The incorporation of iron into apoferritin as mediated by ceruloplasmin

Ceruloplasmin, a copper ferroxidase, promotes the incorporation of Fe(III) into the iron storage protein, apoferritin. The product formed is identical to ferritin as judged by polyacrylamide electrophoresis and iron/protein measurements. Of several proteins examined, only apoferritin accumulates the Fe(III) produced by ceruloplasmin. When ceruloplasmin was replaced by tyrosinase, which we have shown to have ferroxidase activity, no iron incorporation into apoferritin was observed. It is proposed that Fe(III) is transferred directly and specifically to apoferritin. These data support a more specific role for ceruloplasmin in iron metabolism than has previously been proposed.     

Role of phosphate in initial iron deposition in apoferritin

Ferritins from microorganisms to man are known to contain varying amounts of phosphate which has a pronounced effect on the structural and magnetic properties of their iron mineral cores. The present study was undertaken to gain insight into the role of phosphate in the early stages of iron accumulation by ferritin. The influence of phosphate on the initial deposition of iron in apoferritin (12 Fe/protein) was investigated by EPR, 57Fe M?ssbauer spectroscopy, and equilibrium dialysis. The results indicate that phosphate has a significant influence on iron deposition. The presence of 1 mM phosphate during reconstitution of ferritin from apoferritin, Fe(II), and O2 accelerates the rate of oxidation of the iron 2-fold at pH 7.5. In the presence or absence of phosphate, the rate of oxidation at 0 degrees C follows simple first-order kinetics with respect to Fe(II) with half-lives of 1.5 +/- 0.3 or 2.8 +/- 0.2 min, respectively, consistent with a single pathway for iron oxidation when low levels of iron are added to the apoprotein. This pathway may involve a protein ferroxidase site where phosphate may bind iron(II), shifting its redox potential to a more negative value and thus facilitating its oxidation. Following oxidation, an intermediate mononuclear Fe(III)-protein complex is formed which exhibits a transient EPR signal at g' = 4.3. Phosphate accelerates the rate of decay of the signal by a factor of 3-4, producing EPR-silent oligonuclear or polynuclear Fe(III) clusters. In 0.5 mM Pi, the signal decays according to a single phase first-order process with a half-life near 1 min.(ABSTRACT TRUNCATED AT 250 WORDS)     

The consequences of hydroxyl radical formation on the stoichiometry and kinetics of ferrous iron oxidation by human apoferritin

Despite previous detection of hydroxyl radical formation during iron deposition into ferritin, no reports exist in the literature concerning how it might affect ferritin function. In the present study, hydroxyl radical formation during Fe(II) oxidation by apoferritin was found to be contingent on the "ferroxidase" activity (i.e., H subunit composition) exhibited by apoferritin. Hydroxyl radical formation was found to affect both the stoichiometry and kinetics of Fe(II) oxidation by apoferritin. The stoichiometry of Fe(II) oxidation by apoferritin in an unbuffered solution of 50 mM NaCl, pH 7.0, was approximately 3.1 Fe(II)/O(2) at all iron-to-protein ratios tested. The addition of HEPES as an alternate reactant for the hydroxyl radical resulted in a stoichiometry of about 2 Fe(II)/O(2) at all iron-to-protein ratios. HEPES functioned to protect apoferritin from oxidative modification, for its omission from reaction mixtures containing Fe(II) and apoferritin resulted in alterations to the ferritin consistent with oxidative damage. The kinetic parameters for the reaction of recombinant human H apoferritin with Fe(II) in HEPES buffer (100 mM) were: K(m) = 60 microM, k(cat) = 10 s(-1), and k(cat)/K(m) = 1.7 x 10(5) M(-1) x (-1). Collectively, these results contradict the "crystal growth model" for iron deposition into ferritin and, while our data would seem to imply that the ferroxidase activity of ferritin is adequate in facilitating Fe(II) oxidation at all stages of iron deposition into ferritin, it is important to note that these data were obtained in vitro using nonphysiologic conditions. The possibility that these findings may have physiological significance is discussed.     

Catalysis of iron core formation in Pyrococcus furiosus ferritin

The hollow sphere-shaped 24-meric ferritin can store large amounts of iron as a ferrihydrite-like mineral core. In all subunits of homomeric ferritins and in catalytically active subunits of heteromeric ferritins a diiron binding site is found that is commonly addressed as the ferroxidase center (FC). The FC is involved in the catalytic Fe(II) oxidation by the protein; however, structural differences among different ferritins may be linked to different mechanisms of iron oxidation. Non-heme ferritins are generally believed to operate by the so-called substrate FC model in which the FC cycles by filling with Fe(II), oxidizing the iron, and donating labile Fe(III)–O–Fe(III) units to the cavity. In contrast, the heme-containing bacterial ferritin from Escherichia coli has been proposed to carry a stable FC that indirectly catalyzes Fe(II) oxidation by electron transfer from a core that oxidizes Fe(II). Here, we put forth yet another mechanism for the non-heme archaeal 24-meric ferritin from Pyrococcus furiosus in which a stable iron-containing FC acts as a catalytic center for the oxidation of Fe(II), which is subsequently transferred to a core that is not involved in Fe(II)-oxidation catalysis. The proposal is based on optical spectroscopy and steady-state kinetic measurements of iron oxidation and dioxygen consumption by apoferritin and by ferritin preloaded with different amounts of iron. Oxidation of the first 48 Fe(II) added to apoferritin is spectrally and kinetically different from subsequent iron oxidation and this is interpreted to reflect FC building followed by FC-catalyzed core formation.     

High-resolution X-ray structures of human apoferritin H-chain mutants correlated with their activity and metal-binding sites

Ferritins are a family of proteins distributed widely in nature. In bacterial, plant, and animal cells, ferritin appears to serve as a soluble, bioavailable, and non-toxic form of iron provider. Ferritins from animal sources are heteropolymers composed of two types of subunit, H and L, which differ mainly by the presence (H) or absence (L) of active ferroxidase centres. We report the crystallographic structures of four human H apoferritin variants at a resolution of up to 1.5 Angstrom. Crystal derivatives using Zn(II) as redox-stable alternative for Fe(II), allows us to characterize the different metal-binding sites. The ferroxidase centre, which is composed of sites A and B, binds metal with a preference for the A site. In addition, distinct Zn(II)-binding sites were found in the 3-fold axes, 4-fold axes and on the cavity surface near the ferroxidase centre. To study the importance of the distance of the two metal atoms in the ferroxidase centre, single and double replacement of glutamate 27 (site A) and glutamate 107 (site B), the two axial ligands, by aspartate residues have been carried out. The consequences for metal binding and the correlation with Fe(II) oxidation rates are discussed.     

Iron oxidation and transferrin formation by phosvitin.

The catalytic activity of phosvitin in Fe(II) oxidation and the addition of iron to transferrin were studied under various conditions. It was concluded that the Fe(II) oxidized by phosvitin would bind to apotransferrin, although an appreciable fraction of Fe(III) remained bound to phosvitin. Fe(III) also migrated from phosvitin to apotransferrin. This reaction was first-order with respect to Fe(III)-phosvitin concentration with a half-time (t1/2) of 10 min, and a first-order rate constant, k=0.069min-1, in 700 muM-phosphate buffer, pH 7.2, at 30 degrees C. The catalysis of the oxidation of Fe(III) by phosvitin was proportional to O2 concentration, and is quite different from the relative O2 independence of Fe(II) oxidation as catalysed by ferroxidase. A scheme for the mobilization and transfer of iron in the chicken, including the role of ferroxidase, phosyitin and transferrin, is presented.     

Ferroxidase kinetics of horse spleen apoferritin.

Protein ferroxidase site(s), which catalyze the reaction between ferrous ion and dioxygen, have long been thought to play a role in core formation in ferritin; however, the mechanism of the reaction has never been studied in detail. In the present work, the enzymatic activity of ferritin was examined using oximetry, the net Fe2+ oxidation reaction being as follows. [formula: see text] The reaction exhibits saturation kinetics with respect to both Fe2+ and O2 (apparent Michaelis constants: Km,Fe = 0.35 +/- 0.01 mM and Km,O2 = 0.14 +/- 0.03 mM). The enzyme has a turnover number kcat = 80 +/- 3 min-1 at 20 degrees C with maximal activity at pH 7. The kinetics are discussed in terms of two mechanisms, one involving monomeric and the other dimeric iron protein complexes. In both instances Fe(II) oxidation occurs in 1-electron steps. Zinc(II) is a competitive inhibitor of iron(II) oxidation at Zn2+/apoprotein ratios > or = 6 (inhibitor constant KI,Zn = 0.067 +/- 0.011 mM) but appears to be a noncompetitive inhibitor at lower ratios (< or = 2), indicating the presence of more than one type of zinc binding site on the protein. At increments of 50 Fe2+/protein or less, all of the iron is oxidized via the protein ferroxidase site(s), independent of the amount of core already present. However, when larger increments are employed, some iron oxidation appears to occur on the surface of the mineral core. The results of these studies emphasize the role of the protein shell in all phases of core growth and confirm the presence of a functionally important catalytic site in ferritin in addition to other binding sites on the protein for iron.     

Phosphate facilitates Fe(II) oxidative deposition in pea seed (Pisum sativum) ferritin

The iron core within phytoferritin interior usually contains the high ratio of iron to phosphate, agreeing with the fact that phosphorus and iron are essential nutrient elements for plant growth. It was established that iron oxidation and incorporation into phytoferritin shell occurs in the plastid(s) where the high concentration of phosphate occurs. However, so far, the role of phosphate in iron oxidative deposition in plant ferritin has not been recognized yet. In the present study, Fe(II) oxidative deposition in pea seed ferritin (PSF) was aerobically investigated in the presence of phosphate. Results indicated that phosphate did not affect the stoichiometry of the initial iron(II) oxidation reaction that takes place at ferroxidase centers upon addition of ≤48 Fe(II)/protein to apoferritin, but increased the rate of iron oxidation. At high Fe(II) fluxes into ferritin (>48 Fe(II)/protein), phosphate plays a more significant role in Fe(II) oxidative deposition. For instance, phosphate increased the rate of Fe(II) oxidation about 1–3 fold, and such an increase depends on the concentration of phosphate in the range of 0–2 mM. This effect was attributed to the ability of phosphate to improve the regeneration activity of ferroxidase centers in PSF. In addition, the presence of phosphate caused a significant decrease in the absorption properties of iron core, indicating that phosphate is involved in the formation of the iron core.     

M?ssbauer spectroscopic investigation of structure-function relations in ferritins.

Ferritin plays an important role in iron metabolism and our aim is to understand the mechanisms by which iron is sequestered within its protein shell as the mineral ferrihydrite. We present M?ssbauer spectroscopic data on recombinant human and horse spleen ferritin from which we draw the following conclusions: (1) that apoferritin catalyses Fe(II) oxidation as a first step in ferrihydrite deposition, (2) that the catalysis of Fe(II) oxidation is associated with residues situated within H chains, at the postulated 'ferroxidase centre' and not in the 3-fold inter-subunit channels previously suggested as the initial Fe(II) binding and oxidation site; (3) that both isolated Fe(III) and Fe(III) mu-oxo-bridged dimers found previously by M?ssbauer spectroscopy to be intermediates in iron-core formation in horse spleen ferritin, are located on H chains; and (4) that these dimers form at ferroxidase centres. The importance of the ferroxidase centre is suggested by the conservation of its ligands in many ferritins from vertebrates, invertebrates and plants. Nevertheless iron-core formation does occur in those ferritins that lack ferroxidase centres even though the initial Fe(II) oxidation is relatively slow. We compare the early stages of core formation in such variants and in horse spleen ferritin in which only 10-15% of its chains are of the H type. We discuss our findings in relation to the physiological role of isoferritins in iron storage processes.     

Spectroscopic studies on the binding of iron, terbium, and zinc by apoferritin

Ultraviolet difference spectroscopy has been used to study Fe (III)-apoferritin complexes formed after addition of Fe (II) to apoferritin in air. At constant iron, the recorded spectra varied with time after Fe (II) addition and with the number of iron atoms/molecule (protein concentration). The results indicate that after production of an initial complex, rearrangement or migration of Fe (III) atoms occurs, with polynuclear species forming as end-product, probably by hydrolytic polymerization. The presence of Tb3+ or Zn2+ ions affected the Fe (III) spectra and their development in different ways. The combined data suggest that more than one site, or processes, are involved in ferritin iron-core formation and that some of the metal sites are clustered.     

Binding of terbium to apoferritin: a fluorescence study

Luminescence measurements show that apoferritin binds three Tb(III) atoms per subunit in accordance with crystallographic evidence. Fe(II) competes with Tb(III) for at least some of the binding sites. This competition may be the molecular basis for the inhibition of iron incorporation into apoferritin brought about by Tb(III). Ca(II), which is generally replaced by Tb(III) in Ca(II) binding proteins, does not compete with the lanthanide for binding to apoferritin.     

Inhibition and stimulation of formation of the ferroxidase center and the iron core in Pyrococcus furiosus ferritin

Ferritin is a ubiquitous iron-storage protein that has 24 subunits. Each subunit of ferritins that exhibit high Fe(II) oxidation rates has a diiron binding site, the so-called ferroxidase center (FC). The role of the FC appears to be essential for the iron-oxidation catalysis of ferritins. Studies of the iron oxidation by mammalian, bacterial, and archaeal ferritin have indicated different mechanisms are operative for Fe(II) oxidation, and for inhibition of the Fe(II) oxidation by Zn(II). These differences are presumably related to the variations in the amino acid residues of the FC and/or transport channels. We have used a combination of UV–vis spectroscopy, fluorescence spectroscopy, and isothermal titration calorimetry to study the inhibiting action of Zn(II) ions on the iron-oxidation process by apoferritin and by ferritin aerobically preloaded with 48 Fe(II) per 24-meric protein, and to study a possible role of phosphate in initial iron mineralization by Pyrococcus furiosus ferritin (PfFtn). Although the empty FC can accommodate two zinc ions, binding of one zinc ion to the FC suffices to essentially abolish iron-oxidation activity. Zn(II) no longer binds to the FC nor does it inhibit iron core formation once the FC is filled with two Fe(III). Phosphate and vanadate facilitate iron oxidation only after formation of a stable FC, whereupon they become an integral part of the core. These results corroborate our previous proposal that the FC in PfFtn is a stable prosthetic group, and they suggest that its formation is essential for iron-oxidation catalysis by the protein.     

The Amyloid Precursor Protein (APP) Does Not Have a Ferroxidase Site in Its E2 Domain

The ubiquitous 24-meric iron-storage protein ferritin and multicopper oxidases such as ceruloplasmin or hephaestin catalyze oxidation of Fe(II) to Fe(III), using molecular oxygen as oxidant. The ferroxidase activity of these proteins is essential for cellular iron homeostasis. It has been reported that the amyloid precursor protein (APP) also has ferroxidase activity. The activity is assigned to a ferroxidase site in the E2 domain of APP. A synthetic 22-residue peptide that carries the putative ferroxidase site of E2 domain (FD1 peptide) has been claimed to encompass the same activity. We previously tested the ferroxidase activity of the synthetic FD1 peptide but we did not observe any activity above the background oxidation of Fe(II) by molecular oxygen. Here we used isothermal titration calorimetry to study Zn(II) and Fe(II) binding to the natural E2 domain of APP, and we employed the transferrin assay and oxygen consumption measurements to test the ferroxidase activity of the E2 domain. We found that this domain neither in the presence nor in the absence of the E1 domain binds Fe(II) and it is not able to catalyze the oxidation of Fe(II). Binding of Cu(II) to the E2 domain did not induce ferroxidase activity contrary to the presence of redox active Cu(II) centers in ceruloplasmin or hephaestin. Thus, we conclude that E2 or E1 domains of APP do not have ferroxidase activity and that the potential involvement of APP as a ferroxidase in the pathology of Alzheimer’s disease must be re-evaluated.     

Metal ion complexes of apoferritin. Evidence for initial binding in the hydrophilic channels

Metal binding to the iron storage protein apoferritin is the first step in the process by which iron accumulates within the protein shell. In the present study, the stoichiometry of metal binding to apoferritin in solution has been examined using the probe ions Mn(II), VO(IV), and Cd(II) in conjunction with EPR spectroscopic and cadmium ion selective electrode measurements. Binding studies were carried out with the individual ions, in competition with one another, and in competition with Fe(II), Fe(III), and Tb(III). All three probe ions show binding stoichiometries near 0.3 and 0.7 metal ion per subunit, close to the theoretically predicted values of 0.33 and 0.67 for the binding of one and two metal ions, respectively, per three subunits. These results in conjunction with other data are consistent with the binding of one, and possibly two, metal ions within each of the eight hydrophilic channels which are located on 3-fold axes leading to the interior of the protein. Pairs of cadmium binding sites have been located in these channels by x-ray crystallography (Rice, D. W., Ford, G. C., White, J. L., Smith, J. M. A., and Harrison, P. M. (1983) Adv. Inorg. Biochem. 5, 39-49). The possibility that some metal binding occurs elsewhere on the protein is not precluded by the present data, however. In competition experiments between various metal ions, approximately 0.3 metal ion per subunit is readily displaced implying common binding sites in the channels for all of them. The stoichiometry of Mn(II) displacement by Fe(II) is less clear. Oxidation of Fe(II) to Fe(III) by molecular oxygen in the presence of Mn(II) regenerates some Mn(II) binding on the protein, suggesting migration of iron(III) to other protein sites, or perhaps to core.     

Multiple pathways for mineral core formation in mammalian apoferritin. The role of hydrogen peroxide

Human ferritins sequester and store iron as a stable FeOOH((s)) mineral core within a protein shell assembled from 24 subunits of two types, H and L. Core mineralization in recombinant H- and L-subunit homopolymer and heteropolymer ferritins and several site-directed H-subunit variants was investigated to determine the iron oxidation/hydrolysis chemistry as a function of iron flux into the protein. Stopped-flow absorption spectrometry, UV spectrometry, and electrode oximetry revealed that the mineral core forms by at least three pathways, not two as previously thought. They correspond to the ferroxidase, mineral surface, and the Fe(II) + H2O2 detoxification reactions, respectively: [see reactions]. The H-subunit catalyzed ferroxidase reaction 1 occurs at all levels of iron loading of the protein but decreases with increasing iron added (48-800 Fe(II)/protein). Reaction 2 is the dominant reaction at 800 Fe(II)/protein, whereas reaction 3 occurs largely at intermediate iron loadings of 100-500 Fe(II)/protein. Some of the H2O2 produced in reaction 1 is consumed in the detoxification reaction 3; the 2/1 Fe(II)/H2O2 stoichiometry of reaction 3 minimizes hydroxyl radical production during mineralization. Human L-chain ferritin and H-chain variants lacking functional nucleation and/or ferroxidase sites deposit their iron largely through the mineral surface reaction 2. H2O2 is shown to be an intermediate product of dioxygen reduction in L-chain as well as in H-chain and H-chain variant ferritins.     

Iron deposition in apoferritin. Evidence for the formation of a mixed valence binuclear iron complex

A preliminary EPR investigation of iron accumulation in apoferritin has identified paramagnetic species generated during the early stage of iron deposition within the apoprotein shell. A featureless resonance at g' = 4.3, attributable to solitary high spin Fe3+ ions bound to the protein, is generated when Fe(II) is added to apoferritin at a level of 0.5 Fe/subunit (12 Fe/molecule) followed by air oxidation. This resonance accounts for 36% of the added iron. The remainder is EPR-silent and is probably present as oligomeric Fe3+ species. The intensity of the g' = 4.3 signal is reduced 3-fold upon anaerobic addition of 5 Fe(II)/subunit as a new iron resonance with g' values of 1.94, 1.87, and 1.80 is generated. This signal is observable only at temperatures near that of liquid helium and resists saturation at power levels of 100 milliwatts. Its distinctive g-factors, temperature dependence, and saturation characteristics suggest that it arises from a spin-coupled Fe(II)-Fe(III) dimer having a net electron spin of 1/2. In accord with this idea, the signal disappears when air is admitted, presumably because of oxidation of the Fe(II). The proposed mixed valence dimer may be an important intermediate formed during the initiation of core formation within the protein shell.     

Selective adsorption of apoferritin on immobilized Fe(III): demonstration of Fe(III) binding sites

Immobilized metal ion affinity chromatography has been used to demonstrate and partially characterize Fe(III) binding sites on apoferritin. Binding of Fe(III) to these sites is influenced by pH, but not affected by high ionic strength. These results suggest that both ionic and coordinate covalent interactions are important in the formation of the Fe(III): apoferritin complex. This is, to our knowledge, the first demonstration of direct Fe(III) binding to apoferritin. Other immobilized metal ions, including Zn(II), Ni(II), Cu(II), Cr(III), Co(II), and Tb(III), displayed little or no adsorption of apoferritin. The analytical technique of immobilized metal ion affinity chromatography also shows great promise in the purification of apoferritin, ferritin, and other iron-binding proteins.     

Lability of iron at the dinuclear centres of ferritin studied by competition with four chelators

Ferritin molecules contain 24 polypeptide chains folded as four-helix bundles and arranged as a hollow shell capable of storing up to 4500 Fe(III) atoms. H chains contain ferroxidase centres which lie within the bundle, about 12?Å (1.2?nm) from the outside surface and 8?Å from the inner surface of the protein shell. Catalysis of Fe(II) oxidation precedes storage of Fe(III) as ferrihydrite, with the formation of μ-oxo-bridged Fe(III) dimers as intermediates. Factors influencing the movement of μ-oxo-bridged Fe(III) from the ferroxidase centre to the ferritin cavity are uncertain. Assistance by small chelators is one possibility. The aim of this investigation was to determine whether iron at the dinuclear centres of three ferritins (human H chain homopolymer, HuHF, the non-haem ferritin of Escherichia coli, EcFTN, and horse spleen ferritin, HoSF) is accessible to chelators. Forty-eight Fe(II) atoms/molecule were added to the apoferritins followed, 2?min later, by the addition of chelator (1,10-phenanthroline, 2,2-bipyridine, desferrioxamine or 3,4-dihydroxybenzaldehyde). Iron species were analysed by Mössbauer spectroscopy or visible absorbance. Competition between chelators and apoferritin for Fe(II) was also investigated. The main conclusions of the study are that: (1) dinuclear iron and iron in small iron-cores in HuHF and EcFTN is mobilisable by all four chelators; (2) the chelators penetrate the shell; (3) 3,4-dihydroxybenzaldehyde is the most efficient in mobilising Fe(III) but the least successful in competing for Fe(II); (4) Fe(III) is more readily released from EcFTN than from HuHF; (5) 2,2′-bipyridine aids the movement of Fe(III) from ferroxidase centre to core.     

Identification of the iron entry channels in apoferritin. Chemical modification and spectroscopic studies

The knowledge of the route through which iron can enter and leave the apoferritin shell is a prerequisite for the understanding of ferritin's function. The involvement of the hydrophilic 3-fold channels in the iron uptake process has been studied by taking advantage of the reactivity of specific residues that line such channels, i.e., glutamic acid-127 and aspartic acid-130, the major Cd(II) binding sites, and cysteine-126. 113Cd NMR experiments have provided direct evidence for the competition between Fe(II) and Cd(II) binding to major Cd(II) binding sites on the protein and or a higher affinity of Fe(II) for these sites, in line with the well-known inhibitory effect of Cd(II) on iron uptake. Further evidence for the use of the 3-fold channels in the iron entry process has been obtained by means of chemical modification of Cys-126 with different mercurials. In particular, the introduction of the additional carboxylate carried by p-(chloromercuri)benzoate near Asp-127 and Glu-130 increases the initial rate of iron uptake and affects the coordination geometry of the metal in the Fe(III)-apoferritin complex as indicated by optical absorption and EPR data. The assignment of these effects to the carboxylate moiety of p-(chloromercuri)benzoate is brought out by the observation that the introduction in the 3-fold channel of the benzene ring only by means of phenylmercuric acetate has no effect on the initial iron uptake kinetics and on the spectroscopic properties of the Fe(III)-apoferritin complex.     

Iron(III) clusters bound to horse spleen apoferritin: an X-ray absorption and M?ssbauer spectroscopy study that shows that iron nuclei can form on the protein

Ferritin is a complex of a hollow, spherical protein and a hydrous, ferric oxide core of less than or equal to 4500 iron atoms inside the apoprotein coat; the apoprotein has multiple (ca. 12) binding sites for monoatomic metal ions, e.g., Fe(II), V(IV), Tb(III), that may be important in the initiation of iron core formation. In an earlier study we observed that the oxidation of Fe(II) vacated some, but not all, of the metal-binding sites, suggesting migration of some Fe during oxidation, possibly to form nucleation clusters; some Fe(III) remained bound to the protein. Preliminary extended X-ray absorbance fine structure (EXAFS) analysis of the same Fe(III)-apoferritin complex showed an environment distinct from ferritin cores, but the data did not allow a test of the Fe cluster hypothesis. In this paper, with improved EXAFS data and with M?ssbauer data on the same complex formed with 57Fe, we clearly show that the Fe(III) in the distinctive environment is polynuclear (Fe atoms with Fe-Fe = 3.5 A and TB = 7 K). Moreover, the arrangement of atoms is such that Fe(III) atoms appear to have both carboxylate-like ligands, presumably from apoferritin, and oxo bridges to the other iron atoms. Thus the protein provides sites not only for initiation but also for nucleation of the iron core. Sites commodious enough and with sufficient conserved carboxylate ligands to accommodate such a nucleus occur inside the protein coat at the subunit dimer interfaces. Such Fe(III)-apoferritin nucleation complexes can be used to study the properties of the several members of the apoferritin family.