Transition-Metal Storage, Transport, and Biomineralization

1
Transition-Metal Storage, Transport,
and Biomineralization
ELIZABETH C. THEIL
Department of Biochemistry
North Carolina State University
KENNETH N. RAYMOND
Department of Chemistry
University of California at Berkeley
I. GENERAL PRINCIPLES
A. Biological Significance of Iron, Zinc, Copper, Molybdenum, Cobalt,
Chromium, Vanadium, and Nickel
Living organisms store and transport transition metals both to provide appropriate
concentrations of them for use in metalloproteins or cofactors and to protect
themselves against the toxic effects of metal excesses; metalloproteins and
metal cofactors are found in plants, animals, and microorganisms. The normal
concentration range for each metal in biological systems is narrow, with both
deficiencies and excesses causing pathological changes. In multicellular organisms,
composed of a variety of specialized cell types, the storage of transition
metals and the synthesis of the transporter molecules are not carried out by all
types of cells, but rather by specific cells that specialize in these tasks. The
form of the metals is always ionic, but the oxidation state can vary, depending
on biological needs. Transition metals for which biological storage and transport
are significant are, in order of decreasing abundance in living organisms: iron,
zinc, copper, molybdenum, cobalt, chromium, vanadium, and nickel. Although
zinc is not strictly a transition metal, it shares many bioinorganic properties with
transition metals and is considered with them in this chapter. Knowledge of iron
storage and transport is more complete than for any other metal in the group.
The transition metals and zinc are among the least abundant metal ions in
the sea water from which contemporary organisms are thought to have evolved
(Table 1.1).1-5 For many of the metals, the concentration in human blood plasma
2 1 I TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
Table 1.1
Concentrations of transition metals
and zinc in sea water and human
plasma.a
Sea water Human plasma
Element (M) X 108 (M) X 108
Fe 0.005-2 2230
Zn 8.0 1720
Cu 1.0 1650
Mo 10.0 1000
Co 0.7 0.0025
Cr 0.4 5.5
V 4.0 17.7
Mn 0.7 10.9
Ni 0.5 4.4
a Data from References 1-5 and 12.
greatly exceeds that in sea water. Such data indicate the importance of mechanisms
for accumulation, storage, and transport of transition metals and zinc in
living organisms.
The metals are generally found either bound directly to proteins or in cofactors
such as porphyrins or cobalamins, or in clusters that are in tum bound by
the protein; the ligands are usually 0, N, S, or C. Proteins with which transition
metals and zinc are most commonly associated catalyze the intramolecular or
intermolecular rearrangement of electrons. Although the redox properties of the
metals are important in many of the reactions, in others the metal appears to
contribute to the structure of the active state, e.g., zinc in the Cu-Zn dismutases
and some of the iron in the photosynthetic reaction center. Sometimes equivalent
reactions are catalyzed by proteins with different metal centers; the metal
binding sites and proteins have evolved separately for each type of metal center.
Iron is the most common transition metal in biology. 6,7 Its use has created
a dependence that has survived the appearance of dioxygen in the atmosphere
ca. 2.5 billion years ago, and the concomitant conversion of ferrous ion to ferric
ion and insoluble rust (Figure 1.1 See color plate section, page C-1.). All plants,
animals, and bacteria use iron, except for a lactobacillus that appears to maintain
high concentrations of manganese instead of iron. The processes and reactions
in which iron participates are crucial to the survival of terrestrial organisms,
and include ribonucleotide reduction (DNA synthesis), energy production
(respiration), energy conversion (photosynthesis), nitrogen reduction, oxygen
transport (respiration, muscle contraction), and oxygenation (e.g., steroid synthesis,
solubilization and detoxification of aromatic compounds). Among the
transition metals used in living organisms, iron is the most abundant in the
environment. Whether this fact alone explains the biological predominance of
iron or whether specific features of iron chemistry contribute is not clear.
I. GENERAL PRINCIPLES
Many of the other transition metals participate in reactions equivalent to
those involving iron, and can sometimes substitute for iron, albeit less effectively,
in natural Fe-proteins. Additional biological reactions are unique to nonferrous
transition metals.
Zinc is relatively abundant in biological materials. 8
,9 The major location of
zinc in the body is metallothionein, which also binds copper, chromium, mercury,
and other metals. Among the other well-characterized zinc proteins are
the Cu-Zn superoxide dismutases (other forms have Fe or Mn), carbonic anhydrase
(an abundant protein in red blood cells responsible for maintaining the pH
of the blood), alcohol dehydrogenase, and a variety of hydrolases involved in
the metabolism of sugars, proteins, and nucleic acids. Zinc is a common element
in nucleic-acid polymerases and transcription factors, where its role is
considered to be structural rather than catalytic. Interestingly, zinc enhances the
stereoselectivity of the polymerization of nucleotides under reaction conditions
designed to simulate the environment for prebiotic reactions. Recently a group
of nucleic-acid binding proteins, with a repeated sequence containing the amino
acids cysteine and histidine, were shown to bind as many as eleven zinc atoms
necessary for protein function (transcribing DNA to RNA). 10 Zinc plays a structural
role, forming the peptide into multiple domains or "zinc fingers" by means
of coordination to cysteine and histidine (Figure 1.2A See color plate section,
page C-l.). A survey of the sequences of many nucleic-acid binding proteins
shows that many of them have the common motif required to form zinc fingers.
Other zinc-finger proteins called steroid receptors bind both steroids such as
progesterone and the progesterone gene DNA (Chapter 8). Much of the zinc in
animals and plants has no known function, but it may be maintaining the structures
of proteins that activate and deactivate genes. 11
Copper and iron proteins participate in many of the same biological reactions:
(1) reversible binding of dioxygen, e.g., hemocyanin (Cu), hemerythrin (Fe),
and hemoglobin (Fe);
(2) activation of dioxygen, e.g., dopamine hydroxylase (Cu) (important in
the synthesis of the hormone epinephrine), tyrosinases (Cu), and catechol
dioxygenases (Fe);
(3) electron transfer, e.g., plastocyanins (Cu), ferredoxins, and c-type cytochromes
(Fe);
(4) dismutation of superoxide by Cu or Fe as the redox-active metal (superoxide
dismutases).
The two metal ions also function in concert in proteins such as cytochrome
oxidase, which catalyzes the transfer of four electrons to dioxygen to form water
during respiration. Whether any types of biological reactions are unique to copper
proteins is not clear. However, use of stored iron is reduced by copper
deficiency, which suggests that iron metabolism may depend on copper proteins,
3
4 1 / TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
such as the serum protein ceruloplasmin, which can function as a ferroxidase,
and the cellular protein ascorbic acid oxidase, which also is a ferrireductase.
Cobalt is found in vitamin B12 , its only apparent biological site. 12 The vitamin
is a cyano complex, but a methyl or methylene group replaces CN in
native enzymes. Vitamin-B 12 deficiency causes the severe disease of perniCious
anemia in humans, which indicates the critical role of cobalt. The most common
type of reaction in which cobalamin enzymes participate results in the reciprocal
exchange of hydrogen atoms if they are on adjacent carbon atoms, yet not with
hydrogen in solvent water:
b c b c
I I I I
a-C-C-d~a-C-C-d
I I I I
x H H X
(An important exception is the ribonucleotide reductase from some bacteria and
lower plants, which converts ribonucleotides to the DNA precursors, deoxyribonucleotides,
a reaction in which a sugar -OH is replaced by -H. Note that
ribonucleotide reductases catalyzing the same reaction in higher organisms and
viruses are proteins with an oxo-bridged dimeric iron center.) The cobalt in
vitamin B12 is coordinated to five N atoms, four contributed by a tetrapyrrole
(corrin); the sixth ligand is C, provided either by C5 of deoxyadenosine in enzymes
such as methylmalonyl-CoA mutase (fatty acid metabolism) or by a methyl
group in the enzyme that synthesizes the amino acid methionine in bacteria.
Nickel is a component of a hydrolase (urease), of hydrogenase, of CO dehydrogenase,
and of S-methyl CoM reductase, which catalyzes the terminal step
in methane production by methanogenic bacteria. All the Ni-proteins known to
date are from plants or bacteria. 13,14 However, about 50 years elapsed between
the crystallization of jack-bean urease in 1925 and the identification of the nickel
component in the plant protein. Thus it is premature to exclude the possibility
of Ni-proteins in animals. Despite the small number of characterized Ni-proteins,
it is clear that many different environments exist, from apparently direct
coordination to protein ligands (urease) to the tetrapyrrole F430 in methylreductase
and the multiple metal sites of Ni and Fe-S in a hydrogenase from the
bacterium Desulfovibrio gigas. Specific environments for nickel are also indicated
for nucleic acids (or nucleic acid-binding proteins), since nickel activates
the gene for hydrogenase. 15
Manganese plays a critical role in oxygen evolution catalyzed by the proteins
of the photosynthetic reaction center. The superoxide dismutase of bacteria
and mitochondria, as well as pyruvate carboxylase in mammals, are also manganese
proteins. 16,17 How the multiple manganese atoms of the photosynthetic
reaction center participate in the removal of four electrons and protons from
water is the subject of intense investigation by spectroscopists, synthetic inorganic
chemists, and molecular biologists. 17
I. GENERAL PRINCIPLES
Vanadium and chromium have several features in common, from a bioinorganic
viewpoint. 18a First, both metals are present in only small amounts in most
organisms. Second, the biological roles of each remain largely unknown. 18 Finally,
each has served as a probe to characterize the sites of other metals, such
as iron and zinc. Vanadium is required for normal health, and could act in vivo
either as a metal cation or as a phosphate analogue, depending on the oxidation
state, V(lV) or V(V), respectively. Vanadium in a sea squirt (tunicate), a primitive
vertebrate (Figure 1.2B), is concentrated in blood cells, apparently as the
major cellular transition metal, but whether it participates in the transport of
dioxygen (as iron and copper do) is not known. In proteins, vanadium is a
cofactor in an algal bromoperoxidase and in certain prokaryotic nitrogenases.
Chromium imbalance affects sugar metabolism and has been associated with the
glucose tolerance factor in animals. But little is known about the structure of
the factor or of any other specific chromium complexes from plants, animals,
or bacteria.
Molybdenum proteins catalyze the reduction of nitrogen and nitrate, as well
as the oxidation of aldehydes, purines, and sulfite. 19 Few Mo-proteins are known
compared to those involving other transition metals. Nitrogenases, which also
contain iron, have been the focus of intense investigations by bioinorganic chemists
and biologists; the iron is found in a cluster with molybdenum (the iron-molybdenum
cofactor, or FeMoCo) and in an iron-sulfur center (Chapter 7). Interestingly,
certain bacteria (Azotobacter) have alternative nitrogenases, which are
produced when molybdenum is deficient and which contain vanadium and iron
or only iron. All other known Mo-proteins are also Fe-proteins with iron centers,
such as tetrapyrroles (heme and chlorins), Fe-sulfur clusters, and, apparently,
non-heme/non-sulfur iron. Some Mo-proteins contain additional cofactors
such as the Havins, e.g., in xanthine oxidase and aldehyde oxidase. The number
of redox centers in some Mo-proteins exceeds the number of electrons transferred;
reasons for this are unknown currently.
B. Chemical Properties Relative to Storage and Transport
1. Iron
Iron is the most abundant transition element in the Earth's crust and, in
general, in all life forms. An outline of the distribution of iron in the Earth's
crust 20
,21 is shown in Table 1.2. As can be seen, approximately one-third of the
Earth's mass is estimated to be iron. Of course, only the Earth's crust is relevant
for life forms, but even there it is the most abundant transition element. Its
concentration is relatively high in most crustal rocks (lowest in limestone, which
is more or less pure calcium carbonate). In the oceans, which constitute 70
percent of the Earth's surface, the concentration of iron is low but increases
with depth, since this iron exists as suspended particulate matter rather than as
a soluble species. Iron is a limiting factor in plankton growth, and the rich
5
6 1 / TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
Table 1,2
Iron: Its terrestrial distribution.a
One third of Earth's mass, most abundant element by weight
Distribution in crustal rocks (weight %):
igneous 5.6
shale 4.7
sandstone 1.0
limestone 0.4
Ocean (70% of Earth's surface):
0.003--0.1 ppb, increasing with depth; limiting factor in plankton
growth
Rivers:
0.07-7 ppm
Ksp for Fe(OH)3 is approximately 10 -39, hence at pH 7 [Fe3+] 10 -IS M
a Data from References la and 20.
fisheries associated with strong up-welling of ocean depths result at least in part
from the biological growth allowed by these iron supplies. Properties that dominate
the transport behavior of most transition metal ions are: (l) redox chemistry,
(2) hydrolysis, and (3) the solubility of the metal ions in various complexes,
particularly the hydroxides.
As an example of the effects of solubility, consider the enormous variation
in the concentration of iron in rivers, depending on whether the water is from a
clear mountain stream running over rock or a muddy river carrying large amounts
of sediment. However, the amount of dissolved iron in the form of free ferric
ion or its hydrolysis products, whatever the source of water, is extremely low.
As can be seen from the solubility of hydrated Fe(III) (Ks ~ 10 -18 M) (Table
1.2), the concentration of free ferric ion is extraordinarily low at neutral pH; so
significant concentrations of soluble iron species can be attained only by strong
complex formation.
One example of the versatility of iron as a function of its environment is
how the ligand field can strongly alter the structural and ligand exchange properties
of the metal ion (Figure 1.3). The ligand field can also alter the redox
properties. For high-spin ferric ion, as found in the aquo complex or in many
other complexes (including the class of microbial iron-transport agents called
siderophores, to be discussed later), the coordination geometry is octahedral or
pseudo-octahedral. In the relatively weak ligand field (high-spin ground state),
the complex is highly labile. In a strong ligand field, such as an axially ligated
porphyrin complex of ferric ion, or the simple example of the ferrocyanide anion,
the low-spin complex is exchange-inert. Similarly, the high-spin octahedral
ferrous complexes are exchange-labile, but the corresponding axially ligated
porphyrin complexes, or the ferrocyanide complexes, are spin-paired (diamagnetic)
and ligand exchange-inert. Large, bulky ligands or constrained ligands,
such as those provided by metalloprotein and enzyme sites, can cause a tetrahedral
environment, in which both ferrous ion and ferric ion form high-spin
complexes.
7
octahedral
~ ~
Fe3+
~ ~ ~ lL lL ~
high-spin, labile low-spin, inert
~ ~
Fe2+ lL ~ ~ lL lL lL
high-spin, labile low-spin, inert
tetrahedral
~ ~ ~ ~ ~ ~
Fe3+
~ ~
Fe2+ lL ~
Figure 1.3
Versatility of Fe coordination complexes.
The distribution of specific iron complexes in living organisms depends
strongly on function. For example, although there are many different iron complexes
in the average human, the relative amounts of each type differ more than
650-fold (Table 1.3). The total amount of iron in humans is quite large, averaging
more than three and up to five grams for a healthy adult. Most of the iron
is present as hemoglobin, the plasma oxygen-transport protein, where the func-
Table 1.3
Average human Fe distribution.
Protein
Hemoglobin
Myoglobin
Transferrin
Ferritin
Hemosiderin
Catalase
Cytochrome c
Other
Function
Plasma O2 transport
Muscle O2 storage
Plasma Fe transport
Cell Fe storage
Cell Fe storage
H20 z metabolism
Electron transport
Oxidases, other enzymes, etc.
Oxidation state
of Fe
2
2
3
3
3
2
2
3"
Amount of Fe
(g)
2.6
0.13
0.007
0.52
0.48
0.004
0.004
0.14
Percent of total
65
6
0.2
13
12
0.1
0.1
3.6
8 1 / TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
tion of the iron is to deliver oxygen for respiration. A much smaller amount of
iron is present in myoglobin, a muscle oxygen-storage protein. For transport,
the most important of these iron-containing proteins is transferrin, the plasma
iron-transport protein that transfers iron from storage sites in the body to locations
where cells synthesizing iron proteins reside; the major consumers of iron
in vertebrates are the red blood cells. However, at any given time relatively
little of the iron in the body is present in transferrin, in much the same way that
at any given time in a large city only a small fraction of the population will be
found in buses or taxis. Other examples of iron-containing proteins and their
functions are included in Table 1.3 for comparison.
An example of different iron-coordination environments, which alter the
chemical properties of iron, is the difference in the redox potentials of hydrated
Fe3+ and the electron-transport protein cytochrome c (Table 1.4). The co()rdina-
Table 1.4
Fe redox potentials.
Complex
Fe(OHz)63+
Cytochrome G3
HIPIP
Cytochrome c
Rubredoxin
Ferredoxins
Coord. no., type
6, aquo complex
6, heme
4, Fe4S4(SR)4 6,
heme
4, Fe(SR)4
4, Fe4S4(SR)4Z-
770
390
350
250
-60
-400
tion environment of iron in cytochrome c is illustrated in Figure 1.4. For example,
the standard reduction potential for ferric ion in acid solution is 0.77
volts; so here ferric ion is quite a good oxidant. In contrast, cytochrome c has
a redox potential of 0.25 volts. A wide range of redox potentials for iron is
achieved in biology by subtle differences in protein structure, as listed in Table
1.4. Notice the large difference in the potential of cytochrome c and rubredoxin
(Figure 1.5), 0.25 volts vs. -0.06 volts, respectively. In polynuclear ferredoxins,
in which each iron is tetrahedrally coordinated by sulfur, reduction potentials
are near - 0.4 volts. Thus, the entire range of redox potentials, as illustrated
in Table 1.4, is more than one volt.
2. Chemical properties of zinc, copper, vanadium, chromium,
molybdenum, and cobalt
The chemical properties of the other essential transition elements simplify
their transport properties. For zinc there is only the +2 oxidation state, and the
hydrolysis of this ion is not a limiting feature of its solubility or transport. Zinc
is an essential element for both animals and plants. 8,9,20,21 In general, metal ion
uptake into the roots of plants is an extremely complex phenomenon. A crosssectional
diagram of a root is shown in Figure 1.6. It is said that both diffusion
H
H C I C=C-CH -
3 " / 2
/S-F1e-N" I
Met 80 CH C=N
/ 2 H
/CH2 t His 18
heme
group
Figure 1.4 I
Heme group and iron coordination in cytochrome c.
9
cysteine
2 amino acid
cysteine --r-e""'si--:d'--ue-"s-'--'--
Figure 1.5
Fe3+/2+ coordination in rubredoxin.
- stelephloem
tubes pericycle
cortex ------
air space
xylem
vessels
endodermis with
Casparian band
epidermis with
root hair
Figure 1.6
Transverse section of a typical root. 20 The complex features of the root hair surface that regulate
reductase and other activities in metal uptake are only beginning to be understood.
10 1 / TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
and mass flow of the soil solution are of significance in the movement of metal
ions to roots. Chelation and surface adsorption, which -are pH dependent, also
affect the availability of nutrient metal ions. Acid soil conditions in general
retard uptake of essential divalent metal ions but increase the availability (sometimes
with toxic results) of manganese, iron, and aluminum, all of which are
normally of very limited availability because of hydrolysis of the trivalent ions.
Vanadium is often taken up as vanadate, in a pathway parallel to phosphate.
18 However, its oxidation state within organisms seems to be highly variable.
Unusually high concentrations of vanadium occur in certain ascidians (the
specific transport behavior of which will be dealt with later). The workers who
first characterized the vanadium-containing compound of the tunicate, Ascidia
nigra, coined the name tunichrome. 22 The characterization of the compound as
a dicatecholate has been reported. 23
Quite a different chemical environment is found in the vanadium-containing
material isolated from the mushroom Amanita muscaria. Bayer and Kneifel,
who named and first described amavadine,24 also suggested the structure shown
in Figure 1. 7. 25 Recently the preparation, proof of ligand structure, and (by
implication) proof of the complex structure shown in Figure 1.7 have been established.
26 Although the exact role of the vanadium complex in the mushroom
Figure 1.7
A structure proposed for amavadineY
remains unclear, the fact that it is a vanadyl complex is now certain, although
it may take a different oxidation state in vivo.
The role of chromium in biology remains even more mysterious. In human
beings the isolation of "glucose tolerance factor" and the discovery that it contains
chromium goes back some time. This has been well reviewed by Mertz,
who has played a major role in discovering what is known about this elusive
and apparently quite labile compound.27 It is well established that chromium is
taken up as chromic ion, predominantly via foodstuffs, such as unrefined sugar,
which presumably contain complexes of chromium, perhaps involving sugar hydroxyl
groups. Although generally little chromium is taken up when it is administered
as inorganic salts, such as chromic chloride, glucose tolerance in many
adults and elderly people has been reported to be improved after supplementation
with 150-250 mg of chromium per day in the form of chromic chloride.
Similar results have been found in malnourished children in some studies in
Third World countries. Studies using radioactively labeled chromium have shown
that, although inorganic salts of chromium are relatively unavailable to mamcytoplasmic
reductants
Figure 1.8
The uptake-reduction model for chromate carcinogenicity. Possible sites for reduction of chromate
include the cytoplasm, endoplasmic reticulum, mitochondria, and the nucleusY
mals, brewer's yeast can convert the chromium into a usable form; so l:irewer's
yeast is today the principal source in the isolation of glucose tolerance factor
and has been used as a diet supplement.
Although chromium is essential in milligram amounts for human beings as
the trivalent ion, as chromate it is quite toxic and a recognized carcinogen. 30
The uptake-reduction model for chromate carcinogenicity as suggested by Connett
and Wetterhahn is shown in Figure 1.8. Chromate is mutagenic in bacterial
and mammalian cell systems, and it has been hypothesized that the difference
between chromium in the +6 and +3 oxidation states is explained by the' 'uptake-
reduction" model. Chromium(III), like the ferric ion discussed above, is
readily hydrolyzed at neutral pH and extremely insoluble. Unlike Fe 3+ , it
undergoes extremely slow ligand exchange. For both reasons, transport of chromium(
III) into cells can be expected to be extremely slow unless it is present as
specific complexes; for example, chromium(III) transport into bacterial cells has
been reported to be rapid when iron is replaced by chromium in the siderophore
iron-uptake mediators. However, chromate readily crosses cell membranes and
enters cells, much as sulfate does. Because of its high oxidizing power, chromate
can undergo reduction inside organelles to give chromium(m), which binds
to small molecules, protein, and DNA, damaging these cellular components.
11
12 1 / TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
In marked contrast to its congener, molybdenum is very different from chromium
in both its role in biology and its transport behavior, again because of
fundamental differences in oxidation and coordination chemistry properties. In
contrast to chromium, the higher oxidation states of molybdenum dominate its
chemistry, and molybdate is a relatively poor oxidant. Molybdenum is an essential
element in many enzymes, including xanthine oxidase, aldehyde reductase,
and nitrate reductase. 19 The range of oxidation states and coordination geometries
of molybdenum makes its bioinorganic chemistry particularly interesting
and challenging.
The chemistry of iron storage and transport is dominated by high concentrations,
redox chemistry (and production of toxic-acting oxygen species), hydrolysis
(pKa is about 3, far below physiological pH), and insolubility. High-affinity
chelators or proteins are required for transport of iron and high-capacity sequestering
protein for storage. By comparison to iron, storage and transport of the
other metals are simple. Zinc, copper, vanadium, chromium, manganese, and
molybdenum appear to be transported as simple salts or loosely bound protein
complexes. In vanadium or molybdenum, the stable anion, vanadate or molybdate,
appears to dominate transport. Little is known about biological storage of
any metal except iron, which is stored in ferritin. However, zinc and copper are
bound to metallothionein in a fonn that may participate in storage.
II. BIOLOGICAL SYSTEMS OF METAL STORAGE, TRANSPORT,
AND MINERALIZATION
A. Storage
1. The storage of iron
Three properties of iron can account for its extensive use in terrestrial biological
reactions:
(a) facile redox reactions of iron ions;
(b) an extensive repertoire of redox potentials available by ligand substitution
or modification (Table 1.4);
(c) abundance and availability (Table 1.1) under conditions apparently extant
when terrestrial life began (see Section LB.).
Ferrous ion appears to have been the environmentally stable form during
prebiotic times. The combination of the reactivity of ferrous ion and the relatively
large amounts of iron used by cells may have necessitated the storage of
ferrous ion; recent results suggest that ferrous ion may be stabilized inside ferritin
long enough to be used in some types of cells. As primitive organisms
began to proliferate, the successful photosynthetic cells, which trapped solar
energy by reducing CO2 to make carbohydrates (CH20)n and produce O2 , exhausted
from the environment the reductants from H2 or H2S or NH3 . The abilII.
BIOLOGICAL SYSTEMS OF METAL STORAGE, TRANSPORT, AND MINERALIZATION
ity of primitive organisms to switch to the use of H20 as a reductant, with the
concomitant production of dioxygen, probably produced the worst case of environmental
pollution in terrestrial history. As a result, the composition of the
atmosphere, the course of biological evolution, and the oxidation state of environmental
iron all changed profoundly. Paleogeologists and meteorologists estimate
that there was a lag of about 200-300 million years between the first
dioxygen production and the appearance of significant dioxygen concentrations
in the atmosphere, because the dioxygen produced at first was consumed by the
oxidation of ferrous ions in the oceans. The transition in the atmosphere, which
occurred about 2.5 billion years ago, caused the bioavailability of iron to plummet
and the need for iron storage to increase. Comparison of the solubility of
Fe 3+ at physiological conditions (about 10 - 18 M) to the iron content of cells
(equivalent to 10 -5 to 10 -8 M) emphasizes the difficulty of acquiring sufficient
Iron.
Iron is stored mainly in the ferritins, a family * of proteins composed of a
protein coat and an iron core of hydrous ferric oxide [Fe203(H20)n] with various
amounts of phosphate.6,7 As many as 4,500 iron atoms can be reversibly stored
inside the protein coat in a complex that is soluble; iron concentrations equivalent
to 0.25 M [about 10 16-fold more concentrated than Fe(III) ions] can be
easily achieved in vitro (Figure 1.1). Ferritin is found in animals, plants, and
even in bacteria; the role of the stored iron varies, and includes intracellular use
for Fe-proteins or mineralization, long-term iron storage for other cells, and
detoxification of excess iron. Iron regulates the synthesis of ferritin, with large
amounts of ferritin associated with iron excess, small or undetectable amounts
associated with iron deficiency. [Interestingly, the template (mRNA) for ferritin
synthesis is itself stored in cells and is recruited by intracellular iron or a derivative
for efficient translation into protein. 31 Iron does not appear to interact
directly with ferritin mRNA nor with a ferritin mRNA-specific regulatory (binding)
protein; however, the specific, mRNA regulatory (binding) protein has sequence
homology to aconitase, and formation of an iron-sulfate cluster prevents
RNA binding.] Because iron itself determines in part the amount of ferritin in
an organism, the environmental concentration of iron needs to be considered
before one can conclude that an organism or cell does not have ferritin.
Ferritin is thought to be the precursor of several forms of iron in living
organisms, including hemosiderin, a form of storage iron found mainly in animals.
The iron in hemosiderin is in a form very similar to that in ferritin, but
the complex with protein is insoluble, and is usually located within an intracellular
membrane (lysosomes). Magnetite (Fe304) is another form of biological
iron derived, apparently, from the iron in ferritin. Magnetite plays a role in the
behavior of magnetic bacteria, bees, and homing pigeons (see Section II.C).
The structure of ferritin is the most complete paradigm for bioinorganic
chemistry because of three features: the protein coat, the iron-protein interface,
and the iron core. 6,7
* A family of proteins is a group of related but distinct proteins produced in a single organism and usually
encoded by multiple, related genes.
13
14
(A) (B)
Figure 1.9
(A) The protein coat of horse spleen apoferritin deduced from x-ray diffraction of crystals of the
protein. 32 The outer surface of the protein coat shows the arrangement of the 24 ellipsoidal polypeptide
subunits. N refers to the N-terminus of each polypeptide and E to the E-helix (see B).
Note the channels that form at the four-fold axes where the E-helices interact, and at the threefold
axes near the N-termini of the subunits. (B) A ribbon model of a subunit showing the
packing of the four main alpha-helices (A, B, C, and D), the connecting L-loop and the E-helix.
Protein Coat Twenty-four peptide chains (with about 175 amino acids each),
folded into ellipsoids, pack to form the protein coat, * which is a hollow sphere
about 100 A in diameter; the organic surface is about 10 A thick (Figure 1.9).
Channels which occur in the protein coat at the trimer interfaces may be involved
in the movement of iron in and out of the protein. 62,63,65 Since the protein
coat is stable with or without iron, the center of the hollow sphere may be
filled with solvent, with Fe203' H20, or, more commonly, with both small aggregates
of iron and solvent. Very similar amino-acid sequences are found in
ferritin from animals and plants. Sorting out which amino acids are needed to
form the shape of the protein coat and the ligands for iron core formation requires
the continued dedication of bioinorganic chemists; identification of tyrosine
as an Fe(III)-ligand adds a new perspective. 64
Iron-Protein Interface Formation of the iron core appears to be initiated at
an Fe-protein interface where Fe(II)-O-Fe(Ill) dimers and small clusters of Fe(Ill)
atoms have been detected attached to the protein and bridged to each other by
oxo/hydroxo bridges. Evidence for multiple nucleation sites has been obtained
* Some ferritin subunits, notably in ferritin from bacteria, bind heme in a ratio of less than one heme per
two subunits. A possible role of such heme in the oxidation and reduction of iron in the core is being
investigated.
II. BIOLOGICAL SYSTEMS OF METAL STORAGE, TRANSPORT, AND MINERALIZATION 15
from electron microscopy of individual ferritin molecules (multiple core crystallites
were observed) and by measuring the stoichiometry of binding of metal
ions, which compete with binding of monoatomic iron, e.g., VO(IV) and Th(m)
(about eight sites per molecule). EXAFS (Extended X-ray Absorption Fine
Structure) and Mossbauer spectroscopies suggest coordination of Fe to the protein
by carboxyl groups from glutamic (Glu) and aspartic (Asp) acids. Although
groups of Glu or Asp are conserved in all animal and plant ferritins, the
ones that bind iron are not known. Tyrosine is an Fe(III)-ligand conserved in
rapid mineralizing ferritins identified by Uv-vis and resonance Raman spectroscopy.
64
Iron Core Only a small fraction of the iron atoms in ferritin bind directly
to the protein. The core contains the bulk of the iron in a polynuclear aggregate
with properties similar to ferrihydrite, a mineral found in nature and formed
experimentally by heating neutral aqueous solutions of Fe(III)(N03h. X-ray diffraction
data from ferritin cores are best fit by a model with hexagonal closepacked
layers of oxygen that are interrupted by irregularly incomplete layers of
octahedrally coordinated Fe(III) atoms. The octahedral coordination is confirmed
by Mossbauer spectroscopy and by EXAFS, which also shows that the
average Fe(In) atom is surrounded by six oxygen atoms at a distance of 1.95 A
and six iron atoms at distances of 3.0 to 3.3 A.
Until recently, all ferritin cores were thought to be microcrystalline and to
be the same. However, x-ray absorption spectroscopy, Mossbauer spectroscopy,
and high-resolution electron microscopy of ferritin from different sources have
revealed variations in the degree of structural and magnetic ordering and/or the
level of hydration. Structural differences in the iron core have been associated
with variations in the anions present, e.g., phosphate 29 or sulfate, and with the
electrochemical properties of iron. Anion concentrations in tum could reflect
both the solvent composition and the properties of the protein coat. To understand
iron storage, we need to define in more detail the relationship of the
ferritin protein coat and the environment to the redox properties of iron in the
ferritin core.
Experimental studies of ferritin formation show that Fe(n) and dioxygen are
needed, at least in the early stages of core formation. Oxidation to Fe(nI) and
hydrolysis produce one electron and an average of 2.5 protons for iron atoms
incorporated into the ferritin iron core. Thus, formation of a full iron core of
4,500 iron atoms would produce a total of 4,500 electrons and 11,250 protons.
After core formation by such a mechanism inside the protein coat, the pH would
drop to 0.4 if all the protons were retained. It is known that protons are released
and electrons are transferred to dioxygen. However, the relative rates of proton
release, oxo-bridge formation, and electron transfer have not been studied in
detail. Moreover, recent data indicate migration of iron atoms during the early
stages of core formation and the possible persistence of Fe 2+ for periods of
time up to 24 hours. When large numbers of Fe(n) atoms are added, the protein
coat appears to stabilize the encapsulated Fe(n).34a,b Formation of the iron core
16 1 / TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
of ferritin has analogies to surface corrosion, in which electrochemical gradients
are known to occur. Whether such gradients occur during ferritin formation and
how different protein coats might influence proton release or alter the structure
of the core are subjects only beginning to be examined.
2. The storage of zinc, copper, vanadium, chromium, molybdenum,
cobalt, nickel, and manganese
Ions of nonferrous transition metals require a much less complex biological
storage system, because the solubilities are much higher (210 - 8 M) than those
for Fe 3+ . As a result, the storage of nonferrous transition metals is less obvious,
and information is more limited. In addition, investigations are more difficult
than for iron, because the amounts in biological systems are so small. Essentially
nothing is known yet about the storage of vanadium, chromium, molybdenum,
cobalt, nickel, and manganese, with the possible exception of accumulations
of vanadium in the blood cells of tunicates.
Zinc and copper, which are used in the highest concentrations of any of
the non-ferrous transition metals, are specifically bound by the protein
metallothionein 35,36 (see Figure 1.10). Like the ferritins, the metallothioneins
are a family of proteins, widespread in nature and regulated by the metals they
bind. In contrast to ferritin, the amounts of metal stored in metallothioneins are
smaller (up to twelve atoms per molecule), the amount of protein in cells is
less, and the template (mRNA) is not stored. Because the cellular concentrations
of the metallothioneins are relatively low and the amount of metal needed is
relatively small, it has been difficult to study the biological fate of copper and
zinc in living organisms, and to discover the natural role of metallothioneins.
However, the regulation of metallothionein synthesis by metals, hormones, and
growth factors attests to the biological importance of the proteins. The unusual
metal environments of metallothioneins have attracted the attention of bioinorganic
chemists.
Metallothioneins, especially in higher animals, are small proteins 35,36 rich
in cysteine (20 per molecule) and devoid of the aromatic amino acids phenylalanine
and tyrosine. The cysteine residues are distributed throughout the peptide
chain. However, in the native form of the protein (Figure 1.10), the peptide
chains fold to produce two clusters of -SH, which bind either three or four
atoms of zinc, cadmium, cobalt, mercury, lead, or nickel. Copper binding is
distinct from zinc, with 12 sites per molecule.
In summary, iron is stored in iron cores of a complicated protein. Ferritin,
composed of a hollow protein coat, iron-protein interface, and an inorganic core,
overcomes the problems of redox and hydrolysis by directing the formation of
the quasi-stable mineral hydrous ferric oxide inside the protein coat. The outer
surface of the protein is generally hydrophilic, making the complex highly soluble;
equivalent concentrations of iron are :::::0.25 M. By contrast to iron, storage
of zinc, copper, chromium, manganese, vanadium, and molybdenum is relatively
simple, because solubility is high and abundance is lower. Little is known
Figure 1.10
The three-dimensional structure of the a domain from rat cd7 metallothionein-2, determined by
NMR in solution (Reference 36a), based on data in Reference 36b. The four metal atoms,
bonded to the sulfur of cysteine side chains, are indicated as spherical collections of small dots.
A recent description of the structure of the cdsZn2 protein, determined from x-ray diffraction of
crystals, agrees with the structure determined by NMR (Reference 36c).
about the molecules that store these metals, with the possible exception of metallothionein,
which binds small clusters of zinc or copper.
B. Transport
1. Iron
The storage of iron in humans and other mammals has been dealt with in
the previous section. Only a small fraction of the body's inventory of iron is in
transit at any moment. The transport of iron from storage sites in cellular ferritin
or hemosiderin occurs via the serum-transport protein transferrin. The transferrins
are a class of proteins that are bilobal, with each lobe reversibly (and essentially
independently) binding ferric ion. 37-39 This complexation of the metal
cation occurs via prior complexation of a synergistic anion that in vivo is bicarbonate
(or carbonate). Serum transferrin is a monomeric glycoprotein of molecular
weight 80 kDa. The crystal structure of the related protein, lactoferrin,39
has been reported, and recently the structure of a mammalian transferrin 40 has
been deduced.
Ferritin is apparently a very ancient protein and is found in higher animals,
plants, and even microbes; in plants and animals a common ferritin progenitor
17
18 1 / TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
is indicated by sequence conservation. 41 In contrast, transferrin has been in ex"
istence only relatively recently, since it is only found ia the phylum Chordata.
Although the two iron-binding sites of transferrin are sufficiently different to be
distinguishable by kinetic and a few other studies, their coordination environments
have been known for some time to be quite similar. This was first discovered
by various spectroscopies, and most recently was confirmed by crystalstructure
analysis, which shows that the environment involves two phenolate
oxygens from tyrosine, two oxygens from the synergistic, bidentate bicarbonate
anion, nitrogen from histidine, and (a surprise at the time of crystal-structure
analysis) an oxygen from a carboxylate group of an aspartate. 39
The transferrins are all glycoproteins, and human serum transferrin contains
about 6 percent carbohydrate. These carbohydrate groups are linked to the protein,
and apparently strongly affect the recognition and conformation of the native
protein.
Although transferrins have a high molecular weight and bind only two iron
atoms, transferrin is relatively efficient, because it is used in many cycles of
iron transport in its interaction with the tissues to which it delivers iron. Transferrin
releases iron in vivo by binding to the cell surface and forming a vesicle
inside the cell (endosome) containing a piece of the membrane with transferrin
and iron still complexed. The release of the iron from transferrin occurs in the
relatively low pH of the endosome, and apoprotein is returned to the outside of
the cell for delivery of another pair of iron atoms. This process in active reticulocytes
(immature red blood cells active in iron uptake) can tum over roughly
a million atoms of iron per cell per minute. 38 A schematic structure of the
protein, deduced from crystal-structure analysis, is shown in Figure 1.11. Transferrin
is an ellipsoidal protein with two subdomains or lobes, each of which
binds iron. The two halves of each subunit are more or less identical, and are
connected by a relatively small hinge. In human lactoferrin, the coordination
site of the iron is the same as the closely related serotransferrin site. A major
question that remains about the mechanism of iron binding and release is how
the protein structure changes in the intracellular compartment of low pH to release
the iron when it forms a specific complex with cell receptors (transferrin
binding proteins) and whether the receptor protein is active or passive in the
process. Recent studies suggest that the cell binding site for transferrin (a membrane,
glycoprotein called the transferrin receptor) itself influences the stability
of the iron-transferrin complex. The path of iron from the endosome to Feproteins
has not been established; and the form of transported intracellular iron
is not known.
Another major type of biological iron transport occurs at the biological opposite
of the higher organisms. Although almost all microorganisms have iron
as an essential element, bacteria, fungi, and other microorganisms (unlike humans
and other higher organisms) cannot afford to make high-molecular-weight
protein-complexing agents for this essential element when those complexing agents
would be operating extracellularly and hence most of the time would be lost to
the organism. As described earlier, the first life forms on the surface of the
II
19
Figure 1.11
Three-dimensional structure of lactotransferrin. Top: schematic representation of the folding pattern
of each lactoferrin lobe; Domain 1 is based on a beta-sheet of four parallel and two antiparallel
domains; Domain II is formed from four parallel and one antiparallel strand. Bottom: stereo
Ca diagram of the N lobe of lactoferrin; (e) iron atom between domain 1 (residues 6-90 +)
and domain II (residues 91-251); (_) disulfide bridges; (*) carbohydrate attachment site. See
Reference 39.
20 1 I TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
Earth grew in a reducing atmosphere, in which the iron was substantially more
available because it was present as ferrous-containing compounds. In contrast
to the profoundly insoluble ferric hydroxide, ferrous hydroxide is relatively soluble
at near neutral pH. It has been proposed that this availability of iron in the
ferrous state was one of the factors that led to its early incorporation in so many
metabolic processes of the earliest chemistry of life. 6,38 In an oxidizing environment,
microorganisms were forced to deal with the insolubility of ferric hydroxide
and hence when facing iron deficiency secrete high-affinity iron-binding
compounds called siderophores (from the Greek for iron carrier). More than 200
naturally occurring siderophores have been isolated and characterized to date. 42
Most siderophore-mediated iron-uptake studies in microorganisms have been
performed by using cells obtained under iron-deficient aerobic growth conditions.
However, uptake studies in E. coli grown under anaerobic conditions
have also established the presence of siderophore-specific mechanisms. In both
cases, uptake of the siderophore-iron complex is both a receptor- and an energydependent
process. In some studies the dependence of siderophore uptake rates
on the concentration of the iron-siderophore complex has been found to conform
to kinetics characteristic of protein catalysts, i.e., Michaelis-Menten kinetics.
For example, saturable processes with very low apparent dissociation constants
of under one micromolar (l fLM) have been observed for ferric-enterobactin
transport in E. coli (a bacterium), as shown in Figure 1.12. Similarly, in a very
80
c
"E 60
complex concentration (J..lM)
Figure 1.12
Effect of MECAM analogues
on iron uptake from E. coli.
Iron transport by 2 f.LM ferric
enterobactin is inhibited by
ferric MECAM.
OJ
Em
""6
E
Eo
Q)
"§
Q)
C""al
Q.
::J
Q)
LL
en
en
40
20
-5
2 4 6
II. BIOLOGICAL SYSTEMS OF METAL STORAGE, TRANSPORT, AND MINERALIZATION 21
different microorganism, the yeast Rhodoturala pilimanae, Michaelis-Menten
kinetics were seen again with a dissociation constant of approximately 6 JLM
for the ferric complex of rhodotoroulic acid; diagrams of some representative
siderophores are shown in Figure 1. 13. The siderophore used by the fungus
Neurospora crassa was found to have a dissociation constant of about 5 JLM
and, again, saturable uptake kinetics.
o;l :
OH
~ N~
CH3
S /
N
S~COOH
pyochelin
mycobactin P
pseudobactin
UOH
yOH
enterobactin
Figure 1.13
Examples of bacterial siderophores. See Reference 42.
22 4 [Fe(ent)P- (or analogue)
\ I
\ I
tepA protein receptor
~~==:t:I:::::::::::::::::::::::=============~~o:uter membrane .
periplasmic space
cytoplasm
leakage ot
[Fe (ent)] 3to
solution
Figure 1.14
Model for enterobactin-mediated Fe uptake in E. coli.
Although the behavior just described seems relatively simple, transport
mechanisms in living cells probably have several more kinetically distinct steps
than those assumed for the simple enzyme-substrate reactions underlying the
Michaelis-Menten mechanism. For example, as ferric enterobactin is accumulated
in E. coli, it has to pass through the outer membrane, the periplasm, and
the cytoplasm membrane, and is probably subjected to reduction of the metal in
a low-pH compartment or to ligand destruction.
A sketch of a cell of E. coli and some aspects of its transport behavior are
shown in Figure 1.14. Enterobactin-mediated iron uptake in E. coli is one of
the best-characterized of the siderophore-mediated iron-uptake processes in microorganisms,
and can be studied as a model. After this very potent iron-sequestering
agent complexes iron, the ferric-enterobactin complex interacts with
a specific receptor in the outer cell membrane (Figure 1.14), and the complex
is taken into the cell by active transport. The ferric complexes of some synthetic
analogs of enterobactin can act as growth agents in supplying iron to E. coli.
Such a feature could be used to discover which parts of the molecule are involved
in the sites of structural recognition of the ferric-enterobactin complex.
Earlier results suggested that the metal-binding part of the molecule is recognized
by the receptor, whereas the ligand platform (the triserine lactone ring;
see Figure 1.13) is not specifically recognized.
To find out which domains of enterobactin are required for iron uptake and
recognition, rhodium complexes were prepared with various domains of enterobactin
(Figure 1.15) as ligands to use as competitors for ferric enterobactin. 44
The goal was to find out if the amide groups (labeled Domain II in Figure 1.15),
Domain:
(III) metal binding unit
(II) amide linkage
(I) backbone
Figure 1.15
Definition of recognition domains in enterobactin.
which linked the metal-binding catechol groups (Domain III, Figure 1.15) to the
central ligand backbone (Domain I, Figure 1.15), are necessary for recognition
by the receptor protein. In addition, synthetic ligands were prepared that differed
from enterobactin by small changes at or near the catecholate ring. Finally,
various labile trivalent metal cations, analogous to iron, were studied to
see how varying the central metal ion would affect the ability of metal enterobactin
complexes to inhibit competitively the uptake of ferric enterobactin by
the organism. For example, if rhodium MECAM (Figure 1.16) is recognized by
the receptor for ferric enterobactin on living microbial cells, a large excess of
rhodium MECAM will block the uptake of radioactive iron added as ferric enterobactin.
In fact, the rhodium complex completely inhibited ferric-enterobactin
uptake, proving that Domain I is not required for recognition of ferric enterobactin.
However, if only Domain III is important in recognition, it would be expected
that the simple tris(catecholato)-rhodium(III) complex would be an equally
good inhibitor. In fact, even at concentrations in which the rhodium-catechol
complex was in very large excess, no inhibition of iron uptake was observed,
suggesting that Domain II is important in the recognition process.
The role of Domain II in the recognition process was probed by using a
rhodium dimethyl amide of 2,3-dihydroxybenzene (DMB) as a catechol ligand,
with one more carbonyl ligand than in the tris(catecholato)-rhodium(III) complex.
Remarkably, this molecule shows substantially the same inhibition of enterobactin-
mediated iron uptake in E. coli as does rhodium MECAM itself. Thus,
in addition to the iron-catechol portion of the molecule, the carbonyl groups
23
24
OMS
~OH
~OH
catechol
TRIMCAM
Figure 1.16
MECAM and related enterobactin analogues.
II. BIOLOGICAL SYSTEMS OF METAL STORAGE, TRANSPORT, AND MINERALIZATION 25
(Domain II) adjacent to the catechol-binding subunits of enterobactin and synthetic
analogs are required for recognition by the ferric-enterobactin receptor. In
contrast, when a methyl group was attached to the "top" of the rhodium MECAM
complex, essentially no recognition occurred.
In summary, although the structure of the outer-membrane protein receptor
of E. coli is not yet known, the composite of the results just described gives a
sketch of what the ferric-enterobactin binding site must look like: a relatively
rigid pocket for receiving the ferric-catecholate portion of the complex, and
proton donor groups around this pocket positioned to hydrogen bond to the
carbonyl oxygens of the ferric amide groups. The mechanisms of iron release
from enterobactin, though followed phenomenologically, are still not known in
detail.
2. Zinc, copper, vanadium, chromium, molybdenum, and cobalt
As described in an earlier section, transport problems posed by the six elements
listed in the heading are somewhat simpler (with the exception of chromium)
than those for iron. One very interesting recent development has been
the characterization of sequestering agents produced by plants which complex a
number of metal ions, not just ferric ions. A key compound, now well-characterized,
is mugeneic acid (Figure 1.17).45 The structural and chemical similari-
C3 C3
Figure 1.17
Structure and a stereo view of mugeneic acid. See Reference 42.
26 1 I TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
ties of mugeneic acid to ethylenediaminetetraacetic acid (EDTA) have been noted.
Like EDTA, mugeneic acid forms an extremely strong .~omplex with ferric ion,
but also forms quite strong complexes with copper, zinc, and other transitionmetal
ions. The structure of the cobalt complex (almost certainly essentially
identical with that of the iron complex) is shown in Figure 1.18. Like the siderophores
produced by microorganisms, the coordination environment accommodated
by mugeneic acid is essentially octahedral. Although the coordination
properties of this ligand are well laid out, and it has been shown that divalent
metal cations, such as copper, competitively inhibit iron uptake by this ligand,
the detailed process of metal-ion delivery by mugeneic acid and related compounds
has not been elucidated.
0(4)
0(8)
1.939(5) A
0(4)
0(8)
1.941(5) A
1.896(5)
0(3)
(A)
0(5)
N(2)
1.915(6)
0(3)
(B)
0(5)
N(2)
Figure 1.18
Molecular structures of the complexes (molecules A and B) and coordination about the cobalt
ion in molecules A and B of the mugeneic acid-Co(III) complex. Bond lengths in A; angles in
degrees. See Reference 42.
II. BIOLOGICAL SYSTEMS OF METAL STORAGE, TRANSPORT, AND MINERALIZATION 27
As noted in an earlier section, the biochemistry of vanadium potentially
involves four oxidation states that are relatively stable in aqueous solution. These
are V2+, V3+, va2+ , and V02 + (the oxidation states 2, 3, 4, and 5, respectively).
Since even without added sequestering agents, V2+ slowly reduces water
to hydrogen gas, it presumably has no biological significance. Examples of the
remaining three oxidation states of vanadium have all been reported in various
living systems. One of the most extensively investigated examples of transitionmetal-
ion accumulation in living organisms is the concentration of vanadium in
sea squirts (tunicates), which is reported to be variable; many species have vanadium
levels that are not exceptionally high. Others such as Ascidia nigra show
exceptionally high vanadium concentrations. 46
In addition to showing a remarkable concentration of a relatively exotic
transition-metal ion, tunicates are a good laboratory model for uptake experiments,
since they are relatively simple organisms. They possess a circulation
system with a one-chambered heart, and a digestive system that is essentially a
pump and an inlet and outlet valve connected by a digestive tract. The organism
can absorb dissolved vanadium directly from sea water as it passes through the
animal. The influx of vanadate into the blood cells of A. nigra has been studied
by means of radioisotopes. The corresponding influx of phosphate, sulfate, and
chromate (and the inhibition of vanadate uptake by these structurally similar
oxoanions) has been measured. In the absence of inhibitors, the influx of vanadate
is relatively rapid (a half-life on the order of a minute near ODe) and the
uptake process shows saturation behavior as the vanadate concentration is increased.
The uptake process (in contrast to iron delivery in microorganisms, for
example, and to many other uptake processes in microorganisms or higher animals)
is not energy-dependent. Neither inhibitors of glycolysis nor decouplers
of respiration-dependent energy processes show any significant effect on the rate
of vanadate influx.
Phosphate, which is also readily taken up by the cells, is an inhibitor of
vanadate influx. Neither sulfate nor chromate is taken up significantly, nor do
they act as significant inhibitors for the vanadate uptake. Agents that inhibit
transport of anions, in contrast, were found to inhibit uptake of vanadate into
the organism. These results have led to the model proposed in Figure 1.19:
(1) vanadate enters the cell through anionic channels; this process eliminates
positively charged metal ion or metal-ion complexes present in
sea water;
(2) vanadate is reduced to vanadium(III); since the product is a cation, and
so cannot be transported through the anionic channels by which vanadate
entered the cell, the vanadium(III) is trapped inside the cell-the
net result is an accumulation of vanadium. [It has been proposed that
the tunichrome could act either as a reducing agent (as the complex) or
(as the ligand) to stabilize the general vanadium(ill); however, this seems
inconsistent with its electrochemical properties (see below).]
28
vacuole
anionic
channels
x
Figure 1.19
Diagram of a vanadium accumulation mechanism. Vanadium enters the vacuole within the vanadocyte
as mononegative H2Y04-, although it may be possible for the dinegative anion, HYO~-,
to enter this channel as well (X - stands for any negative ion such as Cl- , H2PO,;- , etc., that
may exchange across the membrane through the anionic channel). Reduction to y3+ takes place
in two steps, via a Y(IY) intermediate. The resulting cations may be trapped as tightly bound
complexes, or as free ions that the anionic channel will not accept for transport. The nature of
the reducing species is unknown.
Synthetic models of tunichrome b-] (Figure] .20) have been prepared. Tunichrome
is a derivative of pyrogallol whose structure precludes the formation of
an octahedral complex of vanadium as a simple] : ] metal: ligand complex. The
close analogue, described as 3,4,5-TRENPAMH9 , also cannot form a simple
octahedral ]:] complex. In contrast, the synthetic ligands TRENCAM and 2,3,4TRENPAM
can form pseudo-octahedral complexes. The structure of the vanadium
TRENCAM complex shows that it is indeed a simple pseudo-octahedral
tris-catechol complex.47 The electrochemical behavior of these complexes is
similar, with vanadium(IVfIll) potentials of about - 0.5 to - 0.6 volts versus
NHE. These results indicate that tunichrome b-l complexes of vanadium(IVfIll)
would show similar differences in their redox couples at high pH. At neutral
pH, in the presence of excess pyrogallol groups, vanadium(IV) can be expected
to form the intensely colored tris-catechol species. However, comparison of the
EPR properties reported for vanadium-tunichrome preparations with model vanadium(
lV)-complexes would indicate predominantly bis(catechol) vanadyl coordination.
In any case, the vanadium(III) complexes must remain very highly
reducing. It has been pointed out that the standard potential of pyrogallol is
0.79 V and decreases 60 mV per pH unit (up to about pH 9), so that at pH 7
the potential is about 0.4 V. The potentials of the vanadium couples for the
tunichrome analogs are about - 0.4 V. It has been concluded, therefore, that
tunichrome or similar ligands cannot reduce the vanadium(IV) complex; so the
29
OH
HO OH
OH
HO'~~ N~~)Q(H
HOy 0 0
OH
Tunichrome b-l
H01:
HO
"-0
HN
CofJr:0 : ~ OH OH ~~N NKg)
OH O~ 0
OH
OH
HO~
HO
"-0
HN
« 0 ~ OH OH
HO 0 ~~N N~
OH O~ 0 OH
OH
3,4,5-TRENPAMH6
OH
HO~OH
HN
HO~O ~ OH O N~N----------Nk9t
HO H 0 O~ . OH
OH
OH
TRENCAMH6
2,3,4-TRENPAMH9
Figure 1.20
Structures of tunichrome b-l and synthetic analogues. 43
30 1 / TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
highly reducing vanadium(III) complex of tunichrome must be generated in some
other way. 47
Although a detailed presentation of examples of the known transport properties
of essential transition-metal ions into various biological systems could be
the subject of a large book, the examples that we have given show how the
underlying inorganic chemistry of the elements is used in the biological transport
systems that are specific for them. The regulation of metal-ion concentrations,
including their specific concentration when necessary from relatively low
concentrations of surrounding solution, is probably one of the first biochemical
problems that was solved in the course of the evolution of life.
Iron is transported in forms in which it is tightly complexed to small chelators
called siderophores (microorganisms) or to proteins called transferrins (animals)
or to citrate or mugeneic acid (plants). The problem of how the iron is
released in a controlled fashion is largely unresolved. The process of mineral
formation, called biomineralization, is a subject of active investigation. Vanadium
and molybdenum are transported as stable anions. Zinc and copper appear
to be transported loosely associated with peptides or proteins (plants) and possibly
mugeneic acid in plants. Much remains to be learned about the biological
transport of nonferrous metal ions.
C. Iron Biomineralization
Many structures formed by living organisms are minerals. Examples include
apatite [Ca2(OH)P04] in bone and teeth, calcite or aragonite (CaC03) in the
shells of marine organisms and in the otoconia (gravity device) of the mammalian
ear, silica (Si02) in grasses and in the shells of small invertebrates such as
radiolara, and iron oxides, such as magnetite (Fe304) in birds and bacteria (navigational
devices) and ferrihydrite FeO(OH) in ferritin of mammals, plants, and
bacteria. Biomineralization is the formation of such minerals by the influence of
organic macromolecules, e.g., proteins, carbohydrates, and lipids, on the precipitation
of amorphous phases, on the initiation of nucleation, on the growth
of crystalline phases, and on the volume of the inorganic material.
Iron oxides, as one of the best-studied classes of biominerals containing
transition metals, provide good examples for discussion. One of the most remarkable
recent characterizations of such processes is the continual deposition
of single-crystal ferric oxide in the teeth of chiton. 48 Teeth of chiton form on
what is essentially a continually moving belt, in which new teeth are being
grown and moved forward to replace mature teeth that have been abraded. However,
the study of the mechanisms of biomineralization in general is relatively
recent; a great deal of the information currently available, whether about iron in
ferritin or about calcium in bone, is somewhat descriptive.
Three different forms of biological iron oxides appear to have distinct relationships
to the proteins, lipids, or carbohydrates associated with their formation
and with the degree of crystallinity. 49 Magnetite, on the one hand, often forms
almost perfect crystals inside lipid vesicles of magneto-bacteria. 50 Ferrihydrite,
II. BIOLOGICAL SYSTEMS OF METAL STORAGE, TRANSPORT, AND MINERALIZATION 31
on the other hand, exists as large single crystals, or collections of small crystals,
inside the protein coat of ferritin; however, iron oxides in some ferritins that
have large amounts of phosphate are very disordered. Finally, goethite [aFeO(
OH)] and lepidocrocite [y-FeO(OH)] form as small single crystals in a
complex matrix of carbohydrate and protein in the teeth of some shellfish (limpets
and chitons); magnetite is also found in the lepidocrocite-containing teeth.
The differences in the iron-oxide structures reflect differences in some or all of
the following conditions during formation of the mineral: nature of co-precipitating
ions, organic substrates or organic boundaries, surface defects, inhibitors,
pH, and temperature. Magnetite can form in both lipid and protein/carbohydrate
environments, and can sometimes be derived from amorphous or semicrystalline
ferrihydrite-like material (ferritin). However, the precise relationship between
the structure of the organic phase and that of the inorganic phase has yet to be
discovered. When the goal of understanding how the shape and structure of
biominerals is achieved, both intellectual satisfaction and practical commercial
and medical information will be provided.
Synthetic iron complexes have provided models for two stages of ferritin
iron storage and biomineralization: 51-59 (1) the early stages, when small numbers
of clustered iron atoms are bound to the ferritin protein coat, and (2) the
final stages, where the bulk iron is a mineral with relatively few contacts to the
protein coat. In addition, models have begun to be examined for the microenvironment
inside the protein coat. 54
Among the models for the early or nucleation stage of iron-core formation
are the binuclear Fe(III) complexes with [Fe20(02CR2)]2+ cores;55,56 the three
other Fe(III) ligands are N. The JL-oxo complexes, which are particularly accurate
models for the binuclear iron centers in hemerythrin, purple acid phosphatases,
and, possibly, ribonucleotide reductases, may also serve as models for
ferritin, since an apparently transient Fe(II)-O-Fe(III) complex was detected during
the reconstitution of ferritin from protein coats and Fe(II). The facile exchange
of (02CR) for (02PR) in the binuclear complex is particularly significant
as a model for ferritin, because the structure of ferritin cores varies with the
phosphate content. An asymmetric trinuclear (Fe30) 7+ complex57 and an (FeO)l1
complex (Figure 1.21) have been prepared; these appear to serve as models for
later stages of core nucleation (or growth). 59
Models for the full iron core of ferritin include ferrihydrite, which matches
the ordered regions of ferritin cores that have little phosphate; however, the site
vacancies in the lattice structure of ferrihydrite [FeO(OH)] appear to be more
regular than in crystalline regions of ferritin cores. A polynuclear complex of
iron and microbial dextran (a-l,4-D-glucose)n has spectroscopic (M6ssbauer,
EXAFS) properties very similar to those of mammalian ferritin, presumably
because the organic ligands are similar to those of the protein (-OH, -COOH).
In contrast, a polynuclear complex of iron and mammalian chondroitin sulfate
(a-l ,4-[a-1 ,3-D-glucuronic acid-N-acetyl-D-galactosamine-4-sulfate]n) contains
two types of domains: one like mammalian ferritin [FeO(OH)] and one like
hematite (a-Fe203), which was apparently nucleated by the sulfate, emphasizing
32
06
032"()033
Figure 1.21
The structure of a model for a possible intermediate in the formation of the ferritin iron core.
The complex consists of 11 Fe(III) atoms with internal oxo-bridges and a coat of benzoate ligands;
the Fe atoms define a twisted, pentacapped trigonal prism. See Reference 53.
the importance of anions in the structure of iron cores. 60 Finally, a model for
iron cores high in phosphate, such as those from bacteria, is Fe-ATP (4: 1), in
which the phosphate is distributed throughout the polynuclear iron complex,
providing an average of 1 or 2 of the 6 oxygen ligands for iron. 61
The microenvironment inside the protein coat of ferritin has recently been
modeled by encapsulating ferrous ion inside phosphatidylcholine vesicles and
studying the oxidation of iron as the pH is raised. The efficacy of such a model
is indicated by the observation of relatively stable mixtures of Fe(II)/Fe(III)
inside the vesicles, as have also been observed in ferritin reconstituted experimentally
from protein coats and ferrous ion. 43
,54
Models for iron in ferritin must address both the features of traditional metalprotein
interactions and the bulk properties of materials. Although such modeling
may be more difficult than other types of bioinorganic modeling, the difficulties
are balanced by the availability of vast amounts of information on FeIII.
SUMMARY 33
protein interactions, corrosion, and mineralization. Furthermore, powerful tools
such as x-ray absorption, Mossbauer and solid state NMR spectroscopy, scanning
electron and proton microscopy, and transmission electron microscopy reduce
the number of problems encountered in modeling the ferritin ion core.
Construction of models for biomineralization is clearly an extension of modeling
for the bulk phase of iron in ferritin, since the major differences between
the iron core of ferritin and that of other iron-biominerals are the size of the
final structure, the generally higher degree of crystallinity, and, at this time, the
more poorly defined organic phases. A model for magnetite formation has been
provided by studying the coulometric reduction of half the Fe 3+ atoms in the
iron core of ferritin itself. Although the conditions for producing magnetite have
.yet to be discovered, the unexpected observation of retention of the Fe 2+ by
the protein coat has provided lessons for understanding the iron core of ferritin.
Phosphatidyl choline vesicles encapsulating Fe 2+ appear to serve as models for
both ferritin and magnetite; only further investigation will allow us to understand
the unique features that convert Fe 2+ to [FeO(OH)], on the one hand, and
Fe304, on the other.
III. SUMMARY
Transition metals (Fe, Cu, Mo, Cr, Co, Mn, V) play key roles in such biological
processes as cell division (Fe, Co), respiration (Fe, Cu), nitrogen fixation
(Fe, Mo, V), and photosynthesis (Mn, Fe). Zn participates in many hydrolytic
reactions and in the control of gene activity by proteins with "zinc fingers."
Among transition metals, Fe predominates in terrestial abundance; since Fe is
involved in a vast number of biologically important reactions, its storage and
transport have been studied extensively. Two types of Fe carriers are known:
specific proteins and low-molecular-weight complexes. In higher animals, the
transport protein transferrin binds two Fe atoms with high affinity; in microorganisms,
iron is transported into cells complexed with catecholates or hydroxamates
called siderophores; and in plants, small molecules such as citrate, and
possibly plant siderophores, carry Fe. Iron complexes enter cells through complicated
paths involving specific membrane sites (receptor proteins). A problem
yet to be solved is the form of iron transported in the cell after release from
transferrin or siderophores but before incorporation into Fe-proteins.
Iron is stored in the protein ferritin. The protein coat of ferritin is a hollow
sphere of 24 polypeptide chains through which Fe2+ passes, is oxidized, and
mineralizes inside in various forms of hydrated Fe203. Control of the formation
and dissolution of the mineral core by the protein and control of protein synthesis
by Fe are subjects of current study.
Biomineralization occurs in the ocean (e.g., Ca in shells, Si in coral reefs)
and on land in both plants (e.g., Si in grasses) and animals (e.g., Ca in bone,
Fe in ferritin, Fe in magnetic particles). Specific organic surfaces or matrices of
protein and/or lipid allow living organisms to produce minerals of defined shape
and composition, often in thermodynamically unstable states.
34 1 I TRANSITION-METAL STORAGE, TRANSPORT, AND BIOMINERALIZATION
IV. REFERENCES
1. (a) J. H. Martin and R. M. Gordon, Deep Sea Research 35 (1988), 177; (b) F. Egami, 1. Mol. Evol. 4
(1974), 113.
2. J. F. Sullivan et al., 1. Nutr. 109 (1979),1432.
3. M. D. McNeely et al., Clin. Chem. 17 (1971), 1123.
4. A. R. Byrne and L. Kosta, Sci. Total Env. 10 (1978), 17.
5. A. S. Prasad, Trace Elements and Iron in Human Metabolism, Plenum Medical Book Company, 1978.
6. E. C. Theil, Adv. Inorg. Biochem. 5 (1983), 1.
7. E. C. Theil and P. Aisen, in D. van der Helm, J. Nei1ands, and G. Winkelmann, eds., Iron Transport
in Microbes, Plants, and Animals, VCH, 1987, p. 421.
8. C. F. Mills, ed., Zinc in Human Biology, Springer-Verlag, 1989.
9. B. L. Vallee and D. S. Auld, Biochemistry 29 (1990),5647.
10. J. Miller, A. D. McLachlan, and A. Klug, EMBO J. 4 (1985), 1609.
11. J. M. Berg, 1. Bioi. Chem. 265 (1990), 6513.
12. C. Sennett, L. E. G. Rosenberg, and I. S. Millman, Annu. Rev. Biochem. 50 (1981), 1053.
13. J. J. G. Moura et al., in A. V. Xavier, ed., Frontiers in Biochemistry, VCH, 1986, p. 1.
14. C. T.Walsh and W. H. Grme-Johnson, Biochemistry 26 (1987), 4901.
15. H. Kim and R. J. Maier, J. Bioi. Chem. 265 (1990), 18729.
16. Reference 5, p. 5.
17. V. L. Schramm and F. C. Wedler, eds., Manganese in Metabolism and Enzyme Function, Academic
Press, 1986.
18. (a) D. W. Boyd and K. Kustin, Adv. Inorg. Biochem. 6 (1984), 312; (b) R. C. Bruening et al., J. Nat.
Products 49 (1986), 193.
19. T. G. Spiro, ed., Molybdenum Biochemistry, Wiley, 1985.
20. L. L. Fox, Geochim. Cosmochim. Acta 52 (1988), 771.
21. M. E. Farago, in A. V. Xavier, ed., Frontiers in Bioinorganic Chemistry, VCH, 1986, p. 106.
22. I. G. Macara, G. C. McCloud, and K. Kustin, Biochem. J. 181 (1979), 457.
23. R. C. Bruening et al., 1. Am. Chem. Soc. 107 (1985), 5298.
24. E. Bayer and H. H. Kneifel, Z. Natwforsch. 27B (1972), 207.
25. E. Bayer and H. H. Kneifel, in A. V. Xavier, ed., Frontiers in Bioinorganic Chemistry, VCH, 1986, p.
98.
26. J. Felcman, J. J. R. Frausto da Silva, and M. M. Candida Vaz, Inorg. Chim. Acta 93 (1984), 101.
27. W. Mertz, Nutr. Rev. 3 (1975), 129.
28. E. C. Theil, 1. Bioi. Chem. 265 (1990), 4771; Biofactors 4 (1993), 87.
29. J. S. Rohrer et al., Biochemistry 29 (1990), 259.
30. P. H. Connett and K. Wetterhahn, Struct. Bonding 54 (1983), 94.
31. E. C. Theil, Ann. Rev. Biochem. 56 (1987), 289; Adv. Enzymol. 63 (1990), 421.
32. G. C. Ford et al., Philos. Trans. Roy. Soc. Land. B, 304 (1984),551.
33. G. D. Watt, R. B. Frankel, and G. C. Papaefthymiou, Proc. Natl. Acad. Sci. USA 82 (1985), 3640.
34. (a) J. S. Rohrer et al., J. Bioi. Chem. 262 (1987), 13385; (b) 1. S. Rohrer et al., Inorg. Chem. 28
(1989), 3393.
35. D. H. Hamer, Ann. Rev. Biochem. 55 (1986), 913.
36. A. H. Robbins, D. E. McRee, M. Williamson, S. A. Collett, N. H. Xuong, W. F. Furey, B. C. Want,
and C. D. Stout, J. Mol. Bioi. 221 (1991), 1269.
37. N. D. Chasteen, Adv. Inorg. Biochem. 5 (1983), 201.
38. P. Aisen and I. Listowsky, Annu. Rev. Biochem. 49 (1980), 357.
39. B. T. Anderson et al., Proc. Natl. Acad. Sci. USA 84 (1987), 1768; E. N. Baker, B. F. Anderson, and
H. M. Baker, Int. 1. Bioi. Macromol. 13 (1991), 122.
40. S. Bailey et al., Biochemistry 27 (1988),5804.
41. M. Ragland et al., 1. BioI. Chem. 263 (1990), 18339.
42. B. F. Matzanke, G. Muller, and K. N. Raymond, in T. M. Loehr, ed., Iron Carriers and Iron Proteins,
VCH, 1989, pp. 1-121.
43. L. Stryer, Biochemistry, Freeman, 1981, pp. 110-116.
44. D. J. Ecker et al., J. Am. Chem. Soc. 110 (1988), 2457.
45. Y. Sugiura and K. Nomoto, Struct. Bonding 58 (1984), 107.
46. S. Mann, Struct. Bonding 54 (1986), 125.
IV. REFERENCES
47. A. R. Bulls et al., J. Am. Chern. Soc. 112 (1990), 2627.
48. J. Webb, in P. Westbroek and E. W. de Jong, eds., Biomineralization and Biological Metal Accumulation,
Reidel, 1983, pp. 413-422.
49. K. Kustin et al., Struct. Bonding 53 (1983), 139.
50. R. B. Frankel and R. P. Blakemore, Philos. Trans. Roy. Soc. Lond. B 304 (1984),567.
51. S. J. Lippard, Angew. Chemie 22 (1988), 344.
52. K. E. Wieghardt, Angew. Chemie 28 (1989), 1153.
53. E. C. Theil, in R. B. Frankel, ed., Iron Biomineralization, Plenum Press, 1990.
54. S. Maun, J. P. Harrington, and R. J. P. Williams, Nature 234 (1986), 565.
55. W. H. Armstrong aud S. J. Lippard, J. Am. Chern. Soc. 107 (1985),3730.
56. L. Que, Jr. and R. C. Scarrow, in L. Que, ed., Metal Clusters in Proteins, ACS Symposium Series 372,
American Chemical Society, Washington, DC, 1988, p. 152 and references therein.
57. S. M. Gorun aud S. J. Lippard, 1. Am. Chern. Soc. 107 (1985), 4570.
58. S. M. Gorun et al., J. Am. Chern. Soc. 109 (1987), 3337.
59. Q. Islam et al., J. Inorg. Biochem. 36 (1989), 51.
60. C.-Y. Yang et al., J. Inorg. Biochem. 28 (1986), 393.
61. A. N. Mansour et aI., J. Biol. Chern. 260 (1985), 7975.
62. D. C. Hams and P. Aisen, in T. M. Loehr, ed., Iron Carriers and Iron Proteins, VCH, 1989, pp. 239-
352.
63. P. M. Harrison and T. M. Lilley, in Reference 62, pp. 123-23S.
64. G. S. Waldo et al. Science 259 (1993), 796.
65. J. Trikha, G. S. Waldo, F. A. Lewandowski, Y. Ha, E. C. Theil, P. C. Weber, and N. M. Allewell,
Protein 18 (1994), issue #2, in press.
These references contain general reviews of the subjects indicated:
Chromium: 27, 30
Cobalt: 12
Copper: 8
Iron
Biochemistry; 7, 31, 37, 42
Biomineralization polynuclear models: 6, 42, 56, 57, 58
Siderophores: 42
Structure of storage and transport proteins: 32, 62, 63
Manganese: 17
Molybdenum: 19
Nickel: 13, 14
Vanadium: 18
Zinc: 8, 9, 11, 35
35