See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/8518437

Iron, Ferritin, and Nutrition

Article in Annual Review of Nutrition · February 2004

DOI: 10.1146/annurev.nutr.24.012003.132212 · Source: PubMed

CITATIONS READS
284 1,583

1 author:

Elizabeth C Theil
North Carolina State Univ., Raleigh, NC; & Children's Hospital & Research Center Oakland, CA, United Statese
251 PUBLICATIONS 14,559 CITATIONS

SEE PROFILE

All content following this page was uploaded by Elizabeth C Theil on 27 May 2014.

The user has requested enhancement of the downloaded file.

6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212
(Some corrections may occur before final publication online and in print)
V I E W
E

S
Annu. Rev. Nutr. 2004. 24:327–43

C E
I N

doi: 10.1146/annurev.nutr.24.012003.132212
Copyright

c 2004 by Annual Reviews. All rights reserved

N
A
D V A

IRON, FERRITIN, AND NUTRITION

Elizabeth C. Theil
CHORI (Children’s Hospital Oakland Research Institute), Oakland,
California 94609; email: etheil@chori.org

Key Words iron supplementation, food iron properties, nonheme iron

■ Abstract Ferritin, a major form of endogenous iron in food legumes such as
soybeans, is a novel and natural alternative for iron supplementation strategies where
effectiveness is limited by acceptability, cost, or undesirable side effects. A member of
the nonheme iron group of dietary iron sources, ferritin is a complex with Fe3+ iron in
a mineral (thousands of iron atoms inside a protein cage) protected from complexation.
Ferritin illustrates the wide range of chemical and biological properties among nonheme
iron sources. The wide range of nonheme iron receptors matched to the structure of
the iron complexes that occurs in microorganisms may, by analogy, exist in humans.
An understanding of the chemistry and biology of each type of dietary iron source
(ferritin, heme, Fe2+ ion, etc.), and of the interactions dependent on food sources,
genes, and gender, is required to design diets that will eradicate global iron deficiency
in the twenty-first century.

CONTENTS
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 327
CONSEQUENCES OF THE CHEMICAL FORM OF IRON . . . . . . . . . . . . . . . . . . . 328
Molecular Absorption Mechanisms for Iron . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 328
Extrinsic and Intrinsic Labeling in Studies of Dietary Iron
Bioavailability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 331
FERRITIN IN NATURE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 332
Animals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 333
Plants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
Microorganisms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 334
DIETARY FERRITIN . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 335
Absorption Studies in Rats . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 336
PERSPECTIVE . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 337

INTRODUCTION
Nutritional iron is usually divided into two types: heme, where iron is absorbed as
the stable porphyrin complex unaffected by other food components, and nonheme
iron, which is envisioned as “free” or in weak complexes. Food components such
0199-9885/04/0714-0327$14.00 327
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

328 THEIL

as phytate or tannins can trap iron from weak complexes in other foods during
digestion, altering the bioavailability of food iron. Thus it is not the total amount
of food iron ingested, but the distribution among different chemical forms of iron
and the reactivity among the different chemical forms of iron ingested that will
determine the bioavailability of iron in any one meal. New information empha-
sizes the complexity of nonheme iron in food, and the apparent stability of some
nonheme iron complexes such as ferritin, where the iron is absorbed even from
phytate-rich soybeans.
New approaches to solving iron deficiency are critical because the problem
remains one of great significance in early childhood as well as in menstruating
and pregnant women, where average frequencies of iron deficiency are estimated
at 43% and reach 85% in some populations (47). Iron depletion affects not only
overall health but also cognitive development (78). The problem is not restricted to
those in poverty or to underdeveloped countries. In Japan, for example, ∼15% of
blood donors are rejected because of iron deficiency and estimates of iron deficits
in young women are as high as 25% (Y. Kohgo, personal communication). In the
United States, moreover, approximately 75% of college-aged women report low
iron intake (74), and suboptimal dietary intake of iron occurs in 90% of pregnant
Americans (84). Current iron supplementation regimens, some known for cen-
turies, can have negative consequences and side effects from the oxidative damage
of oxygen and iron chemistry (22, 61). New forms of iron supplementation are
needed, particularly for vulnerable groups such as children and menstruating and
pregnant women (9, 14, 22). Recent studies (68) confirm a 1973 study (79) showing
that soybeans (and possibly other, yet-to-be-discovered ferritin-rich foods) have
the potential to be novel, natural iron supplements that minimize dietary iron defi-
ciency (91). Recent reviews on iron absorption itself and general features of iron
nutrition include References 8, 11, 38, 66, and 100. The theme of the sections that
follow is the properties of ferritin and its chemistry and biology that distinguish it
from other forms of iron in the diet.

CONSEQUENCES OF THE CHEMICAL FORM OF IRON

Molecular Absorption Mechanisms for Iron

Absorption of iron from food requires recognition of the chemical form of the iron
by gut receptors. Both shape and charge are important in the recognition process.
The shape and chemistry of the Fe3+ iron is much more dependent on pH compared
to Fe2+ ions. At the low pH of the stomach, both Fe3+ and Fe2+ ions in solution
will be single ions surrounded by water, much like Na+, Ca2+, Zn2+, Mn2+, Cr3+,
etc., unless the Fe3+ ion is chelated (e.g., citrate, phytate, and heme), or encased
as a multi-ion form in a protein such as ferritin (Figure 1, see color insert); ferritin
is extremely stable (23, 88, 90) and survives in vitro digestion conditions. At the
pH of the intestine, and in other neutral parts of the body, Fe3+ ions form large
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

IRON, FERRITIN, AND NUTRITION 329

insoluble complexes of iron and oxygen, whereas most of the other ions, including
ferrous, are stable as solitary ions surrounded by water. Absorption of the larger
and/or complexed forms of Fe3+ ions in food appears to depend on recognition
by the gut cells and, based on recognition genes in bacteria (5, 6, 16, 105), may
be specific for each Fe3+ complex. Fe2+, Ca, Zn, and Mn ions are recognized
by the gut receptor DMT1, which also functions in other tissues (4, 30, 37). The
iron-lactoferrin complex appears to be recognized by a specific gut cell molecule
for absorption, at least at certain stages of development (83), and may have a
more restricted role in iron trafficking than does DMT1. Heme is a stable iron
complex readily recognized by gut cells, based on absorption properties, but the
gene product that recognizes heme is yet to be identified (66). Finally, Fe3+ phytate
is an Fe3+ complex for which there may not be a gut recognition molecule because
it is poorly absorbed (26, 39), except in the monophytate form (56). The recent
studies on iron availability of ferritin iron in both animals and humans (10, 67,
68, 79), (Table 1 and Figure 2), the absorption of pure ferritin by humans, and the
resistance of ferritin to digestion under conditions similar to those during human
digestion (S.L. Kelleher, E.C. Theil, & B. Lonnerdal, unpublished observations)
indicate the possible existence of yet another gene product required for recognition
and absorption of intact ferritin. Support for the idea that intact ferritin can enter
enterocytes comes from the fact that viruses can cross membranes (101).

TABLE 1 Comparison of soy iron availability in humans in three studies

1973a 1984a 2003a

Number of subjects 5 10 18
Gender Female Male Female
Labeling Intrinsic Extrinsic Intrinsic
Food iron form Bean protein (ferritin) FeCl3 Bean protein (ferritin)b
Test iron form FeSO4 FeSO4 FeSO4
Form of meal Biscuit Soup Soup/muffin
% Uptake (soy meal) 19.8 1.6 25.9
% Uptake (reference 72.8 16 58.9
dose)
% Uptake 27.5 10.0 44.0
(meal/reference dose)
Hctc — 0.47 0.40
Hb (g/L)c 108 — 131
c
Serum ferritin (µg/L) — 61 11.2
a
References for data are: 1973 (79); 1984 (59); and 2003 (68), which is also the source of the table.
b
About 50% of the iron in the soybeans was in the ferritin form.
c
Values on day of soybean meal consumption.
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

330 THEIL

Figure 2 Different forms of dietary iron, at equal Fe content (mg/kg food), are
equally effective in correcting iron-deficiency anemia. Fe-deficient rats, 28 days after
Fe repletion with diets containing equal amounts of iron in different chemical forms:
ferrous sulfate (Fe2+SO4), iron in the Fe3+ mineral of horse spleen ferritin (HSF), and
soybean (SBM). The majority of the iron in SBM is the Fe3+ in the ferritin iron mineral.
Data are from Reference 10.

At the pH of the intestine, Fe3+ and Fe2+ ions have very different chemistry
and effective sizes because of the properties of the water molecules bound or
coordinated to each ion. The size and chemical differences emphasize the need
for different molecular pathways in absorption. (See the next paragraph for a
discussion of the hypothesis that all absorbed nonheme iron is reduced to Fe2+.)
For Fe2+ ions, the coordinated water has a pKa ∼9, similar to that of other metal
ions such as Cu2+, Zn2+, Mn2+, and will be in the form of a small ion surrounded
by four to six water molecules. The water coordinated to the Fe3+ ion has a pKa
∼3, and at neutral pH, will behave as a weak acid, but still stronger than acetic
acid. When the protons dissociate from the waters bound to Fe3+ ions, a reactive
species forms that produces complexes with multiple iron atoms linked by oxygen
bridges (“rust”).
Reduction of Fe3+ to Fe2+ is often proposed as a solution to the problems of
nutrition created by Fe3+ chemistry. The recent discovery of a ferrireductase related
to gut iron absorption (32, 63, 66) supports such an idea. However, other solutions
to the chemical problems of iron nutrition have evolved in microorganisms for
different Fe3+ complexes such as Fe(III)-citrate and heme (5, 6, 46, 82, 105). The
“parsimony of nature” suggests that in humans the multiple receptors evolved
for selective absorption of different forms of iron in microorganisms is conserved.
Support for such an idea comes from the recent identification of three nonheme iron
uptake proteins expressed by cells on the internal side of the gut: DMT1, transferrin
receptor 1, and transferrin receptor 2 (4, 31, 37, 48). In addition to evolutionary
arguments, and cell biology, the hypothesis that reduction of iron in the gut from
Fe3+ to Fe2+ is the sole mechanism for iron absorption has other drawbacks, which
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

IRON, FERRITIN, AND NUTRITION 331

include substrate variability, and electron transfer that can increase free radical
production. An example is the use of ascorbate to convert Fe3+ to Fe2+ in foods.
When Fe3+ is reduced to Fe2+ by ascorbate, the mixture of oxidized ascorbate,
Fe2+, and oxygen or oxidants produces a chain of free radicals, altered enterocyte
function, and gut irritability (24, 44).

Extrinsic and Intrinsic Labeling in Studies of Dietary Iron

Bioavailability
The absorption of added iron will depend on the amount of absorbable iron in
the diet, a value that can vary considerably. In the case of soybeans, for example,
the amount of phytate and the distribution between ferritin and phytate can vary;
compare References 3, 19, and 68. Moreover, variations in the phytate-to-iron ratio
will also affect absorbability because the absorbability of iron in monoferric phytate
is much higher than in the other Fe3+ phytates (56).Results of iron bioavailability
studies with the same foods that appear to conflict with each other may reflect
differences in the distribution of iron in the food among the different forms: Fe2+,
absorbable Fe3+ salts, heme, ferritin and nonabsorbable Fe3+ salts.

EXTRINSIC Fe3+ or Fe2+ ions have been used as extrinsic labels. The addition of
iron salts to foods can have different outcomes depending on the food composition
and the form of the iron added, such as Fe3+ citrate, Fe3+ EDTA, Fe3+(Cl− 3 ), or
Fe2+ sulfate/gluconate.
Iron forms used as labels include Fe3+(Cl−)3 and Fe3+(citrate)−3, a commer-
cially available 1:1 salt, which will dissolve in solutions or in the stomach, with
some of the six Fe3+ binding sites available to form new complexes with chelators
in foods or to form “rust” at the pH of the intestine.
Agents in foods, such as phytates or oxalates, are known to influence iron ab-
sorption (8). When the phytate-to-iron ratio is high during digestion, monophytate
complexes can form, which have a higher absorption based on studies in dogs (56).
Food composition, which will depend on growth conditions of plants, therefore
can influence the outcome of dietary studies with extrinsic labels (59, 79). Iron
availability can also be influenced by mixing foods with varying amounts and
forms of phytate and iron (19).
Fe(III)-EDTA and Fe(III)-citrate are nonheme iron complexes in which all Fe3+
ion binding sites are stably filled by the complexing agent. The efficacy of Fe-EDTA
supplements, even though the complex is an unnatural one, is well established (13,
29, 102), and relates to the stability of the iron complex. In contrast, Fe(III)-citrate,
a more stable iron complex, has been less studied as a dietary source, even though
bacteria have specific receptor recognition and uptake systems for Fe(III)-citrate
(16, 17, 29, 105), and mechanisms for absorption of Fe(III)-citrate may have been
conserved during evolution.
Fe2+ salts, in solution at the pH of outside the gut enterocyte, react with oxygen
and initiate radical chain reactions. Unless rapidly incorporated into cells, via
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

332 THEIL

DMT1 (64, 66) and/or into stable complexes, the reaction products of free Fe2+
ions and oxidants can be toxic (24, 44). Nevertheless, Fe2+ sulfate will continue
to remain a reasonable, if not optimal, standard for absorption because of past use,
until other, more selective standards are developed.

INTRINSIC Studies of iron absorption using intrinsic labels have led to results
that appear to conflict (52, 59, 68, 79, 99). In some seeds such as cereals, the
iron content is relatively low; ferritin is low, and most of the iron is complexed
in phytates or polyphenols (62). On the other hand, in seeds such as soybeans, a
major portion of the seed iron is a solid mineral inside the ferritin protein (3, 20),
which isolates the iron from phytates. In addition, plant ferritin is inside of plastids
(81), which further separates ferritin and iron from phytate. The higher distribution
of ferritin in the hulls (2) may provide a further barrier between ferritin iron and
phytic acid. In the case of ferritin uptake, it needs to be remembered that complexes
as large as viruses can cross cell membranes (101), although the mechanisms are
little understood. Similarly, the mechanism by which ferritin enters the cells in the
gut is not understood at this time, and is a subject for future exploration. Variations
in seed development and composition also appear to influence the distribution of
iron within the seed between phytate and ferritin.
A number of variables are known to influence the iron availability in diets with
added ferritin that was intrinsically labeled (see discussions in References 10, 92).
In the case of pure animal ferritin fed to animals, inflammation, often used to
induce increased ferritin accumulation (52), can produce a type of ferritin with the
iron “locked in” (65). In the case of soybeans, the distribution of labeled iron in the
seed is influenced by the time during plant development when the label is added
(20, 99), and whether or not the plant has formed iron-rich nodules that donate
iron to the seed (20).
The amount of seed phytate in soybeans does not appear to alter the availabil-
ity of endogenous soy iron Fe in ferritin (68, 99). However, the amount of seed
phytate can influence the availability of extrinsic iron added to foods from such
seeds (see the discussion of extrinsic labeling above and Reference 56). Moreover,
developmental changes in the seed can affect the form of the intrinsically labeled
iron in the seed (99). A possible reason for the developmental effect on intrinsic
labeling is the fact that iron continues to accumulate in soybeans (20) as the seed
ripens. Adding labeled iron in a single dose early in plant development and using
nodulating plants appears to produce soybeans with the highest percent of the
labeled iron in ferritin (70% to 90%) (2, 20, 68).

FERRITIN IN NATURE
Ferritin is a very large, unique, and conserved protein encoded in genomes from
Archea to humans and expressed in most cell types. Apparently, the ferritin struc-
tural motif is nature’s only solution to the problem of concentrating iron. Deletion
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

IRON, FERRITIN, AND NUTRITION 333

of a ferritin gene is embryonic lethal in mice (28). The conserved features of the
ferritin protein reside in quaternary and secondary structure: a spherical protein
cage around the solid iron mineral “core” (or around solvent in empty ferritin
protein), assembled from 12 or 24 polypeptides, folded into four helix bundles.
Primary sequence and tertiary structure vary, depending on the main function of
ferritin (concentrating iron or detoxifying hydrogen peroxide or oxygen). In con-
trast to most metalloproteins, where the metal is bound to a specific site on the
protein, in ferritin, iron is a substrate for the catalytic oxidase sites, binding two
iron atoms together to form the mineral (see References 23 and 90 for reviews), in
a transient state involving only a few (<50) iron atoms. Most of the iron in ferritin
is Fe3+ (usually thousands of iron atoms) and is not bound to the ferritin protein
itself, but is in the solid mineral bound through oxygen atoms to other Fe3+ atoms.

Animals
Ferritin is mainly a cytoplasmic protein in animals. In some cells, the large amounts
of iron used by ferrochelatase to make heme appear to be provided by a mito-
chondrial ferritin (54) encoded in the nuclear genome and transported into the
mitochondria. Ferritin gene expression is tightly coupled to iron status in cells and
to cell differentiation. Transcription regulation controls both the total amount of
ferritin mRNA and the relative amount of the two types of ferritin mRNA (H and
L), which are varied at the time of cell differentiation (95).
Translational regulation is a second major level of control for ferritin expression
that is superimposed on the transcriptional regulation. The effect of iron status, for
example, is so dramatic that translational regulation was discovered long before
the genes were cloned, and made ferritin mRNA an early model for translational
mRNA regulation in higher animals. In the noncoding region of ferritin mRNA,
near the cap structure, is a stem-loop structure folded into such a specific 3D shape
that only two proteins bind, IRP1 and IRP2, to repress ribosome binding. The IRPs,
coincidentally, are aconitase homologues and the amounts can vary depending on
the cell type, iron status, oxygen status, etc. Similarly, IRE structures, but with
specific mRNA features and IRP2 binding properties, occur in a number of other
mRNAs involved in iron homeostasis to create a combinatorial array of interactions
(93) that coordinately regulate synthesis of the proteins encoded in IRE-mRNAs
(21, 27, 40, 77, 80, 89, 93). The mRNA-specific differences in IRE structure
and IRP2 binding lead to a range or hierarchy of responses to iron or to anoxia
(93).
Cellular mechanisms for retrieving the iron concentrated and stored in ferritin
is currently the subject of active discussion and investigation. Two current hy-
potheses, which may function in parallel or selectively in different cell types are:
(a) protein degradation after transport to lysosomes (34, 71), which requires iden-
tification of mineral-dissolving mechanisms and iron carriers; and/or (b) protein
unfolding at the evolutionary conserved ferritin pores, which are exquisitely sen-
sitive to changes in solvent, temperature, and amino acid substitution, to increase
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

334 THEIL

access of reductants and chelators to the mineral (43, 57, 85), which requires
identification of the physiological protein unfolders/chelators. Both hypotheses
could occur in parallel, in the same cell, if ferritin in lysosomes is damaged protein
with the iron mineral turning over, and ferritin in the cytoplasm is undamaged pro-
tein with gated pores that control access of reductant and intracellular transporters
to the iron mineral.

Plants
Iron in plants needs to be concentrated in cells for the same functions as in animals.
Ferritin is found in all plant cells at some time during cell development (18), usually
in plastids (chloroplasts in leaves, amyloplasts in tubers and seeds, etc.) (81, 94).
Encoded in the nuclear genome, as in animals, ferritin in plants is targeted to
plastids by an extension peptide, N-terminal to the sequences common to animal
and plant ferritins (72). The ferritin iron-phosphate mineral characteristic of plants
appears to form in plastids after protein transport to the plastid (97). How the iron
in ferritin is retrieved from plastid ferritin for heme synthesis, and synthesized in
cell compartments other than the plastid, is not known. However, in contrast to
animals, where the main use of iron is in heme, photosynthesis in green plants
requires large amounts of nonheme iron in a type of reverse Krebs cycle (33). The
need for plastid iron appears to be dominant over cytoplasmic need in plants and
has influenced regulation and gene structure profoundly, when compared to animal
ferritin genes and regulation (18, 49, 53, 70, 72, 98).
Plant ferritin mRNA is regulated by iron during transcription. There is no IRE
structure in plant ferritin mRNA. In fact, when the animal ferritin mRNA IRE
is linked to plant ferritin mRNA in a chimera, IRE function, even in animal cell
extracts, is inhibited (49). The iron responsive element that regulates transcription
of ferritin mRNA in soybeans, for example, is a bipartite promoter that binds a
protein trans receptor factor to repress transcription when iron is at low levels in
the cell (98), and has no sequence homology to other promoters known at this
time. In plant ferritin genes, the intron number is twice that in animals and appears
to be unrelated to the structure of the exon-encoded proteins (70), suggesting
that in plants regulation of ferritin transcription, perhaps coordinated to plastid
development, dominated the evolution of gene structure.
In soybeans, and possibly in other legumes, the majority of the iron is in ferritin
(3, 20). The high ferritin and iron content of legumes depends, at least in part, on
the large amounts of iron used for nitrogen fixation by the nodules (12, 45, 50, 86,
87). Accumulation of nodule ferritin is developmentally regulated (69, 73). Iron
is recovered by nodule ferritin during nodule senescence and the iron is recycled
to the seed (20) by a mechanism yet to be identified.

Microorganisms
Most microorganisms, archea, algae, protozoa, bacteria, and fungi have ferritin.
Saccharomyces cerevesiae, which has lost many genes (7), has no recognizable
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

IRON, FERRITIN, AND NUTRITION 335

ferritin gene in available genome sequences. In such an exceptional circumstance,

an organism could concentrate and store iron in acidic vacuoles.
The function of ferritin in microorganisms appears to be more variable than in
higher plants and animals. For example, Escherichia coli has three different ferritin
genes (Bfr, FtnA, and dps) (1, 36, 41). FtnA appears to be important in trapping iron
from protein degradation in cells of stationary cultures (1), whereas FtnA and Bfr
together protect cells from oxygen toxicity (96). Bfr, a heme containing ferritin,
also appears to modulate catalase (katA) and peroxide resistance (60). The ferritin
active sites in bacterial ferritins vary from those in ferritins of higher plants and
animals. Differences include a variety of oxy substrates (oxygen, hydrogen perox-
ide) and several different oxygen products (hydrogen peroxide/water). Anaerobes
have ferritins as well (76). Sorting out the various reaction pathways of oxygen in
ferritin reactions is an active area of research (e.g., 42, 55, 103, 104).
DNA in bacteria is protected from hydrogen peroxide damage by dps(DNA
protection in starvation) proteins, which are mini-ferritins with the characteristic
quaternary and secondary structure constructed with 12 rather than 24 polypeptide
subunits (6, 15, 36, 41). Based on crystal structures, the active site (ferroxidase site)
in dps, which oxidizes iron and reduces an oxy substrate, is in a different part of the
molecule than in the ferritin of plants and animals (41), and may relate to whether
hydrogen peroxide is a substrate or a product (15, 42, 104). Dps represents the use
of ferritin by the pathogen to defeat the release of toxic oxygen species produced
by the host, while the host uses ferritin to decrease pathogen access to iron. The
amino acid sequence of dps proteins is very different from the other ferritins, but
the subunit folding and protein cage around a solvent or mineral-filled core are
conserved.

DIETARY FERRITIN
Iron in food ferritin is a “slow-release” form, a solid mineral inside the hollow
protein. In foods, the ferritin iron is chemically and physically different from
the inorganic or organic iron salts or complexes often used as iron supplements,
and different from the natural chelated forms of iron such as Fe(III)-phytate. In
ferritin then, the protein cage protects the mineralized iron from complexing food
agents such as phytates, oxalates, and tannins. Ferritin protein, which is resistant
to proteolysis (25) and further protected in seeds by concentration in the hull, is
likely to survive digestion in the stomach. Because the bioavailability of ferritin
iron and FeSO4 are similar (11, 67), and the DMT1 mechanism for iron uptake
from FeSO4 is restricted to single, divalent cations (reviewed in Reference 66),
novel mechanisms, yet to be identified, are indicated for ferritin interactions with
gut cells.
Currently, the only human genes known to participate in gut iron absorption
are DMT1 and Dcytb, which transport Fe2+ ions into the enterocyte from the gut,
or reduce iron in the gut lumen to Fe2+, respectively (63, 64). In addition, the
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

336 THEIL

lactoferrin receptor will contribute to iron absorption at certain stages of life (58,
83). The small number of genes identified for iron absorption in humans contrasts
with the more than 20 genes known to participate in iron absorption in bacteria
(5, 6, 16, 17), and as they are specific for different iron forms such as heme, Fe2+,
Fe(III)-citrate, FE(III)-enterobactin, transferrin, etc., it is entirely possible that a
gut enterocyte receptor for ferritin occurs in humans.

Absorption Studies in Rats

The use of labeled iron in ferritin or soybean absorption studies in the past has led
to varying results (e.g., 52, 59, 79). However, recent studies emphasize that nature
uses ferritin at critical periods of human, animal, plant, and microbial development
(5, 6, 18, 88, 94), which suggests it might be a good dietary source. To reexamine
the availability of ferritin and soy iron, diets with unlabelled, endogenous iron
in different chemical forms were given to iron-deficient rats. The response of the
erythron (hematocrit) was determined for iron-deficient rats given iron-replete diets
where the amount of iron in each diet was matched, but the chemical form of the
iron was varied: Fe2+SO−2 4 or horse spleen ferritin or soybean seeds (∼80% iron in
ferritin). Iron-deficient rats were divided into three groups and each group received
one of the diets for 14 days. In all three groups the hematocrits had returned to
normal, indicating that in iron deficiency, dietary Fe2+SO−2 4 , pure ferritin, and
soybean (ferritin) iron were equally able to provide iron to the erythron (10, 92).
The availability of iron in rice with increased ferritin, introduced as the soybean
ferritin transgene targeted to the seed (35), has also been demonstrated in iron-
deficient rats. The erythron of iron-deficient rats used the iron in rice transgenic for
soybean ferritin and soybeans with comparable efficiency (10, 67), which indicates
the potential to solving problems of iron nutrition by modifying the ferritin and
iron composition of seeds through engineering or breeding.

ABSORPTION STUDIES IN HUMANS Iron absorption studies using unlabeled fer-

ritin aren’t feasible in humans, as they were in rats (9), because of the large amount
of ferritin required. In a recent study, intrinsically labeled soybeans were used (68).
The combination of three features distinguished the study from earlier studies
(Table 1) in which conflicting results were obtained about the availability in hu-
mans of iron from soybeans (59, 79): (a) subjects were borderline iron deficient
to increase overall iron uptake and increase the sensitivity of the measurements;
(b) dietary uniformity was maximized through the use of a clinical nutrition re-
search center; and (c) developmental studies on iron incorporation and distribution
during growth of the soybean plant (20) were used as a guide for intrinsic labeling
of the iron in soybeans.
Soybean iron, measured as red blood cell 55Fe (68), was efficiently absorbed
by the American women (Table 1). In the women who had blood values for
hemoglobin 131 g/L, hematocrits 40.0%, and serum ferritin mean 7.4 µg/L, ab-
sorption was 25.9% (68). The results confirm those obtained with a group of Indian
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

IRON, FERRITIN, AND NUTRITION 337

and African women with low hemoglobin (10.8 g/100 ml), analyzed 30 years ago,
where absorption was 19.8% of the 55Fe in intrinsically labeled soy in biscuits and
21.2% of 59Fe, added as ferric ammonium citrate (a nonstoichiometric complex of
Fe3+, NH4, and citric acid) in the cooking water (79). Differences with subsequent
experiments such as those in American men (59) may relate to iron status, gender,
and the form of the extrinsic label [Fe3+(Cl−)3], for which the solution chemistry
will lead to some trapping either by phytate or as “rust”.
Direct studies on ferritin iron availability in humans, using purified ferritin from
which iron was removed and replaced in vitro by 59Fe, under conditions known
to give normal iron mineral (75), again showed availabilities of 21% to 27%,
indistinguishable from Fe2+SO−2 4 , for both red blood cell incorporation and whole
body retention. Ferritin was given in apple juice as part of a normal breakfast
(P. Davila-Hicks, S.L. Kelleher, E.C. Theil, & B. Lonnerdal, submitted). In foods
such as legumes the majority of iron is in ferritin, with a significant amount in the
hulls (2) that will provide additional protection to digestion. Combined with the
stability of the isolated protein to denaturation and proteolytic digestion (25, 51),
soybean ferritin is likely to survive digestion intact and to arrive at the intestinal
villi as the protein-coated mineral. The known mechanisms for iron uptake during
digestion have lagged behind those for iron by cells inside the body. Even for the
more intensely studied case of cells in the body, iron, a number of new mechanisms
of iron uptake have been revealed in the last few years, such as a second transferring
receptor and the use of the enterocyte DMT1 protein in cells such as the liver.
Endogenous ferritin is a natural alternative dietary iron source for humans that has
been underutilized in supplementation studies and plans. Questions for the future
include: How much ferritin is in nonsoy food legumes? How available is iron in the
ferritin of nonsoy, food legumes? Can ferritin-rich foods be consumed in quantities
sufficient to minimize dietary iron deficiency? Are there enough variations in the
types of ferritin-rich foods with high acceptability to have a major impact on global
iron deficiency?

PERSPECTIVE
Treatment of iron deficiency anemia with iron supplements has been known for
centuries, yet the problem of iron deficits and anemia still afflicts 30% of the world’s
population. Acceptance of known supplements is a significant issue. Problems of
eradicating nutritional iron deficits include (a) acceptance of known supplements;
(b) confusion when similar experimental designs to test potential iron supple-
ments have apparently conflicting outcomes; and (c) knowledge of the complexity
of molecular mechanisms of iron uptake in humans lags behind that in other or-
ganisms, although the rate of change is promising.
To solve the problem of dietary iron deficits more knowledge is required of iron
absorption mechanisms, the contributions of genetic ethnicities, and the contribu-
tions of gender to iron absorption. In addition, expansion and use of knowledge
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

338 THEIL

about the chemistry and biochemistry of different nonheme iron forms in foods
(chemistry of different iron complexes, natural variability of ratios of iron, com-
plexing agents in plant foods, and alterations during digestion and absorption) are
crucial. Alternative iron supplements and diets will need to be developed to solve
the problem of dietary iron deficiency. A novel, alternative dietary iron source, with
an enormous potential contribution to the eradication of global iron deficiency in
the twenty-first century is ferritin.

ACKNOWLEDGMENT
The support of the NIH (NIDDK and NHLBI) and the USDA-North Carolina Agri-
cultural Research Service; the enthusiastic participation of my graduate students,
research assistants, and postdoctoral fellows; and the insights of my colleagues,
particularly John Beard, Joe Burton, and Bo Lonnerdal, about my work on ferritin
and nutrition are warmly and gratefully acknowledged.

The Annual Review of Nutrition is online at http://nutr.annualreviews.org

LITERATURE CITED
1. Abdul-Tehrani H, Hudson AJ, Chang YS, 9. Beard JL. 2000. Effectiveness and strate-
Timms AR, Hawkins C, et al. 1999. Fer- gies of iron supplementation during preg-
ritin mutants of Escherichia coli are iron nancy. Am. J. Clin. Nutr. 71:1288–94S
deficient and growth impaired, and fur 10. Beard JL, Burton JW, Theil EC. 1996. Pu-
mutants are iron deficient. J. Bacteriol. rified ferritin and soybean meal can be
181:1415–28 sources of iron for treating iron deficiency
2. Ambe S. 1994. Mossbauer study of iron in rats. J. Nutr. 126:154–60
in soybean hulls and cotyledons. J. Agric. 11. Beard JL, Connor JR. 2003. Iron status
Food Chem. 42:262–67 and neural functioning. Annu. Rev. Nutr.
3. Ambe S, Ambe F, Nozuki T. 1987. Moss- 23:41–58
bauer study of iron in soybean seeds. J. 12. Bergersen FJ. 1963. Iron in the developing
Agric. Food Chem. 35:292–96 soybean nodule. Aust. J. Biol. Sci. 16:916–
4. Andrews NC. 2002. A genetic view of iron 19
homeostasis. Semin. Hematol. 39:227–34 13. Bothwell TH. 1999. Iron fortification with
5. Andrews SC. 1998. Iron storage in bacte- special reference to the role of iron EDTA.
ria. Adv. Microb. Physiol. 40:281–351 Arch. Latinoam. Nutr. 49:23–33S
6. Andrews SC, Robinson AK, Rodriguez- 14. Bothwell TH. 2000. Iron requirements in
Quinones F. 2003. Bacterial iron home- pregnancy and strategies to meet them.
ostasis. FEMS Microbiol. Rev. 27:215–37 Am. J. Clin. Nutr. 72:257–64S
7. Aravind L, Watanabe H, Lipman DJ, 15. Bozzi M, Mignogna G, Stefanini S, Barra
Koonin EV. 2000. Lineage-specific loss D, Longhi C, et al. 1997. A novel non-
and divergence of functionally linked heme iron-binding ferritin related to the
genes in eukaryotes. Proc. Natl. Acad. Sci. DNA-binding proteins of the Dps fam-
USA 97:11319–24 ily in Listeria innocua. J. Biol. Chem.
8. Baynes RD, Bothwell TH. 1990. Iron de- 272:3259–65
ficiency. Ann. Rev. Nutr. 10:133–48 16. Braun V, Braun M. 2002. Active transport
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

IRON, FERRITIN, AND NUTRITION 339

of iron and siderophore antibiotics. Curr. 28. Ferreira F, Bucchini D, Martin ME, Levi
Opin. Microbiol. 5:194–201 S, Arosio P, et al. 2000. Early embry-
17. Braun V, Hantke K, Koster W. 1998. Bac- onic lethality of H ferritin gene deletion
terial iron transport: mechanisms, genet- in mice. J. Biol. Chem. 275:3021–24
ics and regulation. In Iron Transport and 29. Fidler MC, Davidsson L, Walczyk T, Hur-
Storage in Microorganisms, Plants, and rell RF. 2003. Iron absorption from fish
Animals, ed. A Sigel, H Sigel, pp. 67–146. sauce and soy sauce fortified with sodium
New York: Marcel Dekker iron EDTA. Am. J. Clin. Nutr. 78:274–
18. Briat J-F, Lobreaux S. 1998. Iron storage 78
and ferritin in plants. Met. Ions Biol. Syst. 30. Fleming MD, Trenor CCI, SU MA, Foern-
35:563–83 zler D, Beier DR, et al. 1997. Microcytic
19. Brunvand L, Henriksen C, Larsson M, anaemia mice have a mutation in Nramp2,
Sandberg A-S. 1995. Iron deficiency a candidate iron transporter gene. Nat.
among pregnant Pakistanis in Norway and Genet. 16:383–86
the content of phytic acid in their diet. Acta 31. Fleming RF, Migas M, Holden CC, Wa-
Obstet. Gynecol. Scand. 74:520–25 heed A, Britton RS, et al. 2000. Trans-
20. Burton JW, Harlow C, Theil EC. 1998. Ev- ferrin receptor 2: continued expression in
idence for reutilization of nodule iron in mouse liver in the face of iron overload
soybean seed development. J. Plant Nutr. and in hereditary hemochromatosis. Proc.
21:913–27 Natl. Acad. Sci. USA 97:2214–19
21. Cairo G, Pietrangelo A. 2000. Iron regu- 32. Frazer DM, Wilkins SJ, Becker EM, Mur-
latory proteins in pathobiology. Biochem. phy TL, Vulpe CD, et al. 2003. A rapid
J. 352:241–50 decrease in the expression of DMT1 and
22. Casanueva E, Viteri FE. 2003. Iron and Dcytb but not Ireg1 or hephaestin explains
oxidative stress in pregnancy. J. Nutr. 133: the mucosal block phenomenon of iron
1700–8S absorption. Gut 52:340–46
23. Chasteen ND, Harrison PM. 1999. Miner- 33. Fromme P, Jordan P, Krauss N. 2001.
alization in ferritin: an efficient means of Structure of photosystem I. Biochim. Bio-
iron storage. J. Struct. Biol. 126(3):182– phys. Acta 1507:5–31
94 34. Garner B, Roberg K, Brunk UT. 1998. En-
24. Courtois F, Suc I, Garofalo C, Ledoux M, dogenous ferritin protects cells with iron-
Seidman E, Levy E. 2000. Iron-ascorbate laden lysosomes against oxidative stress.
alters the efficiency of Caco-2 cells to Free Radic. Res. 29:103–14
assemble and secrete lipoproteins. Am. 35. Goto F, Yoshihara T, Shigemoto N, Toki
J. Physiol. Gastrointest. Liver Physiol. S, Takaiwa F. 1999. Iron fortification of
279:G12–19 rice seed by the soybean ferritin gene. Nat.
25. Crichton RR. 1973. Ferritin. Structure Biotechnol. 17:282–86
Bonding 17:67–134 36. Grant RA, Filman DJ, Finkel SE, Kolter
26. Davidsson L, Galan P, Kastenmayer P, Ch- R, Hogle JM. 1998. The crystal structure
erouvrier F, Juillerat M-A, et al. 1994. Iron of Dps, a ferritin homolog that binds and
bioavailability studied in infants: the in- protects DNA. Nat. Struct. Biol. 5:294–
fluence of phytic acid and ascorbic acid 303
in infant formulas based on soy isolate. 37. Gunshin H, Mackenzie B, Berger UV,
Pediatr. Res. 36:816–22 Gunshin Y, Romero MF, et al. 1997.
27. Eisenstein RS, Ross KL. 2003. Novel Cloning and characterization of a mam-
roles for iron regulatory proteins in the malian proton-coupled metal-ion trans-
adaptive response to iron deficiency. J. porter. Nature 388:482–88
Nutr. 133:1510–16S 38. Hallberg L. 2001. Perspectives on
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

340 THEIL

nutritional iron deficiency. Annu. Rev. PT, Kawano S, et al. 1999. Molecular
Nutr. 21:1–21 cloning of transferrin receptor 2: a new
39. Hallberg L, Rossander L, Skanberg A-B. member of the transferrin receptor-like
1987. Phytates and the inhibitory effect family. J. Biol. Chem. 274:20826–32
of bran on iron absorption in man. Am. J. 49. Kimata Y, Theil EC. 1994. Posttransla-
Clin. Nutr. 45:988–96 tional regulation of ferritin during nodule
40. Hentze MW, Kuhn LC. 1996. Molecu- development in soybean. Plant Physiol.
lar control of vertebrate iron metabolism: 104:263–70
mRNA-based circuits operated by iron, 50. Ko MP, Huang P-Y, Huang J-S, Barker
nitric oxide, and oxidative stress. Proc. KR. 1985. Accumulation of phytoferritin
Natl. Acad. Sci. USA 93:8175–82 and starch granules in developing nodules
41. Ilari A, Ceci P, Ferrari D, Rossi GL, of soybean roots infected with Hereodera
Chiancone E. 2002. Iron incorporation glycines. Phytopathology 75:159–64
into Escherichia coli Dps gives rise to a 51. Lavoie DJ, Ishikawa K, Listowsky I.
ferritin-like microcrystalline core. J. Biol. 1978. Correlations between subunit distri-
Chem. 277:37619–23 bution, microheterogeneity, and iron con-
42. Jameson GNL, Jin W, Krebs C, Per- tent of human liver ferritin. Biochemistry
reira AS, Tavares P, et al. 2002. Stoi- 17:5448–54
chiometric production of hydrogen perox- 52. Layrisse M, Martinez-Torres C, Renzy M,
ide and parallel formation of ferric multi- Leets I. 1975. Ferritin iron absorption in
mers through decay of the diferric-peroxo man. Blood 45:689–98
complex, the first detectable intermedi- 53. Lescure AM, Proudhon D, Pesey H,
ate in ferritin mineralization. Biochem- Ragland M, Theil EC, Briat JF. 1991. Fer-
istry 41:13435–43 ritin gene transcription is regulated by iron
43. Jin W, Takagi H, Pancorbo NM, Theil EC. in soybean cell cultures. Proc. Natl. Acad.
2001. “Opening” the ferritin pore for iron Sci. USA 88:8222–26
release by mutation of conserved amino 54. Levi S, Corsi B, Bosisio M, Invernizzi
acids at inter-helix and loop sites. Bio- R, Volz A, et al. 2001. A human mito-
chemistry 40:7525–32 chondrial ferritin encoded by an intronless
44. Jourdheuil D, Meddings JB. 2001. Ox- gene. J. Biol. Chem. 276:24437–40
idative and drug-induced alterations in 55. Lindsay S, Brosnahan D, Lowery TJ Jr,
brush border membrane hemileaflet flu- Crawford K, Watt GD. 2003. Kinetic stud-
idity, functional consequences for glu- ies of iron deposition in horse spleen fer-
cose transport. Biochim. Biophys. Acta ritin using O2 as oxidant. Biochim. Bio-
1510:342–53 phys. Acta 1621:57–66
45. Kaiser BN, Moreau S, Castelli J, Thom- 56. Lipschitz DA, Simpson KM, Cook JD,
son R, Lambert A, et al. 2003. The soy- Morris ER. 1979. Absorption of mono-
bean NRAMP homologue, GmDMT1, is ferric phytate by dogs. J. Nutr. 109:1154–
a symbiotic divalent metal transporter ca- 60
pable of ferrous iron transport. Plant J. 57. Liu X, Jin W, Theil EC. 2003. Opening
35:295–304 protein pores with chaotropes enhances
46. Kaplan J. 2002. Strategy and tactics in Fe reduction and chelation of Fe from the
the evolution of iron acquisition. Semin. ferritin biomineral. Proc. Natl. Acad. Sci.
Hematol. 39:216–26 USA 100:3653–58
47. Kapur D, Agarwal KN, Agarwal DK. 58. Lonnerdal B, Iyer S. 1995. Lactoferrin:
2002. Nutritional anemia and its control. molecular structure and biological func-
Indian J. Pediatr. 69:607–16 tion. Annu. Rev. Nutr. 15:93–110
48. Kawabata H, Yang R, Hirama T, Vuong 59. Lynch SR, Dassenko SA, Beard JL, Cook
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

IRON, FERRITIN, AND NUTRITION 341

JD. 1984. Iron absorption from legumes inoculated with Bradyrhizobium sp. New
in humans. Am. J. Clin. Nutr. 40:42–47 Phytol. 108:51–57
60. Ma JF, Ochsner UA, Klotz MG, 70. Proudhon D, Wei J, Briat J, Theil EC.
Nanayakkara VK, Howell ML, et al. 1996. Ferritin gene organization: differ-
1999. Bacterioferritin A modulates ences between plants and animals suggest
catalase A (KatA) activity and resistance possible kingdom-specific selective con-
to hydrogen peroxide in Pseudomonas straints. Mol. Evol. 42:325–36
aeruginosa. J. Bacteriol. 181:3730–42 71. Radisky DC, Kaplan J. 1998. Iron in cy-
61. Makrides M, Crowther CA, Gibson RA, tosolic ferritin can be recycled through
Gibson RS, Skeaff CM. 2003. Efficacy lysosomal degradation in human fibrob-
and tolerability of low-dose iron supple- lasts. Biochem. J. 336(Pt. 1):201–5
ments during pregnancy: a randomized 72. Ragland M, Briat J-F, Gagnon J, Laul-
controlled trial. Am. J. Clin. Nutr. 78:145– here J-P, Massenet O, Theil EC. 1990.
53 Evidence for conservation of ferritin se-
62. Matuschek E, Towo E, Svanberg U. 2001. quences among plants and animals and for
Oxidation of polyphenols in phytate- a transit peptide in soybean. J. Biol. Chem.
reduced high-tannin cereals: effect on dif- 265:18339–44
ferent phenolic groups and on in vitro 73. Ragland M, Theil EC. 1993. Ferritin and
accessible iron. J. Agric. Food Chem. iron are developmentally regulated in nod-
49:5630–38 ules. Plant Mol. Biol. 21:555–60
63. McKie AT, Barrow D, Latunde-Dada GO, 74. Ramakrishnan U, Frith-Terhune A,
Rolfs A, Sager G, et al. 2001. An iron- Cogswell M, Kettle Khan L. 2002.
regulated ferric reductase associated with Dietary intake does not account for
the absorption of dietary iron. Science differences in low iron stores among
291:1755–59 Mexican American and non-Hispanic
64. McKie AT, Marciani P, Rolfs A, Brennan white women: Third National Health and
K, Wehr K, et al. 2000. A novel duodenal Nutrition Examination Survey, 1988–
iron-regulated transporter, IREG1, impli- 1994. J. Nutr. 132:996–1001
cated in the basolateral transfer of iron to 75. Rohrer JS, Islam QT, Watt GD, Sayers
the circulation. Mol. Cell 5:299–309 DE, Theil EC. 1990. Iron environment
65. Mertz JR, Theil EC. 1983. Subunit dimers in ferritin with large amounts of phos-
in sheep spleen apoferritin: the effect on phate, from Azotobacter vinelandii and
iron storage. J. Biol. Chem. 258:11719–26 horse spleen, analyzed using extended x-
66. Miret S, Simpson RJ, McKie AT. 2003. ray absorption fine structure (EXAFS).
Physiology and molecular biology of di- Biochemistry 29:259–64
etary iron absorption. Annu. Rev. Nutr. 76. Romao CV, Regalla M, Xavier AV, Teix-
23:283–301 eira M, Liu MY, Le Gall J. 2000. A bacte-
67. Murray-Kolb LE, Theil EC, Takaiwa F, rioferritin from the strict anaerobe Desul-
Goto F, Yoshihara T, Beard JL. 2002. fovibrio desulfuricans ATCC 27774.
Transgenic rice as a source of iron for iron- Biochemistry 39:6841–49
depleted rats. J. Nutr. 132:957–60 77. Rouault TA, Klausner RD. 1996. Post-
68. Murray-Kolb LE, Welch R, Theil EC, transcriptional regulation of genes of iron
Beard JL. 2003. Women with low iron metabolism in mammalian cells. J. Biol.
stores absorb iron from soybeans. Am. J. Inorg. Chem. 1:494–99
Clin. Nutr. 77:180–84 78. Savoie N, Rioux FM. 2002. Impact of ma-
69. O’Hara GW, Dilworth MJ, Boonkerd N, ternal anemia on the infant’s iron status at
Parkpian P. 1988. Iron-deficiency specifi- 9 months of age. Can. J. Public Health
cally limits nodule development in peanut 93:203–7
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

342 THEIL

79. Sayers M, Lynch S, Jacobs P, Charlton Metalloproteins, ed. A Messerschmidt,R

RW, Bothwell TH, et al. 1973. The ef- Huber, K Wieghardt, T Poulos, pp. 771–
fects of ascorbic acid supplementation on 81. Chichester, UK: Wiley
the absorption of iron in maize, wheat and 91. Theil EC. 2003. Ferritin: at the crossroads
soya. Br. J. Haematol. 24:209–18 of iron and oxygen metabolism. J. Nutr.
80. Schneider BD, Leibold EA. 2000. Reg- 133:1549–53S
ulation of mammalian iron homeosta- 92. Theil EC, Burton JW, Beard JL. 1997. A
sis. Curr. Opin. Clin. Nutr. Metab. Care sustainable solution for dietary iron de-
3:267–73 ficiency through plant biotechnology and
81. Seckback J. 1982. Ferreting out the secrets breeding to increase seed ferritin control.
of plant ferritin—a review. J. Plant Nutr. Eur. J. Clin. Nutr. 51:S28–31
5(4–7):369–94 93. Theil EC, Eisenstein RS. 2000. Combina-
82. Stojiljkovic I, Perkins-Balding D. 2002. torial mRNA regulation: iron regulatory
Processing of heme and heme-containing proteins and iso-iron responsive elements
proteins by bacteria. DNA Cell Biol. (iso-IREs). J. Biol. Chem. 275:40659–
21:281–95 62
83. Suzuki YA, Shin K, Lonnerdal B. 2001. 94. Theil EC, Hase T. 1993. Plant and micro-
Molecular cloning and functional expres- bial ferritins. In Iron Chelation in Plants
sion of a human intestinal lactoferrin re- and Soil Microorganisms, ed. LL Barton,
ceptor. Biochemistry 40:15771–79 BC Hemming, pp. 133–56. San Diego:
84. Swensen AR, Harnack LJ, Ross JA. Academic
2001. Nutritional assessment of preg- 95. Torti FM, Torti SV. 2002. Regulation of
nant women enrolled in the Special Sup- ferritin genes and protein. Blood 99:3505–
plemental Program for Women, Infants, 16
and Children (WIC). J. Am. Diet. Assoc. 96. Touati D, Jacques M, Tardat B, Bauchard
101:903–8 L, Despied S. 1995. Lethal oxidative dam-
85. Takagi H, Shi D, Hall Y, Allewell NM, age and mutagenesis are generated by iron
Theil EC. 1998. Localized unfolding at in fur mutants of Escherichia coli: pro-
the junction of three ferritin subunits. J. tective role of superoxide dismutase. J.
Biol. Chem. 273:18685–88 Bacteriol. 177:2305–14
86. Tang C, Robson AD, Dilworth MJ. 1990. 97. Waldo GS, Wright E, Wang Z-H, Briat
A split-root experiment shows that iron is J-F, Theil EC, Sayers DE. 1995. Forma-
required for nodule initiation in Lupinus tion of the ferritin iron mineral occurs in
angustifolius L. New Phytol. 115:61–67 plastids: an x-ray absorption spectroscopy
87. Tang C, Robson AD, Dilworth MJ, Kuo J. (EXAFS) study. Plant Physiol. 109:797–
1992. Microscopic evidence on how iron 802
deficiency limits nodule initiation in Lupi- 98. Wei J, Theil EC. 2000. Identification and
nus angustifoliusL. New Phytol. 121:457– characterization of the iron regulatory el-
67 ement in the ferritin gene of a plant (soy-
88. Theil EC. 1987. Ferritin: structure, gene bean). J. Biol. Chem. 275:17488–93
regulation, and cellular function in ani- 99. Welch RM, Van Campen R. 1975. Iron
mals, plants, and microorganisms. Annu. availability to rats from soybeans. J. Nutr.
Rev. Biochem. 56:289–315 105:253–56
89. Theil EC. 1990. Ferritin mRNA transla- 100. Wessling-Resnick M. 2000. Iron trans-
tion, structure, and gene transcription dur- port. Annu. Rev. Nutr. 20:129–51
ing development of animals and plants. 101. Whittaker GR, Kann M, Helenius A.
Enzyme 44:68–82 2000. Viral entry into the nucleus. Annu.
90. Theil EC. 2001. Ferritin. In Handbook of Rev. Cell Dev. Biol. 16:627–51
6 Feb 2004 18:10 AR AR216-NU24-15.tex AR216-NU24-15.sgm LaTeX2e(2002/01/18) P1: GCE
AR REVIEWS IN ADVANCE10.1146/annurev.nutr.24.012003.132212

IRON, FERRITIN, AND NUTRITION 343

102. World Health Organization. 2000. Eval- L, Laue TM, et al. 2002. Iron and hy-
uation of certain food additives and con- drogen peroxide detoxification proper-
taminants. WHO Tech. Rep. Ser. 896:1– ties of DNA-binding protein from starved
128 cells. A ferritin-like DNA-binding pro-
103. Yang X, Chiancone E, Stefanini S, Ilari tein of Escherichia coli. J. Biol. Chem.
A, Chasteen ND. 2000. Iron oxidation 277:27689–96
and hydrolysis reactions of a novel fer- 105. Zimmermann L, Hantke K, Braun V.
ritin from Listeria innocua. Biochem. J. 1984. Exogenous induction of the iron
349(Pt. 3):783–86 dicitrate transport system of Escherichia
104. Zhao G, Ceci P, Ilari A, Giangiacomo coli K-12. J. Bacteriol. 159:271–77
 Theil.qxd 2/6/2004 5:53 PM Page 1

IRON, FERRITIN, AND NUTRITION C-1

Figure 1 Ferritin pores opened/unfolded or closed/folded. Backbone structure (24

subunits) of a ferritin protein, looking down an axis at the junction of three subunits
that form the gated pore leading from the cavity where the iron mineral forms. There
are eight such pores per ferritin molecule. Photo reproduced from Reference 57.

View publication stats