kazantzis, 1975).
Osteomalacia has been observed in a population environmentally
exposed to cadmium in Japan. Although the skeletal disorder
(Itai-Itai disease) only occurred in postmenopausal women,
evidence of renal tubular dysfunction was found in the popula-
tion as a whole (Friberg et al., 1974) .
Progressive impairment in renal function terminating in
uremia and indistinguishable clinically from chronic nephritis has
been observed following childhood lead poisoning. The wide range
in abnormal response of the kidney to absorbed metals can be
further illustrated by reference to the nephrotic syndrome. Heavy
proteinuria with hypoproteinemia and edema has followed
exposure to inorganic mercury, organic gold and bismuth
preparations.
245
-------
1.4.6 Chronic hemopoietic effects
Chronic arsenic poisoning is associated with an anemia due
to decreased red blood cell formation with increased destruction.
The anemia of chronic lead poisoning also results from decreased
hemopoiesis combined with increased red cell destruction
(Goyer and Rhyne, 1973). By contrast, cobalt has an increased
hemopoietic effect, and has given rise to polycythemia, but
not to increased production of other cellular elements in the
blood.
2. The diagnosis of metal poisoning
As can be seen from the above paragraphs, metal poisoning may
involve any of the organ systems of the body and give rise to a
wide variety of effects. The presenting features may be
entirely non-specific, the clinical examination giving no lead
on the cause of the illness. The principal features which
should be considered in arriving at a correct diagnosis will now
be elaborated upon.
2.1 The history of exposure
In most cases a history of exposure to a toxic metal will
give the necessary clue. In an industrial situation there
may be a clear history of exposure which may be obtained from
patient, relative or co-worker. The clinician should not fail
to take a full and accurate occupational history where a
poisoning of occupational origin is suspected. It may be
necessary to seek additional information from the employer
or his agent, from an occupational hygienist or from a health
and safety representative. Exposure to a toxic metal may
however occur without this being suspected by any of the
persons questioned, and this can only be inferred by an
adequate knowledge of the work process involved. A good example may
be given with exposure to arsine gas, which usually presents
as an acute medical emergency. Arsenic may be a contaminant
in scrap metals, flue dusts etc. and arsine may be formed in
any reducing situation, or when a metallic dross containing
arsenic comes into contact with water. The condition should
therefore be suspected in a scrap metal worker presenting
246
-------
with the appropriate clinical features.
In the domestic environment an appropriate history may give
an exposure inferred from medicaments such as iron pills or
household chemicals. Unlabelled pills or chemicals should,
of course, be kept for analysis. In a homicidal instance, a
psychotic factory worker succeeded in poisoning three fellow
workers in succession by adding a thallium salt to tea, causing
the deaths of two men before the condition was suspected
(Cavanagh et al., 1974). In the general environment diagnosis
of metal poisoning, especially if caused by long-term
absorption of a cumulative metal in the body, can be very
difficult. Methylmercury poisoning in the environs of Minamata
Bay took more than five years to diagnose (Kazantzis, 1971).
However, methylmercury poisoning following the eating of
contaminated bread in Iraq was diagnosed at the beginning of the
outbreak, as the medical officers had previous experience of
the condition.
2.2 The clinical features
The presenting clinical features, especially in acute poisoning,
may be entirely non-specific. Lead colic has been repeatedly
C» 'tv,o
mistaken for the acute surgical abdomen, and eydiiluLomy has
been performed following encephalopathy due to lead. Such
mistakes would not have been made if a careful history had been
taken. Certain clinical sequences should alert the physician
to the possibility of metal poisoning. Gastroenteritis followed
by peripheral neuropathy which is in turn followed by hairfall
is a classical presentation for thallium poisoning. Gastro-
enteritis followed by oliguria or anuria follows the ingestion
of inorganic mercuric and other soluble metal compounds as
mentioned in the previous section. A hemorrhagic gastro-
enteritis, especially if accompanied by collapse and hypotension,
should alert the clinician to an acute metal poisoning by
ingestion. Similarly, an acute pulmonary edema in a subject
free of heart disease should raise the suspicion of a toxic
inhalational exposure. Acute cadmium fume poisoning should be
suspected especially if soldering or welding had been performed,
for iron and steel may be plated with cadmium. Metal poisoning
247
-------
should also be considered in the differential diagnosis of
every case of oliguria or anuria of unknown etiology. A
number of metals give rise to renal tubular dysfunction, in
particular cadmium, uranium and lead. With lead the condition
is seen more often in children. Lead poisoning should be
considered in every case of glucosuria presenting in childhood.
The causes of peripheral neuropathy are often obscure, and many
cases remain incompletely diagnosed. Metal poisoning should be
considered in the differential diagnosis of every case of
peripheral neuropathy.
Few of the signs described in this chapter are specific to metal
poisoning. A patient with a tremor does not necessarily have
mercury poisoning, even if young and working with mercury. The
tremor could be familial in origin, or could be due to other
conditions. There are however certain specific signs which are
indicative of metal absorption, though not necessarily of
poisoning. Among these should be mentioned the characteristic
garlic smell in the breath of selenium and tellurium workers,
the green tongue of the vanadium worker and the pigmented alveolar
margin, the lead line, due to the deposition of insoluble
sulfide in the gums of lead workers. It should be mentioned
that the lead line is not seen in edentulous workers, and that
other metals will deposit sulfide in the same site. A finely
mottled brown pigmentation of the skin, with leukodermia,
a "rain drop pigmentation" affecting in particular the
temples, eyelids and neck is seen in workers in contact with
arsenical dusts. Ulceration and perforation of the nasal
septum should also alert the clinician to previous long-term
exposure to arsenical dusts or to chromates. Chronic ulceration
on the fingers is seen following exposure: to washing soda or to
lime. White striae on the fingernails, Mees lines, are indicative
of previous exposure to arsenic.
2.3 Toxicological analysis
A diagnosis of metal poisoning can be confirmed in the acute
stage and often, but not always, in the chronic stage by
finding an increased concentration of the suspected metal in
248
-------
the appropriate medium. Here it should be mentioned that the
techniques involved in trace metal analysis in body fluids and
tissues present certain difficulties and such analysis should
only be performed in laboratories equipped for this purpose.
Environmental sampling may need to be performed on air, water
or food. Pills, residues in bottles and samples of vomit should
be kept for chemical analysis after an acute poisoning incident.
Blood, urine, feces and tissue samples may also require
chemical analysis. The examination of blood and urine samples
for metal concentration can be considered a routine procedure.
However, it should be borne in mind that some metals disappear
rapidly from the blood and others are sparingly excreted in the
urine. In thallium poisoning, for example, the diagnosis can
be confirmed by taking hair and nail clippings for analysis at
a stage where blood and urine concentrations are not helpful.
Evidence of methylmercury absorption can be obtained from
analysis of mercury content of red blood cells and from
hair samples, but methylmercury is only very slowly excreted
in the urine, and analysis of urinary mercury is not helpful. In
the diagnosis of lead poisoning difficulty may be caused where
exposure has ceased some months previously. In such cases blood
lead levels return to normal by the time diagnosis is attempted.
A calcium EDTA provocation test will increase the urinary
excretion of stored lead in such a situation.
2.4 Biochemical investigation
For most metal poisonings biochemical investigation is required
in addition to estimation of metal concentrations in body
fluids. The profile of biochemical abnormality is indicated in
the chapters on the different metals considered. In general,
biochemical tests of renal and of hepatic function are
always required. Biochemical data are essential for diagnosis
with certain toxic metals but are unhelpful with others. For
example, in the diagnosis of lead poisoning it is essential to
estimate 6-aminolevulinic acid and coproporphyrin in the urine
as well as lead concentration in blood. In contrast, the
diagnosis of poisoning by metallic mercury vapor or by alkyl-
mercury compounds is essentially clinical together with estima-
tion of metal concentrations in blood.
249
-------
2.5 Physiological investigation
Electrocardiography, respiratory function tests, electro-
encephalography and nerve conduction studies provide examples
of possible ancillary investigations for the diagnosis and for
the determination of the extent of poisoning by certain metals.
The principal disturbances of physiological function in such
poisoning are indicated in the chapters on individual metals.
3. Treatment
The treatment of metal poisoning includes treatment of the acute
and chronic effects, and suppression of tissue reaction where
sensitization has occurred.
The management of acute metal poisoning requires emergency
resuscitative procedures which may need to be initiated in the
work or home environment and continued in an acute treatment unit.
Those principles on which the management of the acute case
of poisoning are based, which are aspects of emergency general
medicine (Matthew and Lawson, 1975), will only be briefly
referred to here, while therapy related to metal poisoning
will be discussed in detail.
3.1 Prevention of further absorption
3.1.1 Removal from exposure
Absorption of a toxic metal may follow inhalation, ingestion or
skin or mucosal contamination. Occupational exposure most
frequently gives rise to absorption after inhalation. Following
high-level inhalational exposure, the victim should be removed
immediately from the contaminated atmosphere. Such an emergency
situation may arise with volatile metal compounds such as
stibine, arsine, alkylmercury, alkyltin and alkyllead
compounds. Where removal from exposure cannot be carried out
immediately, oxygen should be given. Following contamination of
the skin with lipid-soluble metal compounds, contaminated clothing
should be removed as soon as possible and the contaminated
area irrigated and washed, but not rubbed, with copious cold
water. Decontamination may also be required for hair and
fingernails.
250
-------
In chronic, cumulative metal poisoning, it may be sufficient
to remove the subject from further exposure. In certain
circumstances, for instance, a worker with mild poisoning
following exposure to inorganic mercury vapor or to lead
fume or dust, may require no therapy other than such removal.
The normal excretory mechanisms will ensure a gradual recovery
from mild toxic effects.
A review of possible routes of absorption in the individual
case should include consideration of the possibility of
exposure by contamination of food, drink, cigarettes or
clothing. Where this possibility exists, food, drink and
cigarettes at the work site should be prohibited, and
facilities provided for showering and for a complete
change of clothing after each shift.
3.1.2 Minimizing absorption from the gastrointestinal tract
3.1.2.1 Removal from the gastrointestinal tract
When the poison has been recently ingested some may be
recovered by emptying the stomach. In a fully conscious
patient this may be induced by vomiting. In an emergency
situation, the simplest procedure is by stimulation of the
back of the throat with a finger, or by giving a readily
available emetic in the form of a spoonful of salt in a
tumbler of water or syrup of ipecacuana, 30-45 ml for an
adult. Where medical help is available, apomorphine 2-8 mg
given intramuscularly produces rapid vomiting without
untoward side effects.
The stomach may be washed out in a conscious patient following
the passage of a gastric tube. If the cough reflex is absent, or
consciousness is impaired, a cuffed endotracheal tube should
first be passed to protect the lungs from inhalation of
gastric contents. The procedure of gastric lavage is a skilled
one which should only be performed in a properly equipped
emergency department. It is probably worth attempting even
if the poison had been swallowed several hours previously.
251
-------
Following the passage of the toxic agent through the
pylorus, absorption may be impeded by speeding its passage
through the gastrointestinal tract. This can best be achieved
by administering an osmotic purgative in the form of sodium
sulfate, 10-30 g.
3.1.2.2 Inactivation in the gastrointestinal tract
While gastric lavage can be effectively performed with warm
water, some poisons may be inactivated in the stomach by a
specific antidote, and following lavage this may be left in
the stomach in a volume not greater than 200-300 ml. However,
lavage should not be delayed if the antidote is not immediately
available. In an emergency situation egg white or milk will
help to inactivate mercury and other heavy metals by precipitation
in the stomach. Activated charcoal will effectively absorb poisons
still present in the stomach (Holt and Holz, 1963). It should
be given as a drink by mixing five to six five-mi spoonfuls
of the powder in a glass of water, and this should be followed
by aspiration 30 minutes later. Sodium bicarbonate solution will
precipitate soluble iron salts in the stomach, or iron may be
effectively chelated and thus inactivated with a solution of
desferrioxamine of which 5-10 g in 100 ml water may be left
in the stomach. Effective precipitation of the toxic barium
ion can be produced by oral administration of gastric lavage
with a 10% solution of sodium or magnesium sulfate. Following
the ingestion of chromic acid, lavage should be performed with
magnesium carbonate. Prussian blue will chelate thallium, and
following its ingestion a colloidal solution should be given
by duodenal tube at a rate of 250 mg per kg per day in divided
doses.
3.2 General supportive therapy
In acute poisoning, resuscitative measures have to take
precedence above all others. If the patient can be kept alive,
excretory and detoxicating mechanisms together with the ad-
ministration of specific antidotes, where these exist, will
be able to ensure eventual elimination of the poison. Careful
symptomatic medical care is always necessary, with an awareness
of delayed effects which may occur.
252
-------
3.2.1 Maintenance of respiration and circulation
Maintenance of respiration and circulation should receive
precedence over all other procedures. The patency of the
airway must be assured, especially at the site of the accident
and on the journey to the treatment center, and artificial
respiration and external cardiac massage may be necessary. At
the treatment center tracheobronchial toilet may be required.
The presence of hypoxia should be assessed by determining
minute ventilation or preferably arterial oxygen and carbon
dioxide tensions. Oxygen may be needed and also mechanical
ventilation. Treatment for acute pulmonary edema may be
necessary following exposure to beryllium or cadmium fume,
nickel carbonyl, or to hydrogen selenide.
In an emergency situation a conscious shocked patient with a
blood pressure below 80-90 mm Hg should be reassured, covered with
blankets and kept supine with the legs elevated. The blood
pressure may be raised as a temporary expedient, by giving a
vaso-pressor drug, e.g. metaraminol, 2-5 mg given slowly intra-
muscularly or intravenously and repeated at intervals to keep
the blood pressure about 100 mm Hg. Once the patient reaches the
treatment center, the circulatory blood volume can be restored
with suitable intravenous fluids. Blood transfusion may be
required for severe hemolytic anemia following exposure to arsine.
3.2.2 Maintenance of water and electrolyte balance
Imbalance may occur from vomiting, diarrhea, tissue damage or from
measures taken to eliminate the poison. It may be sufficient
to give fluids by mouth to prevent dehydration, but intravenous
infusion with appropriate biochemical monitoring is often required.
An adequate urinary flow should be ensured and a catheter may be
necessary.
Acute tubular necrosis may give rise to anuria following
inhalational exposure to arsine or following the ingestion of
mercuric or of uranium salts. Anuria has also followed poisoning
with bismuth, copper and iron salts. Careful water and electrolyte
balance has to be maintained until regeneration of the tubular
epithelium leads to recovery.
253
-------
3.2.3 Control of nervous system effects
The general supportive measures outlined above are usually
adequate for the management of the patient whose level of
consciousness is depressed. There is no indication for the
use of analeptic drugs in acute metal poisoning. Convulsions,
however, require management with a barbiturate or with diazepam
10-30 mg given parenterally. Cerebral edema may occur, as for
example in acute lead poisoning. This should be treated with
an osmotic agent, such as intravenous mannitol or a rapidly
acting diuretic such as frusemide, together with a steroid in the
form of hydrocortisone hemisuccinate or dexamethasone given
intravenously. Prolonged sedation with barbiturates or diazepam
may be necessary to overcome the hyperactivity and other behavioral
disorders which may develop in poisoning with alkyllead compounds.
3.3 Elimination of absorbed poison
3.3.1 Diuresis
The promotion of a diuresis will increase the clearance of
many poisons by decreasing their passive reabsorption from
the proximal renal tubules. Diuresis can be achieved by
giving a fluid load together with osmotic agents like mannitol
or urea or by giving a pharmacological diuretic, e.g.
frusemide. Furthermore, the excretion of some poisons is
influenced by the pH of the urine, for passive tubular
reabsorption is less effective with increased ionization of
the solute in the tubular fluid. While of great value in some
common forms of acute poisoning, these methods have some
limitations in poisoning by metals, although in general,
excretion of the toxic metal can be accelerated if a high
urine flow is maintained, as for example in acute inorganic
mercury and lead poisoning.
3.3.2 Biliary excretion
The action of certain toxic metals is prolonged as a result of
an enterohepatic circulation where excretion in the bile can
occur against a high concentration gradient followed by
intestinal reabsorption. Osmotic purgatives may be of some
254
-------
value in decreasing transit time in the intestine and some of
the poison may also be removed by means of duodenal intubation
or by gall bladder drainage. A promising approach for inter-
rupting this enterohepatic circulation has been described by
Clarkson et al. (1973). This involves a complexing agent
given by mouth which will bind with the metal compound
excreted in the bile, prevent reabsorption, and enhance fecal
excretion of those heavy metal compounds which undergo an ex-
tensive enterohepatic circulation. This synthetic polystyrene
resin containing fixed sulfhydryl groups, when added to food
in a concentration of 1%, doubled the rate of excretion of
methylmercury from mice and lowered blood and tissue levels
compared with untreated controls. In man mercury levels in
blood were reduced and the fecal excretion of methylmercury
enhanced (Bakir et al., 1973).
3.3.3 Dialysis
Extracorporeal hemodialysis will achieve far higher clearance
rates of toxic metals not irreversibly bound to tissues than
can be attained by forced diuresis. Metals known to be dialyzable
are arsenic, copper, iron, lead, lithium, magnesium, mercury,
potassium, sodium, strontium and zinc. The procedure is indicated
where a potentally fatal dose has been absorbed, where the
clinical condition is deteriorating in spite of adequate
treatment by other means, and where a complication exists, such
as aspiration pneumonia or renal insufficiency.
In several acute inorganic mercury poisoning, dialysis may be
performed after giving dimercaprol to remove the mercury
complex.
Cysteine, a sulfhydryl complexing agent, is capable of
reversing the protein binding of methylmercury in the blood, and
its infusion, together with hemodialysis, is effective in
animals and may be effective in man in increasing the rate of
elimination of methylmercury (Kostyniak et al., 1975).
3.3.4 Exchange transfusion
Where there are no facilities for hemodialysis or where the
255
-------
toxic material is poorly dialyzable, exchange transfusion may
be life saving in severe poisoning by agents which remain in the
blood in appreciable concentrations.
3.4 Inactivation of the absorbed poison
There are a limited number of therapeutic agents which may be
administered to counteract the effects of an absorbed toxic
metal. Such a specific antidote may act in different ways. It
may combine with the toxic agent to form a less toxic or a non-
toxic compound, which may be excreted more effectively in the
urine; it may compete with the toxic agent and displace it
from its receptor site; or it may displace the poison into
a tissue where it cannot exert its toxic effects.
The intravenous administration of calcium gluconate will
displace lead from its site of action and will temporarily
relieve the intense pain of lead colic. An infusion of potassium
has been shown to correct the potassium displacing capacity of
absorbed soluble barium salts (Berning, 1975).
Certain antidotes have been designed specifically to compete
for heavy metals with ligands essential for normal physiological
function. These heavy metal antagonists, or chelating agents,
form a stable complex with the metal in the form of a hetero-
cyclic ring (Levine, 1975). The antagonists and the chelates
produced are not themselves without toxic effects and they
should not therefore be administered therapeutically in
those mild cases of poisoning where removal from further
exposure is sufficient to promote recovery.
3.4.1 Dimercaprol (BAL)
2,3 dimercaptopropanol (dimercaprol, British Anti-Lewisite,
BAL) was synthesized as a specific antagonist to the vesicant
arsenical war gases. It is a dithiol compound which success-
fully competes with protein sulfhydryl groups for arsenic
compounds and for other heavy metals, by forming a stable chelate
with them. The other metals for which dimercaprol has been
shown to be effective are mercury in inorganic form, antimony,
256
-------
bismuth, cadmium, chromium, cobalt, gold, nickel. It is also
used as an adjunct to edatate in the treatment of lead
poisoning. However, to be effective it has to be given
as early as possible in the course of poisoning. The half-
life is short, metabolism and excretion being complete within
four hours.
Dimercaprol is given by deep intramuscular injection as a
5 % solution in arachis oil (BP or a 10 % solution with
benzyl benzoate in vegetable oil, USP). It is usually given
in a dose of 3 mg per kg, 4 hourly, for the first two days,
6 hourly on the third day and thereafter once or twice daily
for up to 7 days. In a severe acute case of poisoning 4-5 mg/kg
should be given, 4 hourly for the first 24 hours, but no single
dose should exceed 300 mg. There are a number of minor varia-
tions to this schedule. The drug should be administered at a
lower dosage to patients with impaired renal function. It is
wise to render the urine alkaline during therapy, as the
dimercaprol-metal complex may dissociate in an acidic medium.
Dimercaprol has unpleasant and sometimes alarming side effects
when given in full dosage. One of the most consistent is a
rise in blood pressure accompanied by tachycardia. Other
untoward effects are nausea and vomiting, headache, burning
sensations in the mouth and throat, with paresthesiae of the
hands, a feeling of constriction or of pain in the chest,
lachrymation, salivation, rhinorrhea, sweating and abdominal
pain accompanied by a feeling of anxiety.
Dimercaprol has been shown in the experimental animal to
inhibit the blood pressure lowering effect of intravenously
injected cadmium (Dalhamn and Friberg, 1954) and also of
vanadium and of cobalt (Dalhamn et al., 1953; Dalhamn, 1953).
While the urinary excretion of cadmium is enhanced by
dimercaprol, Tepperman (1947) showed this to be accompanied
by a large increase in cadmium concentration in the kidney
and Oilman et al. (1946) observed severe renal damage as a
sequel. While there may be an indication for giving dimercaprol
in severe acute cadmium poisoning provided careful watch is
257
-------
kept on renal function, dimercaprol should never be given in
chronic poisoning, where the cadmium concentration of the
kidney is already high.
The administration of dimercaprol in mercury poisoning effects
a redistribution of the body burden of the metal without in-
creasing its excretion. In poisoning with inorganic mercuric
salts dimercaprol decreases the renal concentration and thus
protects the kidney. However, with mercuric salts, phenyl
mercury compounds and also with methylmercury compounds,
dimercaprol accelerates the uptake of mercury from blood
into the tissues and in particular its uptake into the brain
(Berlin and Rylander, 1964; Berlin and Lewander, 1965; Berlin
et al., 1965). Dimercaprol is, therefore, contraindicated in
the treatment of poisoning with both aryl- and alkylmercury
compounds.
Dimercaprol enhances the toxicity of selenium and tellurium,
producing kidney damage, and is contraindicated in poisoning
by these metals (Amdur, 1958; Cewenka and Cooper, 1961).
3.4.2 Calcium disodium edatate (Calcium EDTA, Calcium versenate)
Ethylene diamine tetraacetic acid and related compounds are
able to chelate many divalent and trivalent metals in vitro.
Infusion of the sodium salt will chelate calcium from the
body and may result in hypocalcemic tetany. However, the
calcium disodium salt, calcium disodium edatate (Calcium EDTA),
is a valuable therapeutic agent for it will bind lead with the
displacement of calcium from the chelate. However, it has
dangerous toxic effects in the chelation of metals other than
lead. Calcium EDTA is poorly absorbed from the gastrointestinal
tract and so has to be given by intravenous infusion. It is
distributed mainly in the extracellular fluid and is excreted
rapidly by glomerular filtration, about 50 % appearing in the
urine within one hour. Castellino and Aloj (1965) showed in
rats a biphasic pattern of lead excretion, weakly bound
extracellular lead being excreted rapidly, with lead complexed
within cells being removed very slowly, probably as a direct
result of an increased concentration gradient.
258
-------
The toxic effects of calcium EDTA make it necessary to
monitor its administration with care. The most important
of these effects is on the kidney. Friberg (1956) showed
that cadmium excretion could be increased considerably by
giving Ca-EDTA to rabbits previously dosed with cadmium.
However, severe degenerative and necrotic changes were
seen in the tubular epithelium of rabbits with prolonged
exposure to the metal. Reversible hydropic degeneration of
the proximal tubules was produced in rats by Foreman et al.
(1956) and proximal tubular damage has also been observed
in man. It is not known whether this is due to a toxic
effect of the chelating agent itself, to the chelated metal
or to breakdown of the chelate in its transfer through the
tubule, or even to the binding of a trace metal in an essential
enzyme system. Untoward side effects in the treatment of lead
poisoning include a febrile reaction with headache, myalgia,
nausea and vomiting. Lachrymation, nasal congestion, muco-
cutaneous lesions, glucosuria, hypotension and EGG abnormalities
have also been reported. Prolonged courses of calcium EDTA give
rise to trace metal depletion, the most marked being the
excretion of zinc.
Many of the side effects of calcium EDTA have been ascribed
to excessive chelation following the administration of too
high a dose over a short period of time. The drug is now
usually administered in adults by the infusion of 1.0 g in
250 to 500 ml 5 % dextrose over a period of 1 to 2 hours. Two
such infusions may be given daily for 3 to 5 days, but a
daily dose of 50 mg per kg body weight should not be exceeded.
Not more than 75 mg/kg should be given in 6 to 12 hours to a
child. After an interval of a few days, during which time
redistribution of lead from the intracellular to the extra-
cellular compartments occurs, the course may be repeated.
When the blood lead level is high, and especially where there
is encephalopathy, Chisolm (1973) recommends combined BAL and
calcium EDTA therapy. Renal function should be adequate before
commencing treatment. While a patient is on calcium EDTA
treatment, urinalysis should be performed and blood urea,
259
-------
electrolytes, calcium, phosphorus and alkaline phosphatase
should be monitored regularly.
The local application of calcium EDTA in an ointment is of
value in the treatment of chronic ulceration.
3.4.3 Penicillamine
3, f^-dimethylcysteine or penicillamine, is derived from the
hydrolytic degradation of penicillin. The D-isomer is an
effective chelating agent for copper, iron, lead, mercury
and zinc and it will increase the excretion of these metals
in the urine. It is probably the chelating agent of choice
in the less severe forms of lead poisoning. N-acetyl-dl-peni-
cillamine may be even more effective than d-penicillamine in
protecting against the effects of inorganic mercury (Aposhian
and Aposhian, 1959), but there have been few valid comparisons.
Penicillamine is well absorbed from the gastrointestinal tract
and may thus be given by mouth. It is stable once absorbed and
is excreted rapidly in the urine.
The most important side effects of penicillamine are acute
sensitivity reactions manifested by fever, skin rashes, blood
dyscrasias and occasionally renal tubular damage, proteinuria
and the nephrotic syndrome. There is cross sensitivity with
penicillin, so that persons who are allergic to penicillin
should not be given penicillamine. PeniciLlamine antagonizes
pyridoxine, although clinical effects following treatment
with d-penicillamine are unusual.
Penicillamine is given orally, before meals, in four divided
doses at a rate of 0.5 to 1.5 g daily, although up to 5 g
daily has been given without ill effects. The urine should
be tested for protein and a full blood count including a
platelet count should be performed at weekly intervals, but
where treatment continues for longer than 2 to 3 months, as
in the management of Wilson's Disease, these tests may then
be performed monthly.
260
-------
3.4.4 Desferrioxamine mesylate
Desferrioxamine is a chelating agent with a remarkable affinity
for iron but not for other metals. It is poorly absorbed from
the gastrointestinal tract but following injection it complexes
with iron and is rapidly excreted in the urine, increasing the
excretion of iron. The iron in transferrin, the cytochromes
and in hemoglobin is not accessible to desferrioxamine. Given
by mouth it chelates iron remaining in the lumen of the gut
and renders this unabsorbable and thus non-toxic. Its prin-
cipal use is in the treatment of acute iron poisoning in
children. It is also of value in increasing the excretion of
iron in diseases associated with excessive iron storage, and
in transfusion hemosiderosis.
The rapid intravenous administration of desferrioxamine may
lead to an allergic or even anaphylactic reaction with
hypotension and skin rashes. The drug should be used cautiously
in patients with impaired renal function.
Desferrioxamine is given by slow intravenous infusion at a
rate of not more than 15 mg per kg per hour to a maximum of
80 mg per kg per 24 hours. Desferrioxamine may also be given
intramuscularly at a rate of 2 g for an adult, 1 g for a
child, every 12 hours. Gastric lavage may be performed with
5 g desferrioxamine dissolved in 1 liter of water, and 5-10 g
desferrioxamine in 50 to 100 ml water can be left in the
stomach to chelate unabsorbed iron from the gastroinestinal
tract.
3.4.5 Diethylenetriaminopentacetic acid (DTPA)
DTPA is a chelating agent with properties similar to those
of Ca EDTA. It is the most effective agent known for the
chelation of plutonium and its subsequent excretion in the
urine, which may be increased by a factor of one hundred
(Bair and Thompson, 1974). It may be given by slow intra-
venous infusion in a dose of 1 g on alternate days three
times a week for three weeks.
261
-------
3.4.6 Diethyldithiocarbamate (Dithiocarb)
Dithiocarb is a chelating agent which has been found to be
of value in the treatment of acute nickel poisoning (Sunderman,
1971). Dithiocarb can be given orally in moderately severe
poisoning, initially at a rate of 50 mg per kg in divided
doses. In severe poisoning it has been given by Sunderman
parenterally in an initial dosage of 25 mg per kg.
3.5 Modification of response
Under this heading are included examples of therapeutic
measures in metal poisoning which are directed towards a
modification of the tissue response to the poison, or to an
alteration in the biochemical or metabolic state of the
subject.
3.5.1 Modification of tissue response
Chronic beryllium disease is characterized by an inflammatory
reaction which is granulamatous in nature and which appears
to result from a hypersensitive response in certain individuals.
The inflammatory process may be arrested, although not
reversed, by adequate corticosteroid therapy. This treatment
has resulted in a change in the clinical course of the
disease with a reduction in symptoms and a favorable change
in prognosis.
3.5.2 Modification of biochemical status
Chronic manganese poisoning has pathological, biochemical
and clinical features which closely resemble those of Parkinson's
disease, which once established, becomes a permanent disabling
occupational disorder. The observations of Cotzias et al.
(1971) on neurochemical similarities between the two conditions
led to the administration of L-dopa in manganese-induced
parkinsonism. Modification of the parkinsonian state in
chronic manganese poisoning has been obtained with L-dopa
(Mena et al., 1970). In the majority of patients so treated,
hypokinesia and rigidity were markedly reduced and postural
reflexes improved with a restitution of balance. There was
no evidence of an increased body burden of manganese in the
parkinsonian ex-miners treated in this way. Beneficial
effects can be considered solely as a form of replacement
therapy.
262
-------
REFERENCES
Alexander, C.S. (1972). Amer. J. Med. 5_3_, 395-417.
Amdur, M.L. (1958). AMA Arch. Ind. Health 17, 665.
Aposhian, H.V. and Aposhian, M.M. (1959). J. Pharmacol. Exp.
Therap. 126, 131-135.
Bair, W.J. and Thompson, R.C. (1974). Science 183, 715-722.
Bakir, F., Damluji, S.F., Amin-Zaki, L., Murtadha, M.,
Khalid, A., Al-Rawi, N.Y., Tukriti, S., Dhahir, H.I., Clarkson,
T.W., Smith, J.C. and Doherty, R.A. (1973). Science 181,
230-241.
Beattie, A.D., Moore, M.R. and Goldberg, A. (1972). Lancet
2, 12-15.
Beritic, T. (1971). Arch. Environ. Health 23, 289-291.
Beton, D.C., Andrews, G.S., Davies, H.S., Howells, L. and
Smith, G.F. (1966). Brit. J. Ind. Med. 23, 292.
Berlin, M. and Rylander, R. (1964). J. Pharmacol. Exp. Therap.
146, 236-240.
Berlin, M. and Lewander, T. (1965). Acta Pharmacol. Toxicol
_2_2, 1-7.
Berlin, M., Jerksell, L.G. and Nordberg, G. (1965). Acta Pharmacol,
Toxicol. 23, 312-320.
Berning, J. (1975) . Lancet 1, 110 (letter) .
Castellino, H. and Aloj, S. (1965). Brit. J. Ind. Med. 22,
172-180.
Cavanagh, J.B., Fuller, N.H., Johnson, H.R.M. and Rudge, P.
(1974). Quart. J. Med. 4_3, 170, 293-319.
Cerwenka, E.A. and Cooper, W.C. (1961). Arch. Environ. Health
3, 189-200.
Chisolm, J.J. (1973). New Engl. J. Med. 289, 1016-1018.
Clarkson, T.W., Small, H. and Norseth, T. (1973). Arch. Environ.
Health 26, 173-176.
Cotzias, C.G., Papavasiliou, P.S., Ginos, J., Steck, A. and
Diiby, S. (1971). Ann. Rev. Med. 22, 305-326.
Dalhamn, T. (1953). Acta Pharmacol. Toxicol. 9^, 259.
Dalhamn, T. and Friberg, L. (1954). Acta Pharmacol. Toxicol.
1£, 199-203.
Dalhamn, T., Forssman, S. and Sjoberg, S.G. (1953). Acta Pharmacol.
Toxicol. !3, 11.
263
-------
Foreman, H., Finnegan, C. and Lushbaugh, C.C. (1956).
J. Amer. Med. Ass. 160, 1042-1046.
Friberg, L. (1956). AMA Arch. Ind. Health 13_, 18-23.
Friberg, L., Piscator, M., Nordberg, G. and Kjellstrom, T.
(1974). "Cadmium in the Environment." CRC Press, Cleveland.
Oilman, A.F., Philips, S., Allen, P. and Koelle, E. (1946).
J. Pharmacol. 87, 85.
Goyer, R.A. and Rhyne, B.C. (1973). Int. Rev. Exp. Pathol.
12_, 1-77.
Holt, L.E. and Holz, P.H. (1963). In: "Activated Charcoal as
an Antidote for Poisons." pp 1-3. National Clearinghouse
Poison Control Center, USA.
Hunter, D. (1975). "The Diseases of Occupations". (5th edition)
English Universities Press, London.
Jenkins, G.C., Ind, J.E., Kazantzis, G. and Owen, R. (1965).
Brit. Med. J. 2, 78-80.
Kazantzis, G. (1971). Int. J. Environ. Studies .1, 301-306.
Kazantzis, G. (1975). 18th Int. Congr. Occup. Health, Abstracts,
p. 285.
Kostyniak, P.J., Clarkson, T.W., Cestero, R.V., Freeman, R.B.
and Abbasi, A.H. (1975). J. Pharmacol. Exp. Therap. 192, 260-269,
Landrigan, P.J., Whitworth, R.H., Baloh, R.W., Staehling, N.W.,
Barthel, W.F. and Rosenblum, B.F. (1975). Lancet 1, 708-712.
Levine, W.G. (1975). In: "The Pharmacological Basis of
Therapeutics." (L.S. Goodman and A. Gilman, eds) pp 912-923.
(5th edition) Macmillan, New York.
Matthew, H. and Lawson, A.A.H. (1975). "Treatment of Common
Acute Poisonings." (3rd edition) Churchill Livingstone,
Edinburgh.
Mena, I., Court, J., Fuenzalida, S., Papavasiliou, P.S. and
Cotzias, G.C. (1970). N. Engl. J. Med. 282,, 3-8.
Nordberg, G.F. (1976). "Effects and Dose-Response Relationships
of Toxic Metals." Elsevier, Amsterdam.
Stoeckle, J.D., Hardy, H.C. and Weber, A.L. (1969). Amer. J.
Med. j46_, 545-561.
Sunderman, F.W. (1971). Ann. Clin. Res. _3, 182-185.
Teng, C.T. and Brennan, J.C. (1959). Radiology 73, 354-361.
Tepperman, H.M. (1947). J. Pharmacol. 89, 343.
264
-------
MUTAGENIC AND CARCINOGENIC EFFECTS OF METALS
George Kazantzis and Lorna Lilly
1. Mutagenic effects
1.1 Introduction
Genetic damage can be caused by naturally occurring or industrially
produced chemicals, including metallic compounds, in the human
environment. This may have two important consequences. The
first of these is that the damage done to the genes of the
exposed population in the present generation may be manifested
as an increase in abnormal individuals in later generations. The
second important consequence of genetic damage is indicated
by the growing body of evidence which correlates the ability
of an agent to cause genetic damage with its carcinogenicity.
We do not wish to defend the hypothesis that only genetic
damage can result in cancer, but rather to point out the
astonishingly high correlation between agents which are
able to induce mutation and those which are able to induce cancer.
This is reflected in the current attempts to use systems
involving genetic damage for short-term carcinogenicity testing
(see e.g. WHO, 1974; Ames et al., 1975; Bridges, 1976).
Genetic damage may be defined as damage to deoxyribonucleic
acid (DNA). It is usual to classify such damage under three
main headings:
1) Changes at the level of the nucleotide bases in the DNA
("point mutations")
2) Gross chromosomal aberrations
3) Changes in chromosome number
1.2 Types of genetic damage
1.2.1 Point mutations
Mutations may be thought of as a change in the nucleotide bases
of the gene such that replicated copies of that gene give rise
265
-------
to abnormal daughter cells, which if they are gametes, will
pass the abnormality on to the next generation. Such a
change would occur spontaneously only with a very low
frequency. However, in the presence of a mutagen the induced
mutation rate may be many times higher than the spontaneous
rate.
Mutations which involve base substitution such as one purine
for another purine, or a pyrimidine for another pyrimidine
in the DMA are known as transitions, while mutations which
involve the substitution of a purine for a pyrimidine or a
pyrimidine for a purine are known as transversions . Those
mutations which involve the deletion or addition of bases
'called frameshift mutations.
The ability to induce point mutations may be most conveniently
and extensively tested on microorganisms. However, higher
animals (including mammals) may be used. Although some
mutagens may be species specific (for example those that
require the presence of a particular activating enzyme)
others are more generally mutagenic. Since the ultimate
target of a mutagen is DNA, the demonstration of positive
results in any one organism should alert us to the possi-
bility that it may be mutagenic in man.
The "point" mutations which are of most concern to man are
the recessives which may remain hidden for many generations
after their induction, until a chance marriage between two
carriers (heterozygotes) produces an offspring carrying a
double dose of the defective gene (homozygote) who will then
be in some way abnormal. Typical of the many human conditions
produced by recessive genes are the numerous inborn errors
of metabolism such as phenylketonuria, sex-linked recessive
conditions such as hemophilia, and cystic fibrosis which is
probably the commonest of all inherited diseases among
Caucasians, occurring with a frequency of 1 in every 2,000
births .
266
-------
1.2.2 Gross chromosomal aberrations
Chromosomal aberrations involve gross damage to the chromosomes
such as to be readily visible under the light microscope. Such
damage will include chromosome or chromatid breaks resulting
in the loss of fragments of chromosome containing various
amounts of genetic information. More striking abnormalities
may be induced by the joining up of broken ends of two or more
chromosomes or chromatids to form a single configuration
(e.g. exchanges, translocations, rings etc). Such abnormalities
may be unstable and later form anaphase bridges which
interfere with the mechanical separation of the daughter nuclei
and which may cause cell death. Acentric fragments which are
unable to attach to the mitotic or meiotic spindle will be
lost from the daughter nuclei and may also cause cell death.
Many early miscarriages in humans may be due to such unstable
aberrations. In animal experiments, inherited lethality
occurring in the first generation of offspring (dominant
lethality) is usually due to chromosomal aberrations. However,
chromosomal aberrations need not be lethal, and can be passed
on to later generations. Such aberrations may be genetically
unbalanced. For example the translocation form of Down's syndrome
(mongolism) results in 1/3 of live births being affected in
each generation.
Since chromosomal aberrations are only scorable during cell
division, the tissues used for this purpose must either be
dividing in vivo (e.g. bone marrow, or the root tips of
plants) or must be capable of being cultured in vitro
(e.g. human lymphocytes).
1.2.3 Changes in chromosome number
Some environmental agents may cause an increase or decrease in
the number of chromosomes present in a cell (aneuploidy). Such
changes may be due to the chromosomes failing to separate
regularly at cell division (non disjunction). The faulty
distribution of chromosomes may be caused by various events
such as an abnormality in the mitotic spindle, along which
chromosomes travel to the daughter cells, or by a deletion of
the centromeres by which the chromosomes are attached to the
spindle.
267
-------
Agents which destroy the spindle, resulting in a failure to
distribute chromosomes normally to daughter cells, are
said to cause "C-mitosis" (or "C-meiosis") and the con-
sequences of their action can be observed visually under the
microscope. Changes in chromosome number can give rise to a
wide range of inherited abnormality, including for example the
trisomy form of Down's syndrome (mongolism) and the syndromes
of sex chromosome anomaly.
In summary, the three kinds of damage described above could
lead to one of literally hundreds of inherited clinical dis-
orders. It should be emphasized that no single test will
detect all kinds of genetic damage, that not all mutagens are
able to cause all types of damage, and that different organisms
may respond differently to any particular mutagen.
Although most of the papers published on the genetic effects of
environmental agents concern the types of damage described above,
the effects on other genetic processes such as mitotic and
meiotic recombination, gene conversion etc, are occasionally
described. Lack of space will not permit us to discuss these
or other interesting subjects such as the mutations found
in some organisms, which confer resistance to metals
(Tuovinen and Kelly, 1974), the role of metal ions on the
modification of the mutagenicity of non-metallic mutagens
(Shamberger et al., 1973), or the role of metals in the pro-
duction of mutagenic mycotoxins in stored foodstuffs (Gupta and
Venkitasubramanian, 1975).
1.3 Principal metals showing mutagenic effects
A number of metals have been shown to induce mutagenic
effects in biological material ranging from plant cells and
viruses to animal and human cells including those exposed in
vivo. As a result, the possible role of metals as a genetic
hazard to human health is being increasingly recognized.
With most metals the capacity to cause genetic damage to human
germ cells cannot yet be evaluated owing to the inadequate
nature of the data available, but there seems no reason to
268
-------
suppose that these cells are necessarily less sensitive
than other tissues of the body, or that they differ markedly
in this respect from the germ cells of other mammals. Some of
the more important data on the genetic effects of metals are
summarized below.
1.3.1 Mercury (alkyl, aryl and inorganic compounds)
Ceresan-M (containing ethyl (n-phenyl-p-toluene-sulphonanilide)
mercury) was believed to have been responsible for two earlier
outbreaks of mercury poisoning in Iraq in 1956 and 1960
(Mathews and Al-doori, 1976). These authors tested Ceresan-M
for mutagenicity. They fed 30 and 40 mg of the fungicide per
100 mg of food to the larvae of the fruit fly Drosophila
melanogaster. This resulted in a significant increase in
the frequency of sex-linked lethal mutations.
Other fungicides such as Granosan (containing ethylmercuric
chloride) and Agrimix M (containing ethylmercuric chloride and
phenylmercuric-dinaphthyl methanedisulfonate) are known to
affect the mitotic spindle, causing "C-mitosis" (Fishbein,
1974) .
Ramel (1972a) in his extensive review on the genetic effects
of mercury compounds drew attention to their interference with
the mitotic spindle leading to an abnormal distribution of
chromosomes and resulting in aneuploidy. In experiments
upon Allium cepa Ramel compared the "C-mitosis" inducing
ability of methyl- and phenylmercury hydroxide and other
compounds with the similar action of colchicine. From this
and the works of other authors which he reviewed, Ramel
established two very important points. The first of these is
that organic mercurials are even stronger spindle inhibitors
than colchicine, some compounds such as butylmercury bromide
and methylmercury hydroxide being effective at a concentration
as low as 0.1 x 10 M and 2.5 x 10 M respectively, compared
to the 200 x 10 M threshold activity of colchicine. (The
effects of methylmercury on human cells are of the same order
of magnitude as upon Allium cepa.) The second important
point made by Ramel is that whereas low concentrations of
269
-------
colchicine tend to produce a complete blockage of the
spindle fiber mechanisms, the mercurials produce a gradual
transition from normal spindle to complete C-mitosis,
passing through a state where spindle damage is slight,
leading to the defective distribution of single chromosomes.
Both of these observations are of the greatest practical
importance, the former because low levels of environmental
mercury contamination occur widely, and the Latter because the
loss and gain of single human chromosomes are the basic cause
of human abnormalities such as Turner's, Kleinfelter's and
Down's syndromes. Defects caused by a complete C-mitosis, which
concerns the whole chromosome set, are likely to be lethal
at an early stage in development and therefore less important.
Ramel has carried out experiments the results of which support
the hypothesis that organic mercury acts upon the spindle fibers
by interacting with sulfhydryl groups of the protein units
involved in spindle fiber formation.
The effects of mercury are not limited to mitosis. Ramel mentions
his own unpublished experiments which show that methylmercury
can induce "C-meiosis" in the reproductive tissues of flower
buds. He also reports experiments with Drosophila melanogaster
in which he showed that the effects of methylmercury on the
non disjunction of chromosomes can be detected as genetic ab-
normalities in the offspring of these fruit flies. Ramel also
obtained chromosome breakage with both phenyl- and methyl-
mercury in the root tips of Allium cepa.
From his study of the effects of methylmercury on the repair
of radiation-induced chromosome breaks, Ramel assumes that the
chromosomal breakage by methylmercury is due to its action
directly on the chromosomes rather than on enzyme systems of
chromosome repair and DNA synthesis. Umeda et al. (1969) have
reported an increase of abnormal mitotic cells in treatment with
inorganic, phenyl- and alkylmercuric compounds on HeLa cells.
They found "severance of the chromosomes to form rod shaped
masses", polynuclear cells and an increase of the mitotic index.
270
-------
Kato (1976) demonstrated that methylmercuric chloride caused
an increase in chromosomal aberrations in human lymphocytes
after in vitro and in vivo exposure.
The effects of methylmercury on human chromosomes have been
studied by Skerfving et al. (1974). They examined lymphocytes
from the cultured blood of 23 subjects exposed to methylmercury
as a result of a diet of contaminated fish. These were compared
with the lymphocytes of 16 unexposed controls. There were
statistically significant relationships between frequencies of
cells with chromatid type aberrations, "unstable" chromosome
type aberrations, and aneuploidy, and the blood cell mercury
levels in the exposed people. No correlations were found between
stable type chromosome aberrations and blood mercury levels.
1.3.2 Lead
Organic lead compounds are more effective spindle inhibitors than
inorganic forms. In experiments with Allium cepa root tips Ramel
(1972b) demonstrated that the active concentration of organic
compounds of lead, mercury, and tin able to cause spindle
damage was very low, i.e. 10 or 10 M, while with inorganic
metals the doses required are 100-1000 times greater. One of
the interesting possibilities which arises from Ramel's work
is that the effects of different metals may be additive and
therefore the total metal load in the environment may be important.
Although the ability of lead to cause spindle damage (and therefore
euploidy) seems well established, there is some conflicting
evidence as to whether it also causes chromosome aberrations.
Levan (1945) showed that triethyl lead at a concentration of
10 M produced chromosome bridges and fragments in Allium
cepa but that trimethyl lead did not produce a corresponding
effect. No evidence of an increased yield of aberrations was
found by Bauchinger et al. (1972) in a chromosomal analysis
of the lymphocytes of 29 policemen with increased blood levels
of lead. Bauchinger and Schmid (1972) failed to find an increase
in aberrations in cell cultures of the Chinese hamster after
271
-------
treatment with various concentrations of lead acetate.
Schmid et al. (1972) examined 32 workers in the lead manu-
facturing industry, three with acute lead intoxication, and
found no evidence of increased aberration yield compared with
a group of twenty persons from a normal population. They also
found no evidence of increased aberrations in human lymphocytes
treated with lead acetate in vitro.
However, an increase in chromosomal aberrations in occupationally
exposed people has been reported by Schwanitz et al. (1970)
and Forni and Secchi (1972). The latter authors found a signi-
ficant increase in aberrations in 15 workers with preclinical
intoxication and in 37 workers with clinical poisoning, but not
in workers with a history of lead poisoning who had been
unexposed during the previous 18 months. Increased yields
in aberrations have also been found in human lymphocytes
treated in vitro with lead acetate (Schwanitz et al., 1970;
Beek and Obe, 1974; Obe and Beek et al., 1975). Deknudt et
al. (1973) classified the workers which they observed into
three groups, (1) people exposed to high levels of zinc and
low levels of lead and cadmium, (2) people exposed to high
levels of all three metals, (3) people exposed to lead and
cadmium in the absence of zinc. They concluded that exposure to
cadmium and zinc did not appear to increase the number of
cells with severe chromosome abnormalities and that lead was
responsible for the chromosome and chromatid aberrations seen.
Hickey et al. (1974) have applied regression techniques to
data collected from 38 metropolitan regions which showed that
the concentration of atmospheric lead was a significant predictor
of infant mortality rate and congenital Fialformations.
The mutagenic effect of lead on algae (Platymonas subcordiformis)
was investigated by Hessler (1975), who Eound no increased
incidence of mutation in experimental conditions where nitroso-
guanidine (an inducer of point mutations) and U.V. radiation
induced a high proportion of mutants.
272
-------
1.3.3 Cadmium
There is some evidence that cadmium is able to induce genetic
damage. Oehlkers (1952) cites the unpublished experiments of
Glass in which roots of Vivia faba were exposed to cadmium
nitrate. The optimum concentration at which cadmium nitrate
induces chromosomal breakage is about M/10,000 which makes
it about 200 times more active than the chromosome breaker
urethane.
RShr and Bauchinger (1976) treated cell cultures from Chinese
hamsters with various concentrations and various lengths of
-5 -4
exposure to cadmium sulfate. Treatments with 10 to 10
mol/1 for 3 hours caused "C-mitosis" and structural chromosome
aberrations mainly of the chromatid type. After 16 hours of
-4 -8
treatment with 10 to 10 mol/1, marked chromosome stickiness
and picnosis appeared at the higher concentration.
Cadmium is known to kill spermatozoa and cause testicular
interstitial cell tumors in mammals. Gilliavod and Leonard
(1975) investigated the genetic effects of cadmium in mice.
A single intraperitoneal injection of 1.75 mg/kg cadmium
chloride did not increase the dominant lethals in male mice
during the first three weeks after treatment and failed to
induce translocations in the F, male offspring. The treated
males were killed three months later and their testes examined
for translocation configurations. None were found.
Chromatid breaks and exchanges and acentric fragments have
been found in the peripheral lymphocytes in 24 workers in a
zinc melting plant who had increased blood levels of lead
and cadmium (Bauchinger et al., 1976). Because of their
earlier negative findings on lead the authors attribute
these aberrations to either cadmium or the combined effects
of lead and cadmium.
Shiraishi (1975) examined the chromosomes of 12 Japanese
patients with Itai-Itai disease. Eight of these patients had a
high number of chromatid aberrations and the other four had
273
-------
a significant number of stable chromosome type aberrations.
The frequency of aneuploidy in the cells of all 12 was higher
than in cells from controls. Shiraishi also cites earlier
work in which she had shown similar damage in cultured human
lymphocytes treated in vitro.
Bui et al. (1975) in a chromosomal aberration analysis of five
cadmium exposed Swedish workers, four Japanese Itai-Itai patients
and seven Swedish and Japanese controls found no evidence
that cadmium induces chromosome damage in vivo in man. The
Itai-Itai patients as well as the Japanese control subjects
demonstrate a significantly higher frequency of chromosomally
abnormal cells than the cadmium-exposed Swedish workers and the
controls.
The conflict between the findings of Shiraishi and those of
Bui outlined above makes the interpretation of the effects of
cadmium difficult. However, the facts that Glass established
that cadmium was capable of causing chromosome breakage in Vicia,
and that both Shiraishi and Rohr and Bauchinger obtained
effects with mammalian cells treated in vitro (human and
hamster respectively), do suggest that cadmium is capable of
producing chromosome aberrations under some conditions.
Certainly, in the absence of further evidence, the possibility
that it may also do so in vivo in man should not be overlooked.
Cadmium chloride injected in vivo has been shown to cause changes
in chromosome numbers in metaphase II oocytes of mice. These
changes were of such types as would have caused monosomy,
trisomy or triploidy in subsequent progeny (Shimada et al.,
1975, 1976). Similar results were obtained with hamster oocytes
(Watanabe et al., 1976).
1.3.4 Chromium
Venitt and Levy (1974) have demonstrated that simple hexavalent
chromium salts such as Na~CaO., K~CaO. and CaCrO. are able
to induce point mutations in "spot tests" dn which the chemical to
be tested is dropped onto plates prespread with the bacterium
274
-------
E. coli . Three strains were used, two of which lacked certain
DNA repair enzymes. The positive results from these experiments
were confirmed with others in which bacteria in suspension
were treated with Na2Cr02. Soluble salts of other metals such
as tungsten and molybdenum were also tested as was a soluble
trivalent chromium compound Cr_SO .K2SO . • 2H20 . These were
found not to be able to induce mutation. Similar tests with
salts of other metals such as zinc, cadmium and mercury were
also negative. The absence of either the exrA ("error prone-
repair") or the uvrA ("excision repair") pathway did not
modify the mutagenic response to chromate. Venitt et al.
conclude that chromates fall into the transition-causing
class of mutagens. It is highly likely that they specifically
attack guanine-cytosine base pairs in the DNA, causing
guanine-cytosine -> adenine-thymine transitions in the sub-
sequent round of DNA replication.
1.3.5 Selenium
Five selenium compounds were tested for their effects on
human lymphocyte chromosomes by Nakamuro et al. (1976).
Chromosomal aberrations, mainly of the chromatid type,
were found to increase in frequency with increasing dose.
Most were gaps, but there was a substantial increase in
breaks and exchanges as well. The order of efficiency of the
compounds used was H_SeO,, Na2SeO.,, Se02/ H-SeO,, Na2Se04.
Tests with strains of B_._ subtilis having or lacking re-
combination capability were carried out. These experiments
suggested that damage to DNA was produced by selenites but
not by selenates. Experiments to evaluate the effects of
selenium components on the loss of transforming activity
of B. subtilis DNA showed that treatments with H0SeO, and
- - 2 o
^ yield a 30% loss of transforming activity compared to
untreated DNA, but no significant effects were obtained using
or Na
1.3.6 Manganese
Manganese is reported to be able to induce "petite" mutations in
resting and growing cells of yeast (Prazmo et al. , 1975). These
275
-------
mutations involve changes in mitochondria! DNA rather than
in nuclear DNA. Putrament et al. (1975) demonstrated the
induction of antibiotic resistant mutants in yeast mitochondrial
DNA by manganese. From these results the authors suggested that
manganese acts as an error producing factor on replicating
mitochondrial DNA. The authors think the most likely hypothesis is
that the cation acts directly with mitochondrial DNA polymerase.
Dube and Loeb (1975) have shown that manganese acts as a
mutagenic agent during in vitro DNA synthesis and Hall and
Lehman (1968) have shown that manganese reduces the fidelity
of DNA synthesis in vitro.
In 1965 Orgel and Orgel showed that manganese was mutagenic
in bacteriophage T4. The manganous ion induces mutations
that are reversible by 5 Bu or 2 amino-purine and which are
therefore presumably of the transition type.
1.3.7 Zinc
Kroeger (1964) reported an effect of zinc chloride on the
puffing of the larval chromosomes of Chironomus thummi. These
effects are similar to those produced by the mutagen urethane
and by ecdysone. Carpenter and Ray (1969) found no significant
effect of zinc chloride on the production of dominant lethals
or sex linked recessive mutation in Drosophilia. However, there
was an effect with radioactive ZnCl-.
Herich (1969) showed that zinc sulfate inhibited the passage
of cells from interphase to prophase and caused a disturbance
in the differentiation of the chromosome matrix and precocious
dispiralization of the chromosomes. The effect of zinc ions
on mitosis was also studied by Barlow (1966), using cells
of wing anlagen of Ephestia kuhniella z. in vitro. The
numbers of nuclei in prophase rose steeply to a peak after
three hours of treatment in 0.02M ZnCl-. Treatments of more
than two hours significantly raised the number of metaphases.
Barlow concluded that zinc blocks metaphase by disrupting the
mitotic spindle.
276
-------
Organic zinc compounds may be used as fungicides. Pilinskaya
(1970) reported that chromosomal aberrations were present in
the blood of 9 industrial workers who had worked in a ziram
plant for 3-5 years. An increased incidence of both chromosome
and chromatid type aberrations was found in their lymphocytes.
Blood from unexposed persons (Pilinskaya, 1971) treated in
vitro with ziram also contained chromosomal aberrations.
In a survey of patients treated with a closely related
organic compound disulfiram (Antabuse) containing no zinc,
and with ethyl instead of methyl groups, Lilly (1975) was
unable to find a significant increase in aberrations. It is
possible that the difference between the action of ziram
and disulfiram may be due to the presence and absence of
zinc molecule respectively.
1.3.8 Nickel
Cultured rat embryo muscle cells exposed to nickel showed a
marked decrease in mitotic index and in induced abnormal mitotic
figures such as multipolar spindles, C-metaphase-like shapes,
disoriented bipolar spindles, lagging chromosomes and unequal
chromosome divisions (Swierenga and Basrur, 1968). Histo-
chemical staining for protein-bound sulfhydryl groups showed that
nickel could be causing mitotic abnormalities by interfering
in the sulfhydryl groups involved in the spindle mechanism.
Levan (1945) pointed out unusual results with roots of Allium
cepa exposed to nickel nitrate. The chromosomes behave as in
"C-mitosis", but nevertheless the spindle remains clearly
visible.
Nickel chloride, which does not give tumors in animals, also
gave negative results in two mutagenicity tests with Salmonella
typhimurium in host-mediated assay (Buselmaier et al., 1972).
Corbett et al. (1970) showed that nickel sulfate at 300/Ug/ml
is toxic but not mutagenic to intracellular bacteriophage T4.
277
-------
1.3.9 Arsenic
Arsenical interference with normal DNA repair processes in
E. coli has been reported by Rossman et al. (1975).
Petres et al. (1970) examined the lymphocytic chromosomes
of 13 people who had been exposed to arsenic and also the
chromosomes of unexposed controls. They found a large
increase in acentric fragments, chromatid bridges and
dicentric chromosomes in the exposed group, and they point
out that this is in agreement with their earlier in vitro
work. Petres and Hagedorn (1977) found a similar increase
in patients treated with arsenicals.
Elevated sister chromatid exchange rates were found by Burgdorf
et al. (1977) in the lymphocytes of six individuals who had
taken arsenic and later developed stigmata of arsenic use,
including multiple cutaneous malignancies. Sister chromatid
exchanges are a type of aberration which can only be observed
after differential staining of the chromatids. Their biological
importance is not fully understood, but they may be induced
by many chromosome-breaking substances.
Oppenheim and Fishbein (1965) were able to induce chromosome
aberrations in cultured human lymphocytes by treatment with
potassium arsenite.
1.3.10 Other metals
Levan (1945) has shown that "colchicine mitosis" (C-mitosis)
is caused by large numbers of inorganic metals salts such as
those of lithium, beryllium, chromium, iron, cobalt,
nickel, copper, arsenic, rubidium, palladium, cadmium, barium,
gold, mercury, thallium, lead. With many o~ these metals the
threshold of "C-mitosis" was considerably Lower than that for
colchicine. For example the threshold for Lead nitrate and
copper nitrate was as low as 0.000005-0.00005 M. There was
a tendency for the threshold to fall with Increasing
molecular weight. Levan also showed that pronounced chromosome
stickiness was caused by titanium, lithium, beryllium, aluminum,
iron, cobalt, nickel, bismuth, molybdenum, and thallium.
278
-------
Radiation can cause genetic damage, so radioactive isotopes of
metals such as uranium can be mutagenic in this way. However,
it has been pointed out (United Nations, 1972) that high
concentrations of uranium isotopes may present a toxicological
rather than a radiological hazard. We do not know of any
reports of genetic experiments specifically designed to
distinguish between the radiation and chemical effects of
uranium. However, Malenkho and Bakun (1973) have reported
aneuploidy in human lymphocytes and mouse muscle cells after
treatments with up to 0.5/ug/ml and 0.3,ug/ml respectively
of uranium nitrate. Chromosome stickiness has also been
reported by Levan.
Cobalt is known to cause "petite" (mitochondrial) mutations in
yeast (Prazmo et al., 1975).
Oehlkers (1952) referred to his earlier work on meiotic
chromosomes and to the work of Deufel (1951) with mitotic
chromosomes of Vicia faba both of which had shown that
aluminum chloride was a potent chromosome breaker.
Many chemicals used for anti-tumor therapy are able to
induce mutation or chromosome breaks. One such agent is the
platinum(II) coordination complex cis-PtC!0 (NH.,) 0 which
causes DNA cross links. Monti-Bragadin et al. (1975) showed
that PtCl2(NHO2 causes mutation in Salmonella typhimurium.
The strain of Salmonella used is specifically reverted by base-
pair substitution mutagens.
Sub-toxic doses of antimony sodium tartrate and ammonium tellurate
have been reported to induce chromosome aberrations in vitro
in cultured human lymphocytes (Paton and Allison, 1972).
2. Carcinogenic effects
2.1 Definition, mechanism of action
A carcinogen is a substance exposure to which results in
the eventual appearance of a cancer which would not have
279
-------
occurred otherwise. A chemical carcinogen may initiate
malignant change by direct action or by a variety of indirect
mechanisms. The only common mechanisms of action known to
exist between widely varying carcinogenic chemicals relate
to their electrophilic reaction, which makes them capable of
reacting with various nucleophilic centers in the cell such
as nucleic acids and proteins, to their ability to be
converted by metabolic processes to electrophilic reagents or
to their properties as free radicals. The mechanism of action
of the carcinogenic metals is far from clear at present. They
appear to be direct acting or ultimate carcinogens, although
their oxidation state may influence their carcinogenic potential
(Mancuso, 1975) and salts of the same oxidation state may have
different carcinogenic effects (Sunderman, 1973) .
2.2 Recognition and evaluation of carcinogenic activity
2.2.1 Clinical observation
That some metals may act as carcinogens was first realized
from clinical observation of industrial workers. The carcinogenic
potential of nickel was recognized some 30 years after the
opening of the Mond Nickel Refinery in South Wales, when 10
cases of nasal cancer, a rare tumor in the general population,
were reported. Observations by astute clinicians of an associa-
tion between cancer and an environmental, or behavioral factor
have on several occasions been the starting point of detailed
epidemiological enquiry. Such observations are likely to
be as valuable in the future as they have been in the past.
2.2.2 Epidemiological enquiry
The carcinogenic potential of chemicals can only be investigated
in humans by means of epidemiological study. Such study is
capable of identifying high risk groups together with their
antecedent host and environmental factors. However, inter-
actions between environment and host factors, and even between
environmental exposures, are complex and require sophisticated
methods for their investigation (Fraumeni, 1975). With regard
to the metals, exposure in relation to occupation is likely to
280
-------
have been greater than general environmental exposure and the
study of occupational groups is thus of particular value.
However, exposure patterns in industry up to the present time
have only been available retrospectively, rendering them but
approximations. Furthermore, the exposure patterns involving the
suspected carcinogenic metals are difficult to characterize
and to record. Epidemiological. study can tell us that a number
of industrial processes involving exposure to certain metals
carry an increased risk of cancer, but it may only be possible to
define this risk in terms of an occupation, an industrial plant
or a process (Norseth, 1976). Apart from the problems involved
in complex exposures over long periods of time, further diffi-
culties in evaluating the carcinogenic effect of one specific
metal are caused by host and behavioral variables which may
increase or decrease the carcinogenic response.
Smoking histories have been deficient in past epidemiological
studies, and the carcinogenic effect of tobacco smoke could
not be assessed. Dietary pattern, exposure to other chemicals
or drugs, hormonal and immune responses may also provide
important modifications. The multifactoral nature of the
risks, then, mean that sophisticated methods are required for
the evaluation of the carcinogenic potential of an individual
metal. At the present time, the only metals which have been
shown convincingly by epidemiological study to cause human
cancer are arsenic, chromium and nickel, while beryllium and
cadmium are highly suspected.
2.2.3 Bioassay
Many metals have been tested for carcinogenic activity in
experimental animals but the animal models used have often
borne no relation to the human exposure situation in terms of
dose or route of administration. Since human exposure occurs
only by inhalation or ingestion it is difficult to interpret the
significance of cancer induced in the rat by a large dose of
an insoluble compound of the metal given by intramuscular or
subcutaneous injection. For example, cadmium sulfide suspensions
will produce sarcoma in a high proportion of rats injected in
281
-------
this way (Kazantzis, 1966) but there is no evidence at present
that the inhalation of insoluble cadmium sulfide dust can give
rise to cancer in man. Nevertheless, in our present state of
knowledge, it is probably wise to work on the assumption that
certain individuals would be at risk oi: developing cancer if
exposed to a chemical carcinogen definitely active in one or
more animal models. Basic biochemical systems in man and animals
are sufficiently similar to make it likely that man would not
react differently from animal, models except as regards
quantitative aspects and tissues affected. Conversely most
chemical agents shown by epidemiological study to give rise
to cancer in humans have also demonstrated carcinogenic
activity in one or more animal models (Weisburger and Rail,
1972). At an international symposium at the Karolinska
Institute in 1977 (Task Group on Air Pollution and Cancer, 1978)
on risk assessment methodology and epidemiological evidence of
cancer, it was concluded that if a substance has been shown to
be carcinogenic in an adequate animal test system, it should
normally be dealt with as if it is also carcinogenic in man,
unless adequate epidemiologocal evidence exists to the contrary.
In borderline cases, the report concludes, error should be on the
side of safety.
To give reliable information, test systems using animal models
have to be very carefully designed and controlled. Various
factors have to be taken into account Ln the selection of
animal species, strains, sex and age. Because of the possibility
of interactions and of biologic effects the purity of the metal
compound to be tested must be assured. Careful thought should
be given to the correct use of control animals, the route of
administration and the evaluation of results, taking into
account the experimental design. The correct use of animal
models is further considered by Weisburger (1976).
2.2.4 In vitro test systems
Much of the information we have at present on the carcinogenic
effects of metals has been obtained by bioassay using animal
models. However, the technique is always time-consuming and
282
-------
expensive and it may even be misleading, especially when
applied to the detection of weak carcinogens. The need to
obtain a rapid, if only provisional, answer to the
question as to whether a chemical in current or
proposed use has a carcinogenic potential has led to intense
activity in an attempt to devise an appropriate indicator system
which in vitro will mimic molecular receptors (Stoltz et al.,
1975). Such tests, briefly indicated below and reviewed by
Weisburger (1976) , should at present be regarded as not more than
screening procedures.
2.2.4.1 Mutagenicity tests
It has been found empirically that many chemical carcinogens
in their reactive form are also powerful mutagens in microbial
and mammalian test systems. However, certain carcinogens
have not been shown to exert mutagenic activity and some
powerful mutagens do not appear to be carcinogenic. The
congruence between mutagenic and carcinogenic effects
became evident with the development of techniques for converting
a precarcinogen into the ultimate carcinogen, and a premutagen
to the ultimate mutagen (Ames, 1971; Ames et al., 1973). The
present state of knowledge on the value of mutagenicity
tests in evaluating the carcinogenic effect of chemicals can
be summed up as follows (see WHO, 1974; Lancet, 1977).
While the relationship between carcinogenesis and mutagenesis
requires further investigation, the association between
mutagenicity and carcinogenicity of many compounds is
sufficiently great to justify the use of mutagenicity tests
in prescreening for possible carcinogenesis. The results of
mutagen testing, to be evaluated, will require correlating
with figures for the incidence of cancer in populations
known to be exposed to particular hazards.
2.2.4.2 Cell culture transformation
Cell cultures can be transformed in vitro by the addition
of a carcinogen and these will give rise to abnormal cell
lines when reimplanted into their host (Williams et al.,
283
-------
1973). This technique may not allow for adequate biochemical
activation, but a further development, in >rfhich the carcinogen-
is administered to a pregnant animal, and tissue cultures
prepared from the fetus, will do so (DiPaolo and Nelson, 1973).
2,2.4.3 DNA damage and repair
There are differences in the type of DNA damage which can
be induced by carcinogenic and non-carcinogenic chemicals.
Study of these breaks and their repair processes may be of
value in the identification of carcinogenic activity (Stich arid
San, 1973).
2.2.4.4 Serum protein indicators
Carcinogens which act primarily on the liver have been shown
to induce serum a-feto protein in the rat long before the
appearance of the tumor (Kroes et al., 1973) . This observa-
tion is being developed as a screening test for the carcinogenic
potential of a limited group of chemicals.
2.3 Principal metals showing carcinogenic effects
Only three metals have been shown in epidemiological studies
to exert an unequivocal carcinogenic effect in humans. These
metals are nickel, chromium and arsenic. However, adequate
epidemiological studies and in particular prospective studies,
have not been performed and human data are at present greatly
lacking. There is some evidence, at present inconclusive, that
cadmium may exert a carcinogenic effect and possibly also beryllium.
Most of the available epidemiological data have been obtained
retrospectively on occupationally exposed groups. Of the
three metals known to have exerted a carcinogenic effect in
humans, only nickel and chromium have been confirmed as
carcinogens in experimental animals, there being no conclusive
evidence in the case of arsenic. Cadmium and beryllium, which
are suspect in humans, have been shown to have an unequivocal
carcinogenic effect in animals, although tae exposure situations
have not been comparable. In addition to the above, iron,
cobalt, zinc, titanium and lead have produced tumors in
experimental animals but the doses used and modes of
284
-------
administration have been entirely foreign to any conceivable
human exposure situation. These ten metals will be considered
in greater detail below.
2.3.1 Nickel
2.3.1.1 Epidemiological observations
A highly significant excess of both lung cancer and cancer
of the nasal sinuses has been found in nickel refinery workers
in different countries (Mastromatteo, 1967; Doll et al.,
1970; Virtue, 1972; Pedersen et al., 1973; Saknyn and
Shabynina, 1973; Sunderman and Mastromatteo, 1975). In the
Mond Nickel Works in South Wales, where exposure to nickel
carbonyl gas had occurred in the past, Doll et al. (1970)
estimated the excess mortality from lung cancer to be ten times
that expected for men first employed before 1914 and about
five times that expected for those first employed between
1915 and 1924. For nasal cancer, those first employed between
1910 and 1914 experienced an almost 900-fold increase, while
those first employed between 1920 and 1924 had a mortality
almost 100 times the national average. Recent observations
by Doll (1975) showed that the risk to the refinery workers
continued at least until 1929. While all the men studied were
manual workers in the refinery, only some of these were
employed on the specific process. Others were general laborers
who changed their job frequently. It was thus not possible to
collect detailed information on exposure.
The theory that nickel carbonyl had acted as a specific
carcinogen had to be abandoned when a similar hazard was found
in nickel refinery workers in Ontario, where only the electrolytic
process had been used (Mastromatteo, 1967). Sunderman (1973)
collected data on 327 cases of lung cancer and 115 cases of
cancer of the nasal sinuses in nickel refinery workers with a
variety of exposure to nickel in Norway, France, Germany,
USA and Japan as well as in Wales and Canada.
285
-------
Apart from nickel refinery workers, isolated cases • " 1nnq
and nasal cancer have been reported in workers in n :- i1
plating, nickel grinding and nickel puLishing opcrn' : • •-
but the significance of these cases cannot be a;:'v~? *.j -...
no definitive epidemiological study has yet beet.> pertormed,
However, from the extent of exposure in the old i-, i'c!:sl Defining
processes and frorr the fall in cance* mortality \'-,&^ thl.i was
reduced, it seems unlikely that the rruch lowt r ~:t. ./c"• ; of exposure
encountered in nickel plating would constitute a rjsk.
The exact identity of the carcinogenic agent or agents in
nickel refineries is open to speculation. Sunderman and
Mastromatteo (1975) consider that fresh nickel dnsbs from sintering
and roasting processes are especially carcinogen-_c, but also
speculate on respirable particles of tretallic nickel
subsulfide and nickel oxide as significant carcinogens. The
available data, which are inadequate, suggest thai- the cancer
risk is greater in heavy cigarette smokers with an inter-
action between two carcinogens as in the case of asbestos and
smoking.
The latent period from first exposure to nickel to diagnosis of
cancer is long, averaging about 25 years with a range of some
10 to 40 years. In those cases for which information is
available, epithelial tumors predominated but anaplastic
and pleomorphic tumors have also been observed (Sunderman, 1973).
2.3.1.2 Animal models
The experimental models which have been used to study nickel
carcinogenesis are reviewed by Sunderman (1973). Carcinogenesis
has been adequately demonstrated in several animal species
following inhalational and parenteral administration of nickel
but not following oral or cutaneous exposure. Nickel sulfide,
Ni_.S~, is the most potent metallic compound yet tested with
regard to carcinogenesis, subcutaneous or intramuscular
injection, having given rise to fibrosarcoma or rhabdomyosarcoma
(Heath and Daniel, 1964; Sunderman, 1976). (The high yield
of tumors and the ease with which they can be produced and
286
-------
propagated makes this method a conveniently simple one for
experimental tumor induction.) Metallic nickel powder in
suspension and in the form of pellets or sponge has produced
similar tumors following implantation. Lung cancer has been
produced in guinea pigs following the inhalation of nickel
dust (Hueper, 1958) and in rats following the inhalation
of nickel carbonyl (Sunderman and Donnelly, 1965).
Carcinoma of the kidney and cholangiosarcoma of the liver
in addition to sarcoma have been observed following multiple
intravenous injections of nickel carbonyl (Lau et al., 1972).
While comparison of the relative potency of the various nickel
compounds tested is difficult because of variation in
experimental design from study to study, potency appears to
be inversely related to solubility of the nickel compound in
aqueous media, the highly soluble salts like nickel chloride
having no apparent carcinogenic activity. Various possible
mechanisms of nickel carcinogenesis have been proposed. Beach
and Sunderman (1970) have observed an inhibition of RNA synthesis
in hepatocytes, believed to be an inhibitory effect of nickel
on RNA polymerase activity, thus altering the expression of
genetic information.
2.3.2 Chromium
2.3.2.1 Epidemiological observations
A lung cancer hazard in workers producing chromates from the
raw material, chromite or chrome iron ore was first noted
in Germany in the 1930's. The observation was confirmed
in the USA by Machle and Gregorius (1948) and in Britain by
Bidstrup and Case (1956). Similar observations have been made
in the 1970's in the Soviet Union, Japan and Norway (Pokrovskaya
and Shabynina, 1973; Ohsaki et al., 1974; Langard and Norseth,
1975). In the American series the observed mortality for
lung cancer was about 30 times the number expected from the
national mortality data of that time, but in the British series
the increased risk was computed at three times that expected.
287
-------
In the industrial pioc-SL Liie : aw <_; •.. . _..: • c<3, riirliud.
;aixed with calcium and J<:"!JUP! or sxr ., ~. i... • bc-iiate and
roasted in a rotor; 1'a± n •<„••' i.r --i .-- " •• -.' K-.phc.ri~ 'r!~v.
JLUJOl^': I oxidv IS C..1U" ;t .. • ,-7-.-. • '>' 'I'um "J; !.•" .'•"" te
cind the dichromate pioduced by ureaz;i .g \.,~.is with dilute
sulfuric acid. As the lung cancer hazard has only been
demonstrated convincingly in workers producing chromates and
not in workers exposed to dichromate or chromic acid, it
seems reasonable to suppose that the carcinogenic agent is
present in the primary ore or in one of the furnace products
either as an intermediate or as a component of the residue
(Kazantzis, 1972). However, Pokrovskaya (1973) reported an
increased incidence of lung cancer in chrome alloy production
and Mancuso (1975) considers all forms of chromium to be
carcinogenic. There are no studies available today which
unequivocally show that chromium compounds other than
hexavalent ones are carcinogenic (Langard and Norseth,
1977) .
The time interval from the first exposure to chromium
compounds to the findings of lung cancer has varied from
10 to 47 years, with a mean latent interval from 10.6 to
21 years in different studies. About 80% of the lung cancers
examined were squamous cell carcinomas, but undifferentiated
cancers and adenocarcinomas have also been seen.
2.3.2.2 Animal models
Animal models have not contributed so far to the precise
identification of the carcinogen in the chromate production
industry although they confirm the carcinogenic potential of the
metal. A small number of tumors have been observed in the rat
following the intramuscular or intrapleural administration of
chromite ore roast (Hueper, 1958), and more tumors followed
the intramuscular implantation of calcium chromate (Hueper and
Payne, 1959). A small number of rats developed lung cancer
(6 squamous, 2 adenocarcinoma in 100 rats) following
bronchial implantation of calcium chromate in the form of a
pellet (Laskin, 1968). Inhalation exposure to calcium chromate
288
-------
dust produced bronchial adenocarcinoma in mice in a single
experiment (Nettesheim et al., 1971). Levy and Venitt (1975)
produced squamous cell carcinoma of the rat lung by implanting
medium solubility hexavalent chromium salts into the bronchial
tree. In their animals without lung tumors, a significant
increase in squamous metaplasia was seen in those rats exposed
to the pure hexavalent compounds but not in those exposed
to others. These authors also tested a number of hexavalent and
trivalent chromium compounds for mutagenic activity. Again
they observed significant levels of induced mutation with
hexavalent chromium compounds, but none with soluble trivalent
compounds.
2.3.3 Arsenic
2.3.3.1 Epidemiological observations
The evidence for the carcinogenicity of arsenic is based
on observations made in occupational groups and also in
general population groups exposed to arsenic following
ingestion in food, water or medicaments or following inhalation
in arsenic polluted atmospheres.
A number of different occupational groups with exposure to
arsenic have shown an excess cancer risk. An increase in
observed lung cancer deaths over those expected was found
in workers engaged in the production of sodium arsenate
(Hill and Faning, 1948) , in refinery and smelter workers
(Lee and Fraumeni, 1969; Milham and Strong, 1974; Tokudome and
Kuratsune, 1976; Pinto, 1977), and in miners (Osburn, 1969).
Arsenic is a component of ores containing copper, lead, zinc,
silver and gold, nickel or cobalt and is also associated with
natural radioactive sources as in the Erzgebirge. Exposure
patterns are therefore complex and often involve other suspected
carcinogens. Although exposure may occur to a variety of
arsenic compounds, refinery workers and others are pre-
dominantly exposed to airborne arsenic trioxide. German and
French vineyard workers who used a pesticide spray containing
arsenic developed chromic arsenic poisoning over a 12- to 14-
289
-------
year exposure period. Roth (1958) investigated 27 deaths
and found 12 cases of lung cancer, one of which was bilateral,
among these workers. These deaths occurred 8 to 14 years
after the pesticides had been banned.
Epidemiological observations have been made in retrospective
and in correlation studies, and the relative risk of developing
respiratory cancer has been computed at 3.3, based on 147 observed
and 44 expected deaths (IARC, 1973).
Skin cancer following chronic arsenic poisoning and following
medication with arsenical preparations, in particular with
Fowler's solution, has been consistently reported (Neubauer,
1947) . The characteristic pathogenesis consists of. keratotic
lesions persisting for many years, distributed particularly
on palms and soles, and eventually progressing to epithelioma,
which may be multifocal in origin. Chronic lesions resembling
eczema or psoriasis which progress to a low grade basal cell
carcinoma may also occur (Bowen's disease). Bowen's disease
may be associated with an increased incidence of other malignant
lesions, but this has not been confirmed. There have been a
number of case reports of multicentric arsenic-induced epithelioma
associated with lung cancer, sometimes bilateral in origin.
Cancers of the lymphatic and hemopoietic systems, and hemangio-
sarcoma of the liver, the latter associated with hepatic cirrhosis
and splenomegaly (Banti's disease), hcive also been reported
to occur more frequently following arsenic exposure, both with
and without epithelioma (Lander et al., 1975; Popper and
Thomas, 1975). However, the excess risk has not yet been
adequately confirmed epidemiclogically.
The daily intake of arsenic in food and water is of the order
of 100 to 800 /ug, with an average value of 190,ug/day (Schroeder
and Balassa, 1966j Somers and Smith, 1971). It is usually present
in its pentavalent form, although traces of trivalent arsenic
derived from pesticides may also be found (Grasso and O'Hare,
1976). Contamination of food and drinking water from industrial
sources and subsequent observations of chronic arsenic
290
-------
poisoning have been widely reported (see Tsuchiya et al., 1977).
In one such episode in Taiwan (Yeh and How, 1963; Tseng, 1977)
various clinical manifestations, including hyperkeratotic
skin lesions were seen in the exposed population. Arsenic
levels in drinking water were of the order of 0.8 to 2.5 mg/1.
The rate of skin cancer was very high, 10,6/1000 population,
and evidence of a dose-response relationship was found.
Arsenic is one of the large number of carcinogens identified in
tobacco smoke, but its significance remains to be determined.
Raised levels of arsenic have been detected in the atmosphere
and in soil in the vicinity of non-ferrous smelters and
refineries, and raised levels of urinary arsenic in the
neighborhood population (Milham and Strong, 1974). In a
large scale epidemiological study in the USA average mortality
rates from lung cancer were significantly increased in
countries with copper, lead or zinc smelting or refining
industries compared with countries where other non-ferrous ores
were processed (Blot and Fraumeni, 1975) . After taking con-
founding variables into account, the authors concluded that the
excess mortality could well have been caused by industrial
emissions containing inorganic arsenic.
As with nickel and chromium, the latent interval between first
exposure to arsenic and development of cancer is long. Average
latent intervals of 34, 39 and 41 years were estimated for
heavy, medium and light exposure categories (Lee and Fraumeni,
1969) .
Further references to skin and lung cancer following
arsenic exposure are given by Hernberg (1966) and by Tsuchiya
et al. (1977) .
2.3.3.2 Animal models
Surprisingly, there is little if anv convincina evidence for
the carcinogenic effect of arsenic in experimental animals.
An increased incidence of lymphocytic leukemia and of lymphoma
was observed in female Swiss mice and their young following
291
-------
the subcutaneous injection of sodium arsenate during pregnancy
(Osswald and Goettler, 1971). However, repeated application
of an aqueous solution of sodium arsenate failed to elicit
any skin tumors (Baroni et al., 1963). Many other negative
findings have been reported.
2.3.4 Cadmium
2.3.4.1 Epidemiological observations
In contrast to arsenic, there is little epidemiological
evidence at present to support the contention that cadmium
is a cancer hazard for man, while animal models show cadmium
to have a considerable carcinogenic potential.
Three out of eight deaths in 74 men with at least 10 years
of exposure to cadmium oxide dust were noted by Potts (1965) to be
from cancer of the prostate. Kipling and Waterhouse (1967) surveyed
248 workers exposed to cadmium oxide dust for a minimum
period of one year. They found four cases of prostatic cancer,
where the expected figure was computed at 0.58, from annual
incidence rates supplied by the Birmingham Regional Cancer
Registry. They concluded that the increased incidence of
prostatic cancer was highly significant (probability of
occurrence 0.003) , while the incidence of other forms of
cancer was close to that expected. However, the numbers
involved were not large (they included the 3 cases noted
by Potts) and the authors were unable to infer the existence
of an industrial hazard. Since then, only one study has
supported the above findings, a retrospective cohort mortality
investigation of workers exposed to cadmium fume and cadmium
oxide and sulfide dusts (Lemen et al., 1976) . Atmospheric
samples taken in the premould department showed concentrations
of 74.8 and 90.3,ug/m for cadmium and 0.3 and l.l.ug/m for
arsenic respectively. Samples in the retort department showed
11.05 ,ug/m for cadmium and 1.4 ug/m for arsenic. The study
involved 292 workers with a minimum of 2 years of exposure
during a 30-year period from 1940. The group experienced a
total of 92 known deaths (99.3 expected). A significantly
292
-------
excessive risk was demonstrated for malignant neoplasms,
with 27 observed and 17.6 expected (p<0.05). Twelve of
these deaths were due to respiratory cancer (5.11 expected,
p<0.05) and 4 were due to prostatic cancer (1.15 expected,
N.S.). The greatest risk for respiratory cancer was apparent
30 years after initial employment and for prostatic cancer
20 years after initial exposure. At that time interval, the
increased risk was significant as the expected figure was
only 0.88 (p<0.05). However, in this study there was con-
comitant exposure to arsenic, and smoking histories have not
yet been made available.
2.3.4.2 Animal models
Finely divided cadmium metal suspended in fowl serum injected
into the thigh muscle of the rat gave rise to rhabdomyosarcoma
and fibrosarcoma (Heath and Daniel, 1964) . Similar tumors
followed the intramuscular and subcutaneous injection of
cadmium sulfide and of cadmium oxide suspensions (Kazantzis, 1963;
Kazantzis and Hanbury, 1966). The tumors were metastasizing
and pleomorphic and the tumor yield was high. In the series with
cadmium oxide, the tumors were comparable with the nickel
sulfide-induced tumors described earlier. Subcutaneous and
intramuscular injection of soluble cadmium chloride or sulfate,
in addition to producing sarcomata at the injection site, also
induced interstitial cell tumors of the testis (Haddow et al.,
1964; Roe et al., 1964; Reddy et al., 1973), even after a
single dose (Gunn et al., 1963). The interpretation of the
specific role of cadmium in initiating interstitial cell tumors
of the testis is difficult. The initial effect of cadmium is a
necrotizing one on the interstitial cells, probably acting
through the vascular supply to the testis, and it is possible
that the neoplastic change is secondary to this. Furthermore,
certain strains of laboratory rats have since been shown to
have a high spontaneous incidence of interstitial cell tumors
(Levy et al., 1973).
In an experimental series in which cadmium sulfate was given
by subcutaneous injection or by gastric instillation repeatedly
293
-------
over a prolonged time interval to rats and to mice, in
the latter by gastric instillation only, no morphological change
attributable to cadmium was seen in the prostate gland or
in other parts of the genitourinary system. The dose range
was based on exposure levels calculated to produce low molecular
weight proteinuria in man. The rats were given weekly sub-
cutaneous injections of 0.05, 0.1 or 0.2 mg 3 CdSO. • 8H2°
weekly gastric instillation of 0.2, 0.4 or 0.8 mg/kg for
2 years and the mice were given weekly gastric instillations
up to 4 mg/kg for 18 months, equivalent to 1.75 mg Cd/kg body
weight (Levy et al., 1973; Levy and Clark, 1975; Levy et al.,
1975). This dose range was considered to be too low for
adequate carcinogenic evaluation (IARC, 1973).
2.3.5 Beryllium
2.3.5.1 Epidemiological observations
There is very little epidemiological evidence to support
the contention that exposure to beryllium compounds provides
a cancer risk in man. Mancuso (1970) examined the records
of the Beryllium Case Registry (Hardy et al., 1967) and,
from a total of 800 cases, found no overall evidence of an
increased mortality from lung cancer, compared with a control
cohort. There did appear to be a small increase in lung cancer
mortality in workers with a short exposure and in particular
in those with a previous history of respiratory disease.
Bayliss (1972), who investigated the causes of death among
3900 men who had worked in American beryllium factories,
found no increase in mortality from respiratory cancer.
2.3.5.2 Animal models
The carcinogenic activity of beryllium compounds, in particular
of the oxide, sulfate, phosphate and zinc beryllium silicate,
has been confirmed by a number of investigations in rabbits, rats,
mice and monkeys. Intravenously administered beryllium oxide and
zinc beryllium silicate have given rise to osteogenic sarcoma
in mice and rabbits (Gardner and Heslington, 1946; Barnes et
al., 1950). Inhalational exposure has also been followed by
294
-------
osteosarcoma (Dutra et al., 1951). In other series where
beryllium oxide, sulfate or phosphate was administered by
inhalation to rats and monkeys, lung cancer was produced
(Vorwald and Reeves, 1959; Schepers, 1964; Reeves et al.,
1967). The cell types included adenocarcinoma as well as
squamous cell carcinoma.
Taking the various experimental observations together, a dose-
response relationship for the production of malignant tumors
in animal models becomes apparent, and the quantitative aspects
of this are given by Reeves (1977) .
Beryllium has been shown to block thymidine incorporation into
hepatic DNA (Witschi, 1970) and also certain enzymes needed
for DNA synthesis. It has been suggested that beryllium
binding to nucleoproteins leads to specific inhibition of
events leading to DNA replication (Reeves, 1977) .
2.3.6 Iron
2.3.6.1 Epidemiological observations
Speculation has been put forth on the role of iron oxide
as a possible carcinogen common to hematite, asbestos, nickel
and chrome workers but very little evidence exists in support
of such a hypothesis. Boyd et al. (1970) observed 36 deaths
from lung cancer in hematite miners in Cumberland, England,
compared with 21 expected, based on regional mortality figures,
this being a significant difference. An increased lung cancer
mortality has also been noted in the iron miners of Kiruna,
Sweden (Jorgensen, 1973). However, in both occupational
environments there are also increased concentrations of radon
daughters. Cigarette smoking and diesel fumes have also been
queried as confounding variables.
2.3.6.2 Animal models
Iron dextran and a few other polysaccharide complexes have
induced local sarcomas in mice, rats, hamsters and rabbits
following subcutaneous or intramuscular injection in high
295
-------
dosage (e.g. Richmond, 1959; Haddow and Horning, 1960;
Roe and Carter, 1967). An increased incidence of non-
lymphoreticular neoplasms at sites distant to the injection has
also been observed in mice (Langvad, 1968) but this obser-
vation requires confirmation. The significance of these
observations in relation to iron therapy in man has been
debated. The possible hazard is reviewed by Roe (1967) . While
parenteral iron polysaccharide therapy is common, sarcoma
at the injection site has been rarely leported.
2.3.7 Lead
2.3.7.1 Epidemiological observations
Mortality studies of lead workers have not shown any evidence
of an excess risk for malignant neoplasms (Dingwall-Fordyce and
Lane, 1963; Robinson, 1974) .
2.3.7.2 Animal models
Numerous investigators have confirmed the observation
initially made by Zollinger (1953) on the induction of renal
carcinoma and adenoma in rats given lead phosphate by
subcutaneous injection. Similar results have been obtained
with lead acetate given subcutaneously or by mouth to rats and
also to mice (Van Esch and Kroes, 1969; Coogan, 1973) .
Interstitial cell tumors of the testis have followed prolonged
feeding with lead acetate in rats (Zawirska and Medras, 1968).
The doses used in all these experiments were high and the
significance with regard to human environmental exposure,
while unknown, is unlikely to be great.
2.3.8 Zinc
2.3.8.1 Epidemiological observations
There are no epidemiological observations to suggest that
zinc may present an increased cancer hazard to man.
2.3.8.2 Animal models
The production of teratomas by the injection of zinc chloride
into the testes of chickens was the first experimental
296
-------
observation to be made on the carcinogenic action of metal
compounds (Sunderman, 1976). Since then the observation has
been amply confirmed by several investigators. The subject
is reviewed by Sunderman (1976) who could find no evidence
that zinc compounds are carcinogenic after administration
by any route other than intratesticular injection.
2.3.9 Cobalt
2.3.9.1 Epidemiological observations
There is no evidence to suggest that cobalt may present an
increased cancer hazard to man.
2.3.9.2 Animal models
Metallic cobalt powder, cobalt oxide and cobalt sulfide have
been shown to induce local sarcomas following subcutaneous intra-
muscular or intraosseous injection (Vollman, 1940; Heath, 1956;
Oilman, 1962). Injection into muscle gave rise to rhabdomyosarcoma,
as seen following the administration of nickel and of
cadmium. Heath et al. (1971) were able to produce such neoplasms
with fine particles from prostheses made from a cobalt
chromium alloy.
2.3.10 Titanium
2.3.10.1 Epidemiological observations
There is no evidence to suggest that titanium may present
an increased cancer hazard to man.
2.3.10.2 Animal models
An increased incidence of fibrosarcoma and of lymphosarcoma
has been observed in rats given titanium metal powder in
long-term survival studies (Furst, 1971). Tumors were also
obtained in rats and mice with titanocene injected intra-
muscularly. Fibrosarcoma was induced at the injection site
and hepatomas and malignant lymphomas were also obtained
(Furst and Haro, 1969).
297
-------
REFERENCES
Ames, B.N. (1971). In: "Chemical Mutagens: Principles and
Methods for Their Detection." (A. Hollander, ed) p 267, Vol.
2, New York.
Ames, B.N., Dunston, W.E., Yamasaki, E. and Lee, F.D. (1973).
Proc. Natl. Acad. Sci. 70, 2281.
Ames, B.N., McCann, E., Choi, E. and Yamasaki, E. (1975).
Proc. Natl. Acad. Sci. T2, 5135-5139.
Barlow, C.A. (1966). Nature 209, 1045-1046.
Barnes, J.M., Denz, F.A. and Sissons, H.A. (1950). Brit. J.
Cancer 4_, 212-222.
Baroni, C., van Esch, G.J. and Saffiotti, U. (1963). Arch.
Environ. Health 1_, 668-674.
Bauchinger, M. and Schmid, E. (1972) . Mutat. Res. L4_, 95-
100.
Bauchinger, M., Schmid, E. and Schmid, T. (1972). Mutat.
Res. 16_, 407-412.
Bauchinger, M., Schmid, E., Einbrodt, H.J. and Dresp, J.
(1976). Mutat. Res. _40, 57-62.
Bayliss, D. (1972). In: "NIOSH Criteria for a Recommended
Standard: Occupational Exposure to Beryllium." US Department
of Health, Education, and Welfare, Washington, D.C.
Beach, D.J. and Sunderman, F.W. (1970). Cancer Res. _3Q_,
1645-1650.
Beek, B. and Obe, G. (1974). Experientia 30_, 1006-1007.
Bidstrup, P.L. and Case, R.A.M. (1956). Brit. J. Ind. Med.
_13, 260-264.
Blot, W.J. and Fraumeni, J.F. (1975). Lancet 2, 142-144.
Boyd, J.T., Doll, R., Faulds, J.S. and Leiper, J. (1970).
Brit. J. Ind. Med. 27, 97-105.
Bridges, B. (1976). Nature 261, 195-200.
Bui, T., Lindsten, J., Nordberg, G.F. (1975). Environ. Res.
9, 187-195.
Burgdorf, W., Kurvink, K. and Cervenka, J. (1977) Hum.
Genet. 36, 69-72.
Buselmaier, von W., Rohrborn, G. and Propping, P. (1972).
Biol. Zentralbl. 9_I, 311-325.
Carpenter, J.M. and Ray, J.H. (1969). Amer. Zool. 9_/ H21.
298
-------
Coogan, P.S. (1973). Proc. Inst. Med. Chicago 29, 509.
Corbett, T.H., Heidelberger, C. and Dove, W.F. (1970). Mol.
Pharmacol. 6_, 667-679.
Deknudt, Gh., Leonard, A. and Ivanov. B. (1973). Environ.
Physiol. Biochem. _3 , 132-138.
Deufel, J. (1951). Chromosoma _4, 239-272.
Dingwall-Fordyce, I. and Lane, R.E. (1963). Brit. J. Ind.
Med. _20, 313-315.
DiPaolo, J.A. and Nelson, R.L. (1973). Arch. Pathol. 95,
380.
Doll, R. (1975). Personal communication.
Doll, R., Morgan, L.G. and Speizer, F.E. (1970). Brit. J.
Cancer 24, 623-632.
Dube, D.K. and Loeb, L.A. (1975). Biochem. Biophys. Res.
Commun. (T7, 1041-1046.
Dutra, F.R., Largent, E.J. and Roth, J.L. (1951). AMA Arch.
Pathol. 51, 473-479.
Fishbein, L. (1974). Sci. Total Environ. 2, 341-371.
Forni, A. and Secchi, G.C. (1972). In: "Environmental Health
Aspects of Lead." pp 473-485. Published by the Commission of
European Communities Directorate General for Dissemination
of Knowledge,.Center for Information and Documentation CID,
Luxembourg.
Fraumeni, J.F. (1975). In: "Persons at High Risk of Cancer.
An Approach to Cancer Aetiology and Control." (J.F. Fraumeni,
ed) Academic Press, New York.
Furst, A. (1971) Geol. Soc. Amer. Mem. 123, 109-130.
Furst, A. and Haro, R.T. (1969). Prog. Exp. Tumour Res. 12_,
102-133.
Gardner, L.U. and Heslington, H.F. (1946). Fed. Proc. 5,
221.
Gilliavod, N. and Leonard, A. (1975) . Toxicology _5, 43-47.
Oilman, J.P.W. (1962). Cancer Res. 22, 158-162.
Grasso, P. and O'Hare, C. (1976). In: "Chemical Carcinogens."
(C.E. Searle, ed) p 721. American Chemical Society, Washington
D.C.
Gunn, S.A., Gould, T.C. and Anderson, W.A.D. (1963). J.
Natl. Cancer Inst. 31, 745-759.
299
-------
Gupta, S.K. and Venkitasubramanian, T.A. (1975). Appl.
Microbiol. 29, 834-836.
Haddow, A. and Horning, E.S. (1960). J. Natl. Cancer Inst.
24., 139-147.
Haddow, A., Roe, F.J.C., Dukes, C.E. andMitchley, B.C.V.
(1964). Brit. J. Cancer IQ, 667-673.
Hall, Z.W. and Lehman, I.R. (1968). J. Mol. Biol. 36 , 321-
333.
Hardy, H.C., Rabe, E.W. and Lurch, S. (1967). J. Occup. Med.
9_, 271-276.
Heath, J.C. (1956). Brit. J. Cancer 10, 668-673.
Heath, J.C. and Daniel, M.R. (1964). Brit. J. Cancer 18,
124.
Heath, J.C. and Daniel, M.R. (1964). Brit. J. Cancer 18,
251-264.
Heath, J.C., Freeman, M.A.R. and Swanson, S.A.V. (1971).
Lancet 1, 564-566.
Herich, R. (1969). Nucleus 12, 81-85.
Hernberg, S. (1976). Paper presented to Symposium on Origins
of Human Cancer, Cold Spring Harbour Laboratory, New York.
Hessler, A. (1975). Mutat. Res. 31, 43-47.
Hickey, R.J., Clelland, R.C., Boyce, D.E. and Bowers, E.J.
(1974). Mutat. Res. 26_, 445-446.
Hill, A.B. and Faning, E.L. (1948). Brit. J. Ind. Med. 5, 2-
6.
Hueper, W.C. (1958). AMA Arch. Ind. Health 18, 284-291.
Hueper, W.C. (1958). Arch. Pathol. 65^ 600-607.
Hueper, W.C. and Payne, W.W. (1959). Amer. Ind. Hyg. Ass. J.
20, 274.
IARC (1973). "Evaluation of Carcinogenic Risk of Chemicals
to Man." Vol. 2, pp 1-181. International Agency for Research
on Cancer, WHO, Geneva.
IARC Working Group on the Evaluation of the Carcinogenic
Risk of Chemicals in Man (1973). IARC Monograph, Vol. 2,
Lyon.
Jorgensen, H.S. (1973). Work Environ. Health 10, 107-125.
300
-------
Kato, R. (1976). Jap. J. Hum. Genet. 20_, 256-257.
Kazantzis, G. (1963). Nature 198, 1213.
Kazantzis, G. (1966). Brit. J. Cancer 20, 190.
Kazantzis, G. (1972). Ann. Occup. Hyg. 15, 25-29.
Kazantzis, G. and Hanbury, W.J. (1966). Brit. J. Cancer 20,
190-199.
Kipling, M.D. and Waterhouse, J.A.H. (1967) . Lancet ^, 730-
731.
Kroeger, H. (1964). Chromosoma 15, 36-70.
Kroes, R., Williams, G.M. and Weisburger, J.H. (1973).
Cancer Res. 33, 613.
Lancet (1977). 2, 340-341.
Lander, J.J., Stanley, R.J., Sumner, H.W., Boswell, D.C. and
Aach, R.D. (1968). Gastroenterology 6JB, 1582-1586.
Langvad, E. (1968). Brit. J. Cancer 3^» 415-423.
Langard, S. and Norseth, T. (1975). Brit. J. Ind. Med. 32,
62-65.
Langard, S. and Norseth, T. (1977) . In: "Toxicology of
Metals." Volume II, pp 164-187. Environmental Health Research
Series, 600/1-77-022, Environmental Protection Agency,
Research Triangle Park.
Laskin, S. (1968). American Industrial Health Association,
St. Louis. Quoted by Roe, F.J.C. and Carter, R.L. (1969).
Brit. J. Cancer 23, 172.
Lau, T.C., Hackett, R.L. and Sunderman, F.W. (1972). Cancer
Res. 32., 2253-2258.
Lee, A.M. and Fraumeni, J.F. (1969). J. Natl. Cancer Inst.
421, 1045-1052.
Lemen, R.A., Lee, J.S., Wagoner, J.K. and Blejer, H.P.
(1976). Ann. N.Y. Acad. Sci. 271, 273-279.
Levan, A. (1945). Nature 156, 751-752.
Levy, L.S. and Clark, J. (1975). Ann. Occup. Hyg. 17, 205-
211.
Levy, L.S. and Venitt, S. (1975). British Association for
Cancer Research, 16th Annual General Meeting and Symposium
"Industrial Carcinogenesis." Birmingham, England.
301
-------
Levy, L.S., Roe, F.J.C., Malcolm, D., Kazantzis, G., Clark,
J. and Platt, H.S. (1973). Ann. Occup. Hyg. 16_, 111-118.
Levy, L.S., Clark, J. and Roe F.J.C. (1975). Ann. Occup.
Hyg. JL7, 213-230.
Lilly, L.J. (1975). Toxicology 4_, 331-340.
Machle, W. and Gregorius, F. (1948). Publ. Health Rep. Wash.
6J3, 1114-1127.
Malenkho, A.F. and Bakun, N.A. (1973). Dokl. Akad. BSSR
1T_, 375-377.
Mancuso, T.F. (1970) . Environ. Res. _3, 251-275.
Mancuso, T.F. (1975). Paper presented at the International
Conference on Heavy Metals in the Environment, Toronto,
Canada, Oct. 27-31.
Mastromatteo, E. (1967). J. Occup. Med. _9, 127-136.
Mathews, C. and Al-Doori, Z. (1976). Mutat. Res. 40, 31-36.
Milham, S. and Strong, T. (1974). Environ. Res. ^7, 176-182.
Monti-Bragadin, C., Tamaro, M. and Banfi, E. (1975). Chem.
Biol. Interact. 11, 469-472.
Nakamuro, K., Yoshikawa, K., Sayato, Y. , Kurata, H., Tonomura,
M. and Tonomura, A. (1976). Mutat. Res. 4^0_, 177-183.
Nettesheim, P., Hanna, M.G., Dougherty, D.G., Newell, R.F.
and Hellman, A. (1971). J. Natl. Cancer Inst. 47, 1129-1144.
Neubauer, 0. (1947) . Brit. J. Cancer 1, 192-251.
Norseth, T. (1976). Paper presented at the Symposium on the
Origins of Human Cancer, Cold Spring Harbour Laboratory, New
York.
Obe, G., Beek, B. and Dudin, G. (1975). Mutat. Res. 29, 283.
Oehlkers, F. (1952). Symposium on Chromosome Breakage, June
9-11, 1952. Supplement to Heredity 6^
Ohsaki, Y., Abe, S., Hamma, Y., Yozawa, K., Kishi, F. ,
Murao, M., Sato, H., Date, F., Kawanchi, F., Kobayanshi, T.
and Fujita, I. (1974). J. Jap. Soc. Intern. Med. 63, 1198-
1203.
Oppenheim, J.J. and Fishbein, W.N. (1965). Cancer Res. 25,
980-985.
Orgel, A. and Orgel, L.E. (1965). J. Mol. Biol. 14, 453-457.
302
-------
Osburn, H.S. (1969) . South Afr. Med. J. 4_3_, 1307-1312.
Osswald, H. and Goettler, K. (1971). Verb. Dtsch. Ges.
Pathol. j>5, 289-293.
Paton, G.R. and Allison, A.C. (1972). Mutat. Res. 16, 332-
336.
Pedersen, E., Hogelveit, A. and Candersson, A. (1973). Int.
J. Cancer 12, 32-41.
Petres, J. and Hagedorn, M. (1977). Environ. Health Perspect.
JL9_ (In press) .
Petres, J., Schmid, T., Ulrich, K. and Wolf, U. (1970).
Dtsch. Med. Wochenschr. P. 95, 79-80.
Pilinskaya, M.A. (1970). Genetika £, 157-163.
Pilinskaya, M.A. (1971) Genetika 1_, 138-143.
Pinto, S.S. (1977). Environ. Health Perspect. _19_ (In press).
Pokrovskaya, L.V. (1973). Gig. Tr. Prof. Zabol. 10, 23-26.
Pokrovskaya, L.V. and Shabynina, N.K. (1973). Gig. Tr. Prof.
Zabol. _17, 23-26.
Popper, H. and Thomas, L.B. (1975). Ann. N.Y. Acad. Sci.
246, 172-193.
Potts, C.L. (1965). Ann. Occup. Hyg. JS, 55-61.
Prazmo, W., Balbin, E., Barznowski, H., Ejchart, A. and
Putrament, A. (1975). Genet. Res. 26, 21-29.
Putrament, A., Baranowska, H., Ejchart, A. and Prazmo, W.
(1975). Mol. Gen. Genet. 140, 339-347.
Ramel, C. (1972a). "Mercury in the Environment." (L. Friberg
and J. Vostal, eds) pp 169-181. CRC Press, Cleveland.
Ramel, C. (1972b). 2nd meeting EEMS Zinkovy Castle, Czechoslovakia,
p 48 (Abstract).
Reddy, J., Svoboda, D., Azamoff, D. and Dumas, R. (1973). J.
Natl. Cancer Inst. 51, 891-903.
Reeves, A.L. (1977). In: "Toxicology of Metals." Volume II,
pp 85-109. Environmental Health Research Series, 600/1-77-022,
Environmental Protection Agency, Research Triangle Park.
Reeves, A.L.. Deitch, D. and Vorwald, A.J. (1967). Cancer
Res. 21_, 439-445.
Richmond, H.G. (1959). Brit. Med J. 1, 947-949.
303
-------
Robinson, T.R. (1974). J. Occup. Med. 16, 601-605.
Roe, F.J.C. (1967). In: "Practical Carcinogenic Hazards from
Drugs." (R. Truhaut, ed) pp 105-118. Springer-Verlag, Berlin.
Roe, F.J.C. and Carter, R.L. (1967). Int. J. Cancer 2, 370-
380.
Roe, F.J.C., Dukes, C.E., Cameron, K.M., Pugh, R.C.B. and
Mitchley, B.C.V. (1964). Brit. J. Cancer _18_, 674-681.
Rossman, T., Meyn, M.S. and Troll, W. (1975). Mutat. Res.
3CI, 157-162.
Roth, F. (1958). Virchows Arch. (Pathol. Anat.) 331, 119-
137.
Rohr, G. and Bauchinger, M. (1976). Mutat. Res. _40_, 125-130.
Saknyn, A.V. and Shabynina, N.K. (1973). Gig. Tr. Prof.
Zabol. 17, 25-28 (In Russian).
Schepers, G.W.H. (1964). Ind. Med. Surg. _33_, 1-16.
Schmid, E., Bauchinger, M., Pietruck, S. and Hall, G. (1972)
Mutat. Res. 16, 401-406.
Schroeder, H.A. and Balassa, J.J. (1966). J. Chron. Dis. 19,
85-106.
Schwanitz, G., Lehnert, G. and Gebhart, E. (1970). Dtsch
Med. Wochenschr. 9j>, 1636-1641.
Shamberger, R.J., Baughmann, F.F., Kalchert, S.L., Willis,
C.E. and Hoffman, G.C. (1973). Proc. Natl. Acad. Sci.
TQ_, 1461-1463.
Shimada, T,, Watanabe, T. and Endo, A. (1976). Mutat. Res.
•4jO, 389-395.
Shimada, T., Watanabe, T. and Endo, A. (1975). Teratology
12_, 210.
Shiraishi, Y. (1975). Humangenetik 27, 31-44.
Skerfving, S., Hansson, K. and Ryman, N.J. (1974). Environ.
Res. 1, 83-98.
Somers, E. and Smith, D.M. (1971). Food Cosmet. Toxicol. 9_,
185.
Stich, H.F. and San, R.H.C. (1973). Proc. Soc. Exp. Biol.
14_2, 155.
Stoltz, D.R., Poirier, L.A., Irving, C.C., Stich, H.F.,
Weisburger, J.H. and Grice, H.C. (1975). Toxicol. Appl.
Pharmacol. 29, 157.
304
-------
Sunderman, F.W. (1973). Ann. Clin. Lab. Sci. 3, 156-180.
Sunderman, F.W. (1976). Prev. Med. 5, 279-294.
Sunderman, F.W. (1976). In: "Advances in Modern Toxicology."
(R.A. Goyer and M.A. Mehlman, eds) Hemisphere Publishing
Corp., Washington, D.C.
Sunderman, F.W. and Donnelly, A.S. (1965). Amer. J. Pathol.
46_, 1027-1041.
Sunderman, F.W. and Mastromatteo, E. (1975). In: "Nickel."
(F.W. Sunderman, F. Coulston, G.L. Eichorn, J.A. Fellows, E.
Mastromatteo, H.T. Reno and M.H. Samitz, eds) pp 144-188.
National Academy of Sciences, Washington, D.C.
Swierenga, S.H. and Basrur, P.K. (1968). Lab. Invest. 19,
663-674.
Task Group on Air Pollution and Cancer - Risk Assessment
Methodology and Epidemiological Evidence. (1978). Report
from an International Symposium at the Karolinska Institute,
Stockholm, 1977. Environ. Health Perspect. (In press, 1978).
Tokudome, S. and Kuratsune, M. (1976). Int. J. Cancer 17,
310-317.
Tseng, W.P. (1977). Environ. Health Perspect. 19, 109-119.
Tsuchiya, K., Ishinishi, N. and Fowler, B. (1977). In:
"Toxicology of Metals." Volume II, pp 30-70. Environmental
Health Research Series, 600/1-77-022, Environmental Protection
Agency, Research Triangle Park.
Tuovinen, O.H. and Kelly, D.P. (1974) . Arch. Microbiol. 95,
153-164.
Umeda, M., Saito, K., Hirose, K. and Saito, M. (1969). Jap.
J. Exp. Med. 39, 47-59.
United Nations (1972). Ionizing Radiation Levels and Effects.
Vol. 1, Levels. United Nations Scientific Committee Report
on the Effects of Atomic Radiation.
Van Esch, G.J. and Kroes, R. (1969) . Brit. J. Cancer 23,
765-770.
Venitt, S. and Levy, L.S. (1974). Nature 250, 493-495.
Virtue, J.A. (1972). Can. J. Otolaryngol. 1, 37-42.
Vollman, J. (1940). Schweitz A. Allg. Pathol. Bakteriol. _2,
440-443.
Vorwald, A.J. and Reeves, A.L. (1959). Arch. Ind. Health 19,
190-199.
305
-------
Watanabe, T., Shimada, T. and Endo, A. (1976). Teratology
14, 259.
Weisburger, J.H. (1976). In: "Chemical Carcinogens." (C.E.
Searle, ed) pp 1-23. ACS Monograph 173, American Chemical
Society, Washington, D.C.
Weisburger, J.H. and Rail, D.P. (1972). In: "Environment and
Cancer." (R.L. Clark, ed) p 437. Williams andWilkins,
Baltimore.
WHO (1974). Assessment of the carcinogenicity and mutagenicity
of chemicals. WHO Technical Report Series No. 546.
Williams, G.M., Elliot, J.M. and Weisburger, J.H. (1973).
Cancer Res. _33_, 606.
Witschi, H.P. (1970). Biochem. J. 120, 623-634.
Yeh, S. and How, S.W. (1963). Inst. Pathol. Natl. Taiwan
Univ. L4, 25-73.
Zawirska, B. and Medras, K. (1968). Zentralbl. Allg. Pathol.
Ill, 1-12.
Zollinger, H.U. (1953). Virchows Arch. Pathol. Anat. 323,
694-710.
306
-------
TECHNICAL REPORT DATA
J !-'<±tji,, Hi '>'• or '/,'(• /t: i / M "t /I"L v t
REPORT NO
EPA-600/1-78-016
4 TITLE ANDSuBTlTLE
TOXICOLOGY OF METALS
- Volume III
7 AUTHORlS)
Lars Friberg, Chairman
8 PERFORMING ORGANIZA riON REPORT NO
9 PERFORMING ORGANIZATION NAMI AND ADDRESS
Subcommittee on the Toxicology of Metals
Permanent Commission and International Association
of Occupational Health
12 SPONSORING AGENCY NAME AND ADDRESS
Health Effects Research Laboratory
Office of Research and Development
U.S. Enironmental Protection Agency
Research Triangle Park, N.C. 27711
RTP,NC
CV'tMT S ACCESS 1C"* NO
REPORT J A T E
February i
6 PE RFORMING ORGANIZATION CODE
10 PROGRAM ELEMENT NO.
1AA601
11 CONTRACT/GRANT NO
68-02-1287
13 TYPE OF REPORT AND PERIOD COVERED
14 SPONSORING AGENCY CODE
EPA 600/11
15 SUPPLEMENTARY NOTES
Prepared in cooperation with the Swedish Environmental Protection Board, and the
Karolinska Institute
16. ABSTRACT
This report on metal toxicology contains reviews on the general aspects of
metal toxicology. These have been written for inclusion in a Handbook on the
Toxicology of Metals: Environmental and Occupational Aspects which is being
prepared by the Scientific Committee on the Toxicology of Metals of the Permanent
Commission and International Association of Occupational Health.
DESCRIPTOR'
KEY WORDS AND DOCUMENT ANALYSIS
'DENTIFIERS'OPEN ENDED TERMS
metals
toxicology
reviews
DiSTRIBUF J ^ STATEMENT
RELEASE TO PUBLIC
EPA Form 2220-t (3 73)
19 SECURITY CLASS I I h:\ Report I
'0 SECURITY CLASS i rim -tagi'l
UNCLASSIFIED
L COSATI 1 idd/Croup
06 T
PAGES
22 PRICE
307
------- |