Ferritin forms dynamic oligomers to associate with microtubules in vivo: implication for the role of microtubules in iron metabolism

Ferritin, a ubiquitously distributed iron storage protein, has been reported to interact with microtubules in vitro (Hasan et al., 2005, FEBS journal 272:822-831). Here, we demonstrate that ferritin binds with the microtubules in an oligomeric form and that the microtubule-bound ferritin contains more than two-fold amount of iron compared to the unbound ferritin fraction in vitro. Indirect immunofluorescence microscopy showed that a significant fraction of the ferritin molecules colocalized with the microtubules as oligomers in a wide variety of cell lines. These findings are consistent with the immediate oligomerization of rhodamine-labeled ferritin, microinjected in living human hepatoma cells. Ferritin oligomers were dynamic in the cytoplasm, and an anti-microtubule drug significantly inhibited their intracellular movement. Treatment of cells with an iron donor, ferric ammonium citrate, remarkably increased the number of cells containing ferritin oligomers. On the other hand, when the cells, such as mouse neuroblastoma cells, were deprived of iron, ferritin oligomers were localized in the microtubule dense, neurite shafts, but were disappeared from the microtubule deficient neurite tips. These data indicate that the microtubules provide a scaffold for the cytoplasmic distribution and transport of the iron-rich ferritin and implicate the role of microtubules in iron metabolism.     

Crosslinks between intramolecular pairs of ferritin subunits: effects on both H and L subunits and on immunoreactivity of sheep spleen ferritin

Ferritin is a multisubunit protein, controlling iron storage, with a protein coat composed of 24 subunits (up to three distinct types) in different proportions depending on cell type. Little is known about the subunit interactions in ferritin protein coats composed of heterologous subunits, despite the relevance to ferritin structure and ferritin function (iron uptake and release). Synthetic crosslinking is a convenient way to probe subunit contacts. Crosslinks between subunit pairs in ferritin protein coats are also a natural post-translational modification which coincides with different iron content in ferritin from sheep spleen; ferritin from sheep spleen also contains H and L subunits. Crosslinks synthesized by the reaction of ferritin low in natural crosslinks with difluorodinitrobenzene (F2DNB) reproduced the effects of the natural crosslinks on iron uptake and release. We now extend our observations on the structural effects of natural and synthetic crosslinks to include immunoreactivity of the assembled protein, with monoclonal antibodies as a probe. We also demonstrate, for the first time, ferritin peptides involved in an apparent H- and L-subunit contact: two peptides decreased 4X in cyanogen bromide peptide maps after F2DNB crosslinking were residues L-96-138 and H-66-96; the major DNP-dipeptide was Lys-DNP-Lys. Using the structure of an all L-subunit ferritin as a model, the most likely site for the H-L DNP crosslink is L-Lys 104 (C helix) and H-Lys 67 (B helix). The B helix forms the internal subunit dimer interface, a putative site of iron core nucleation. Alteration by crosslinks of the B helix could, therefore, explain the effect of crosslinks on ferritin iron uptake, release, and iron content.     

Titanium mineralization in ferritin: a room temperature nonphotochemical preparation and biophysical characterization

The incremental addition of titanium(III) citrate to H-chain homopolymers of human ferritin results in the formation of 1.5–6.5-nm particles of amorphous TiO₂ within the nanocage of the protein. The mineralization conditions are mild, featuring ambient temperature and no need for photochemical activation. Low ratios of titanium to protein favor intraprotein mineralization, and the products are characterized by stained and unstained transmission electron microscopy, UV–vis spectroscopy, dynamic light scattering, analytical ultracentrifugation, and metal analysis. With up to 1,000 equiv of metal, there is no change to the protein hydrodynamic radius or diffusion constant. There is, however, a systematic shift in the sedimentation coefficient, which confirms mineralization within the protein core.     

Characterization and Distribution of Ferritin Binding Sites in the Adult Mouse Brain

Abstract : Studies on iron uptake into the brain have traditionally focused on transport by transferrin. However, transferrin receptors are not found in all brain regions and are especially low in white matter tracts where high iron concentrations have been reported. Several lines of research suggest that a receptor for ferritin, the intracellular storage protein for iron, may exist. We present, herein, evidence for ferritin binding sites in the brains of adult mice. Autoradiographic studies using ¹²⁵I-recombinant human ferritin demonstrate that ferritin binding sites in brain are predominantly in white matter. Saturation binding analyses revealed a single class of binding sites with a dissociation constant ( K _D) of 4.65 × 10^-9 M and a binding site density ( B _max) of 17.9 fmol bound/μg of protein. Binding of radiolabeled ferritin can be competitively displaced by an excess of ferritin but not transferrin. Ferritin has previously been shown to affect cellular proliferation, protect cells from oxidative damage, and deliver iron. The significance of a cellular ferritin receptor is that ferritin is capable of delivering 2,000 times more iron per mole of protein than transferrin. The distribution of ferritin binding sites in brain vis-à-vis transferrin receptor distribution suggests distinct methods for iron delivery between gray and whi     

Ferritin in liver, plasma and bile of the iron-loaded rat

Rats were loaded with iron. With overload, up to a 10-fold increase of the iron and ferritin protein content of the livers was measured. The plasma ferritin concentration increased gradually with the ferritin concentration in the liver. The ferritin concentration in the bile increased also and was in the same range as in the plasma. The ratio plasma ferritin concentration to bile ferritin concentration in individual rats decreased in the case of considerable iron overload. After intravenous injection of liver ferritin, less than 2% of the ferritin concentration that disappeared from the blood was found to be in the bile. Isoelectric focussing revealed that the microheterogeneity of liver and bile ferritin were identical, but slightly different from plasma ferritin. These results indicate that ferritin was not solely leaking from the plasma to the bile. Together with ferritin, iron accumulated in the bile. The iron content of the bile ferritin was in the same range as in fully iron-loaded liver ferritin. It is likely that ferritin in the bile is excreted by the liver and consists of normal iron-loaded liver ferritin molecules. In all circumstances, the amount of iron in the bile was much higher than could be accounted for by transport by the bile ferritin. The ferritin protein to iron ratio in the bile was 0.1-1.2, which was in the same range as was measured in isolated lysosomal fractions of the liver. Those results agree with the supposition that ferritin and iron in the bile are excreted by the liver though lysosomal exocytosis.     

Ferritin as an exogenous yolk protein in snails

Summary A main yolk component in the oocytes of the pulmonate snailPlanorbarius corneus L. has been isolated and identified as the iron storage protein ferritin by its ultrastructure, iron content, immuunological properties and behaviour in disc electrophoresis. As judged from acrylamide electrophoresis data and ultrastructural observations, yolk ferritin is an exogenous protein which is synthesised in the hepatopancreas and taken up by the oocytes by endocytosis.     

Quantitative analysis of immunogold labelling for ferritin in liver from control and iron-overloaded rats

Summary The distribution of ferritin antigenicity in control and iron-loaded rat hepatocytes was investigated with an immunogold-ferritin antibody technique. Antibody to horse spleen ferritin showed immunoreactivity as determined by dot blotting with immunogold/silver staining with purified rat liver ferritin but not with rat haemosiderin. The initial site of ferritin degradation was studied by analysing the density of gold labelling in the cytosol and lysosomes in combination with pre-embedding acid phosphatase cytochemistry.Immunoreactive ferritin was present in the cytosol, cytosolic clusters and lysosomes of normal hepatocytes. After iron-loading, the labelling density increased over tenfold in parenchymal cell cytosol with a smaller increase in Kupffer cells. Ferritin clusters contained substantially more immunoreactive ferritin than equivalent areas of lysosomes or cytosol. Analysis of the labelling density in hepatocyte lysosomes showed that, despite a striking increase in iron content, one-quarter of the lysosomes showed less immunolabelled ferritin than the cytosol. The existence of a wide range of ferritin labelling densities in the lysosomes with a large proportion unlabelled suggests that the ferritin protein shell is not degraded at a significant rate either in the cytosol or in clusters but only after incorporation into lysosomes.     

Small-scale isolation of ferritin for the assay of the incorporation of 14C-labelled amino acids

1. A new procedure is described for the isolation of pure ferritin from small amounts of tissue. After the removal of most of the tissue proteins by heat coagulation, the ferritin fraction was chromatographed successively on CM-cellulose and Sephadex G-200. 2. The isolated ferritin appeared to be free from other proteins, as judged by its sedimentation pattern in the ultracentrifuge, by electrophoresis on polyacrylamide gel and by immunoelectrophoresis. 3. After the injection of [¹⁴C]leucine into a series of rats, the specific activity of liver ferritin isolated by the new procedure bore a constant relationship to that of liver ferritin separated by antigen–antibody precipitation. The procedure can thus be used to obtain ferritin of suitable purity for studies of amino acid incorporation. 4. The new procedure can be used to measure the total ferritin protein content of a tissue. It is not possible to use ultraviolet absorption for the measurement of ferritin protein because of considerable interference from the iron that it contains.     

Modification of ferritin during iron loading

Recombinant human ferritin loaded with iron via its own ferroxidase activity did not sediment through a sucrose-density gradient as a function of iron content. Analysis of the recombinant ferritin by native PAGE demonstrated an increase in altered migration pattern of the ferritins with increasing sedimentation, indicating an alteration of the overall charge of ferritin. Additionally, analysis of the ferritin by SDS-PAGE under nonreducing conditions demonstrated that the ferritin had formed large aggregates, which suggests disulfide bonds are involved in the aggregation. The hydroxyl radical was detected by electron spin resonance spectroscopy during iron loading into recombinant ferritin by its own ferroxidase activity. However, recombinant human ferritin loaded with iron in the presence of ceruloplasmin sedimented through a sucrose-density gradient similar to native ferritin. This ferritin was shown to sediment as a function of iron content. The addition of ceruloplasmin to the iron loading assay eliminated the detection of the DMPO-*OH adduct observed during loading using the ferroxidase activity of ferritin. The elimination of the DMPO-*OH adduct was determined to be due to the ability of ceruloplasmin to completely reduce oxygen to water during the oxidation of the ferrous iron. The implications of these data for the present models for iron uptake into ferritin are discussed.     

Heart tissue contains small and large aggregates of ferritin subunits

Ferritin purified from horse heart and applied to nondenaturing polyacrylamide gel electrophoresis migrated as a single band that stained for both iron and protein. This ferritin contained almost equal amounts of fast- and slow-sedimenting components of 58 S and 3-7 S, which could be separated on sucrose density gradients. Iron removal reduced the sedimentation coefficient of the fast-sedimenting ferritin to 18 S, and sedimentation equilibrium gave a molecular weight 650,000, with some preparations containing ferritin of 500,000 molecular weight as well. Sedimentation rates of the 3 S and 7 S ferritins were not affected by iron removal, and sedimentation equilibrium data were consistent with Mr's 40,000 and 180,000, respectively. Preparations of ferritin extracted from horse spleen contained only 67 S (holo) or 16 S (apo) ferritin and no slow-sedimenting species. When examined by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, all of the ferritins contained the usual H and L subunits (23 and 20 kDa, respectively), but the slow-sedimenting (3 S and 7 S) heart apoferritins also contained appreciable quantities (ca 25%) of three larger subunits of 42, 55, and 65 kDa. All the subunits reacted positively in Western blots to polyclonal antibodies made against specially purified large heart or spleen ferritins containing only 20- and 23-kDa subunits. Similar results were obtained for ferritins from rat heart. The results indicate that mammalian heart tissue is peculiar not just in having an abnormally large iron-rich ferritin but also in having iron-poor ferritins of much lower molecular weight, partly composed of larger subunits.     

FERRITIN IN THE FUNGUS PHYCOMYCES

The iron-protein ferritin has been purified from mycelium, sporangiophores, and spores of the fungus Phycomyces blakesleeanus. It has a protein-to-iron ratio of 5, a sedimentation coefficient of 55S, a buoyant density in CsCl of 1.82 g/cm³, and the characteristic morphology of ferritin in the electron microscope. Apoferritin prepared from Phycomyces ferritin has a sedimentation coefficient of 18S and consists of subunits of molecular weight 25,000. In the cytoplasm of Phycomyces, ferritin is located on the surface of lipid droplets (0.5–2.0 µ in diameter) where it forms crystalline monolayers which are conspicuous in electron micrographs of sporangiophore thin-sections. Ferritin is found in all developmental stages of Phycomyces but is concentrated in spores. The level of ferritin iron is regulated by the iron level in the growth medium, a 50-fold increase occurring on iron-supplemented medium.     

The formation of ferritin from apoferritin. Kinetics and mechanism of iron uptake

Ferritin has a high capacity as an iron store, incorporating some 4500 iron atoms as a microcrystalline ferric oxide hydrate. Starting from apoferritin, or ferritin of low iron content, Fe²⁺ and an oxidizing agent, the uptake of iron can be recorded spectrophotometrically. Progress curves were obtained and the reconstituted ferritin was shown by several physical methods to be similar to natural ferritin. The progress curves of iron uptake by apoferritin are sigmoidal; those for ferritins of low iron content are hyperbolic. The rate of iron uptake is dependent on the amount of iron already present in the molecule. The distribution of iron contents among reconstituted ferritin molecules is inhomogeneous. These findings are interpreted in terms of a crystal growth model. The surface area of the crystallites forming inside the protein increases until the molecule is half full, and then declines. This surface controls the rate at which new material is deposited. The experimental results can best be accounted for by a two-stage mechanism, an initial slow `nucleation' stage, which is apparently zero order with respect to [Fe<sup>2+</sup>], followed by a more rapid `growth' stage. The rate of Fe²⁺ oxidation is increased in the presence of apoferritin as compared with controls. Ferritin can therefore be regarded as an enzyme to which the product remains firmly attached. The protein appears to increase the rate of `nucleation'. The apparent zero order of this stage suggests the presence of binding sites on the protein, which are saturated with respect to Fe²⁺. These sites are presumed also to be oxidation sites. The oxidation and subsequent formation of the ferric oxide hydrate may proceed according to one of three alternative models.     

Iron content and degree of lipid peroxidation in liver mitochondria isolated from iron-loaded rats

1.The content of non-heme iron and the degree of lipid peroxidation were measured in liver mitochondria isolated from rats injected with either Jectofer (an iron-sorbitol-citric acid complex) or iron-nitrilotriacetate. 2. The sedimentation profiles of the mitochondria from controls and iron-treated rats as revealed by analytical differential centrifugation, indicated single population of mitochondria with $\text{s}{}_{\mathrm{4,B}}$ values of 13200± 560 S and 14200±590 S for controls and iron-loaded animals, respectively. In contrast, the sedimentation profiles of the acid phosphatase activity and the non-heme iron revealed marked polydispersities with at least three populations of particles for both controls and iron-loaded animals. 3. The mitochondria and iron-rich lysosomes were separated by density-gradient centrifugation in an isotonic medium of Percoll and sucrose. With this technique, the amount of non-heme iron in a mitochondrial fraction by differential centrifugation decreased from 69±28 nmol/mg protein to 5.6±1.1 nmol/mg protein and from 19.3±5.6 nmol/mg protein to 3.3±0.6 nmol/mg protein for Jectofer and iron-nitrilotriacetate injected rats, respectively. For control rats the amount of mitochondrial non-heme iron was about 2.7 nmol/mg protein both before and following density gradient centrifugation. The extra amount of non-heme iron still present in the purified mitochondrial fraction from iron-loaded rats, as compared to controls, was further characterized by the reactivity towards bathophenanthroline sulfonate. The results suggest that the extra iron was due to a small amount of either ferritin or hemosiderin still contaminaning the mitochondrial fraction. The amount of mitochondrial heme iron was the same in iron-loaded rats and controls. 4. The degree of lipid peroxidation in the mitochondria was estimated from the amount of malondialdehyde. The thiobarbituric acid method used for the quantitation of malondialdehyde was modified so that it was insensitive to variable amounts of iron present in the samples. No difference in the degree of lipid peroxidation was observed between the mitochondria from iron-loaded rats and controls. 5. In contrast to recent proposals (Hanstein, E.G. et al. (1981) Biochim. Biophys. Acta 678, 293–299), the present study showed that the amounts of non-heme iron and the degrees of lipid peroxidation are the same in mitochondria isolated from iron-loaded and control animals.     

Musca domestica hemolymph ferritin

We describe a method for the purification of ferritin from Musca domestica larval hemolymph. Musca ferritin occurs in hemolymph predominantly as a native protein with molecular weight equal to 550,000 and subunits of 26,000. The average iron content of purified ferritin was determined to be 3,000 ± 600 iron atoms per molecule. The iron contents of ferritin was heterogeneous; both fully iron loaded molecules and apoferritin are probably present in the Musca hemolymph. The anti-ferritin serum raised in rabbit was able to recognize native ferritin but was not reactive with the protein subunits isolated by SDS-PAGE. The ferritin concentration in hemolymph attains a maximum of 0.28 mg/ml in the wandering stage larvae, decreasing to 0.13 mg/ml at the middle of pupal stadium. The ferritin contents of midgut and fat bodies were also determined. Fat body ferritin content is greatly reduced when the feeding larva passes into wandering stage. © 1996 Wiley-Liss, Inc.     

Inflammatory macrophages in the dog contain high amounts of intravesicular ferritin and are associated with pouches of connective tissue fibers

We have studied the subcellular distribution of ferritin in inflammatory macrophages present in regional lymph nodes from dogs subjected to a pulmonary inflammatory reaction. The inflammatory reaction was induced by intrabronchial instillation of calcium tungstate (CaWO4), a water-insoluble powder. Ferritin was identified by electron microscopy, and its electron density was enhanced by the use of a modified Perls method. From day 14 on after the CaWO4 deposition, tungsten-positive lymph node macrophages showed a massive accumulation of ferritin. Most of the ferritin was stored in membrane-bounded vesicles that showed heterogeneous concentrations of the protein. A significant complement of ferritin was also detected in the cytoplasmic ground substance of phagocytes. The cell surface of the ferritin-rich, tungsten-positive macrophages showed deep infoldings that encompassed small pockets of connective tissue fibers. These features were not observed in control samples or in lymph nodes from dogs subjected to CaWO4-induced inflammation for periods shorter than 1 week. Our data indicate that inflammatory macrophages greatly increase their content of ferritin macrophages greatly increased their content of ferritin and that ferritin is stored predominantly by a membrane-bounded vesicular compartment. This is in contrast with suggestions that the inflammation-induced increase in macrophage iron is restricted to the labile pool of iron and it does not involve the iron bound to ferritin molecules. Our observation of nodules of connective-tissue fibers in intimate topographical association with ferritin-rich macrophages may indicate that the increase in intracellular ferritin in the macrophage is in some way related to the secretion of factors by the phagocyte that will stimulate fibrillogenesis by neighboring fibroblats.     

The UV and visible spectral properties of ferritin

The results of this investigation show that the visible and near ultraviolet extinction coefficients of the ferritin iron core increase as the content of iron per ferritin molecule increases. Additionally, the absorption spectrum of ferritin undergoes a red shift as the content of iron per molecule increases. These results suggest that use of the visible absorbance of ferritin to attempt to quantitate iron content or rate of iron exchange is subject to question. The experimentally determined coefficients of ferritin were used to calculate guidelines for the determination of the protein concentration of low-iron apoferritin or ferritin solutions by either absorbance or differential refractometry or a combination of both.     

Local conformational changes in ferritins with variable iron content

Conformational changes were induced in human spleen ferritin by partial or complete removal of iron, and the immunoreactivity of the ferritin samples with variable iron content was analyzed. We established that a decrease in iron content resulted in bimodal changes in immunoreactivity of the epitopes recognized by the monoclonal antibodies G10 and F11. Immunoreactivity demonstrated a 3-6-fold decrease on lowering iron content from 800 to 40 atoms per protein molecule, followed by a sharp (4-14-fold) increase that was observed when low-iron ferritin was converted to iron-free apoferritin. These bimodal changes suggest the presence of more than two conformational states of ferritin with local alterations of the epitopes recognized by the monoclonal antibodies. The global conformation of ferritin, however, remained essentially unaltered, as demonstrated by ferritin interaction with polyclonal antibodies. Together, the results indicate that local conformational changes in the ferritin protein shell occur on progressive iron removal that results in low-iron and iron-free forms of ferritin. These changes are most clearly seen in apoferritin when compared to low-iron ferritin.     

Studies on iron uptake and micelle formation in ferritin and apoferritin

Summary Iron uptake and micelle formation in ferritin and apoferritin have been followed both spectrophotometrically and by means of sedimentation velocity experiments. Information was thus obtained on the molecular weight distribution of the reconstitution product.To achieve incorporation native ferritin (whole ferritin as purified from horse spleen), native apoferritin (apoferritin prepared by fractionation of ferritin preparations) and reduced apoferritin (apoferritin prepared by reduction of ferritin by dithionite or ascorbic acid) have been incubated with ferrous salts in the presence of oxidizing agents under different experimental conditions.Although some iron is incorporated in native ferritin, full saturation is not achieved and the molecular weight distribution of the incubated products remains heterogeneous.Native and reduced apoferritin show a similar iron incorporation, but the reconstitution products markedly differ in terms of their iron distribution.Ferritin reconstituted from native apoferritin has a broad molecular weight distribution, while that reconstituted from reduced apoferritin is characterized by a narrow, homogeneous molecular weight distribution. However treatment of apoferritin with reducing or oxidizing agents prior to the incubation alters the characteristics of the iron distribution without changing the iron incorporation properties.These results point to a role of the protein moiety not only in iron oxidation, but also in micelle formation.