Comparative Fe and Zn K-edge X-ray absorption spectroscopic study of the ferroxidase centres of human H-chain ferritin and bacterioferritin from Desulfovibrio desulfuricans.

Iron uptake by the ubiquitous iron-storage protein ferritin involves the oxidation of two Fe(II) ions located at the highly conserved dinuclear "ferroxidase centre" in individual subunits. We have measured X-ray absorption spectra of four mutants (K86Q, K86Q/E27D, K86Q/E107D, and K86Q/E27D/E107D, involving variations of Glu to Asp on either or both sides of the dinuclear ferroxidase site) of recombinant human H-chain ferritin (rHuHF) in their complexes with reactive Fe(II) and redox-inactive Zn(II). The results for Fe-rHuHf are compared with those for recombinant Desulfovibrio desulfuricans bacterioferritin (DdBfr) in three states: oxidised, reduced, and oxidised/Chelex-treated. The X-ray absorption near-edge region of the spectrum allows the oxidation state of the iron ions to be assessed. Extended X-ray absorption fine structure simulations have yielded accurate geometric information that represents an important refinement of the crystal structure of DdBfr; most metal-ligand bonds are shortened and there is a decrease in ionic radius going from the Fe(II) to the Fe(III) state. The Chelex-treated sample is found to be partly mineralised, giving an indication of the state of iron in the cycled-oxidised (reduced, then oxidised) form of DdBfr, where the crystal structure shows the dinuclear site to be only half occupied. In the case of rHuHF the complexes with Zn(II) reveal a surprising similarity between the variants, indicating that the rHuHf dinuclear site is rigid. In spite of this, the rHuHf complexes with Fe(II) show a variation in reactivity that is reflected in the iron oxidation states and coordination geometries.

Iron-reductases in the yeast Saccharomyces cerevisiae.

Several NAD(P)H-dependent ferri-reductase activities were detected in sub-cellular extracts of the yeast Saccharomyces cerevisiae. Some were induced in cells grown under iron-deficient conditions. At least two cytosolic iron-reducing enzymes having different substrate specificities could contribute to iron assimilation in vivo. One enzyme was purified to homogeneity: it is a flavoprotein (FAD) of 40 kDa that uses NADPH as electron donor and Fe(III)-EDTA as artificial electron acceptor. Isolated mitochondria reduced a variety of ferric chelates, probably via an 'external' NADH dehydrogenase, but not the siderophore ferrioxamine B. A plasma membrane-bound ferri-reductase system functioning with NADPH as electron donor and FMN as prosthetic group was purified 100-fold from isolated plasma membranes. This system may be involved in the reductive uptake of iron in vivo.

Iron transfer form transferrin to ferritin mediated by polyphosphate compounds.

We have studied iron transfer from transferrin to ferritin in the presence of ATP, GTP, ADP, AMP and 2,3-diphosphoglycerate. These compounds, with the exception of AMP, can release iron from transferrin at pH 7.4 and form a stable Fe(III)-phosphate complex. From these complexes, only a limited number of Fe(III) atoms can be incorporated into ferritin. Ascorbate enhances iron transfer from transferrin to ferritin at the beginning of the process but subsequently inhibits further iron deposition in ferritin.

Mobilization of iron from ferritin by isolated mitochondria. Effects of species compatibility between ferritin and mitochondria and iron content of ferritin.

Mitochondria mobilize iron from ferritin by a mechanism that depends on external FMN. With rat liver mitochondria, the rate of mobilization of iron is higher from rat liver ferritin than from horse spleen ferritin. With horse liver mitochondria, the rate of iron mobilization is higher from horse spleen ferritin than from rat liver ferritin. The results are explained by a higher affinity between mitochondria and ferritins of the same species. The mobilization of iron increases with the iron content of the ferritin and then levels off. A maximum is reached with ferritins containing about 1 200 iron atoms per molecule. The results represent further evidence that ferritin may function as a direct iron donor to the mitochondria.

Iron transfer from ferritin to transferrin. Effect of serum factors.

The transfer of iron from 59Fe-labelled human spleen ferritin to human apotransferrin occurs in the absence of either reducing or chelating agents. The reaction is first order with respect to ferritin and zero order to apotransferrin. The transfer is enhanced by low-Mr substances from human serum such as ascorbate, citrate, bicarbonate and lactate. A mixture of the four molecules at their normal physiological concentrations can increase the iron exchange to the same extent as that observed with an ultrafiltrate of serum. A pathway of intracellular iron mobilization is considered.

Over-expression of natural and variant human H-chain ferritins in E. coli.

The natural human H-chain ferritin was expressed in E. coli using a multi-copy expression vector containing the lambda pL promoter. A variant H-ferritin, having an altered N-terminus, was also produced. These proteins are overproduced (greater than 30% of the soluble protein), correctly assembled into its 24-subunit shell, and able to bind iron. The identity of the products was confirmed using an antibody specific for H-ferritin.

Iron storage in Saccharomyces cerevisiae.

A ferritin-like molecule was purified from iron-loaded cells of Saccharomyces cerevisiae, but its iron content was very low and was not representative of the cellular iron content. A study of the intracellular distribution of iron has shown that the vacuoles are involved in the storage of iron in the yeast cell. Moreover, it seems that this vacuolar iron can be further utilised by the cells for iron-requiring processes such as mitochondriogenesis.

Stabilisation and immobilisation of penicillin amidase.

Penicillin amidase was coupled to a periodate-oxidised dextran by reductive alkylation in the presence of sodium cyanoborohydride. A loss of activity (25%) was observed but the conjugate enzyme dextran was more thermostable than the native enzyme. Native and dextran-conjugated penicillin amidase were immobilised on amino activated silica (Promaxon, Spherosil, Aerosil) by a classical method using glutaraldehyde for the native enzyme and reductive alkylation for the modified enzyme. Good relative activity of the enzymes was obtained after insolubilisation. Immobilisation of both native and modified enzymes resulted in the thermostabilisation of the penicillin amidase.

Transferrin protein and iron uptake by cultured hepatocytes.

The binding and uptake of 59Fe-loaded 3H-labelled rat transferrin by cultured rat hepatocytes was investigated. At 4 degrees C, there is no evidence for a specific binding of transferrin which could be related to the association of neo-synthesized transferrin with plasma membrane receptors. At 37 degrees C, iron uptake is much more important than transferrin uptake; it proceeds linearly over the time of incubation, is largely proportional to the extracellular transferrin concentration, and is compatible with uptake by fluid phase endocytosis. The difference observed between iron and transferrin uptake implies the existence of a mechanism allowing the reutilization of transferrin after iron delivery.

Identification of the nuclear transcription factor NFkappaB in rat after in vivo ethanol administration.

NFkappaB, a nuclear transcription factor, was induced in the brain nuclear fraction of naive rats after an acute injection of ethanol, 2 g/kg. In contrast, rats which had been chronically alcoholised showed the constitutively active NFkappaB-like complex only after a further acute dose of ethanol. Hepatic nuclear fractions did not exhibit the specific NFkappaB-like complex during the first 45 min after acute ethanol injection, beyond that which was normally constitutively present. Such activation of NFkappaB-like complex in the brains of the naive rats may play an important role in the cellular protective response.

An animal model of iron overload and its application to study hepatic ferritin iron mobilization by chelators.

Administration of 3,5,5-trimethylhexanoyl ferrocene in the diet of male Wistar rats results in a substantial increase in hepatic ferritin protein (greater than 2-fold) and of ferritin iron (4-8-fold). The iron-loading in liver, under the conditions used, appears to be essentially in parenchymal cells rather than in reticulo-endothelial cells. It is suggested that the model represents a useful system for the study of the potential efficacy of new iron chelators for the mobilization of hepatic storage iron. The ability of desferal (DFO) and of a new siderophore, desferrithiocin (DFT), to mobilize hepatic ferritin iron is observed in this model of iron overload. Desferrithiocin stimulates ferritin iron mobilization, when administered either by gavage or by intraperitoneal injection, whereas desferal is active intraperitoneally but inactive orally. Our studies lead to the conclusion that DFT merits further examinations, for its activity as an orally active iron chelator.

Studies of in vivo iron mobilization by chelators in the ferrocene-loaded rat.

The oral efficacy of the oral iron chelators 1,2-dimethyl-3-hydroxypyrid-4-one (CP20), 1,2-diethyl-3-hydroxypyrid-4-one (CP94) and desferrioxamine B (DFO) has been compared with intraperitoneal DFO in an experimental model of iron overload with similar biochemical and biophysical characteristics to those observed for human genetic haemochromatosis. The hepatic iron stores in the ferrocene-loaded rat were relatively stable and did not decrease at the end of the loading period. In contrast, the iron dextran rat model showed a rapid depletion of its iron stores 2 weeks after cessation of intraperitoneal injection. When CP20 and CP94 were administered to the ferrocene-loaded rat model in combination with an iron-free diet there were significant decreases in (i) total homogenate iron and (ii) hepatic ferritin iron when compared to the iron-loaded rat receiving the iron-free diet alone. Desferrioxamine, when administered by gavage, only showed chelation of ferritin iron, while intraperitoneal injection of desferrioxamine showed significant depletion of iron both in the total homogenate and ferritin. Subcellular fractionation of the hepatic organelle clearly showed that where there was depletion of homogenate iron there was a net decrease in the lysosomal fraction, while changes in ferritin iron were reflected by decreases in the cytosolic iron content. Although no assessment of net iron excretion was made, we suggest that the use of this animal model should ascertain the site of chelation by iron chelators.

Iron mobilization from cultured hepatocytes: effect of desferrioxamine B.

When cultured rat hepatocytes prelabelled for different times at 37 degrees with 59Fe are reincubated for 1 hr in a fresh medium, radiolabelled iron is released in the washout medium as a function of the prelabelling time, and behaves like low molecular weight material on isokinetic centrifugation in sucrose gradients. When apotransferrin or desferrioxamine B are present in the reincubation medium, the kinetics of iron release are similar but the absolute amounts of radiolabelled iron found in the culture medium are much greater. In the presence of apotransferrin, most of the 59Fe released from the cells distributes as transferrin whereas with desferrioxamine B, almost all the 59Fe is extracted by benzyl alcohol indicating its chelation by the drug. Cell fractionation data indicate that iron accumulated by hepatocytes is rapidly incorporated into cytosol ferritin, and this seems to be a preferred source of iron for the chelator.

Effect of chronic chloroquine administration on iron loading in the liver and reticuloendothelial system and on oxidative responses by the alveolar macrophages.

The ability of chloroquine to alter iron loading in the liver, spleen, and alveolar macrophages was investigated in iron-loaded or -depleted rats. Chloroquine significantly reduced incorporation of iron into the liver, spleen, and alveolar macrophages of animals loaded in vivo with iron dextran. The ability of these macrophages to respond to oxidative stress was assayed by their capacity to release reactive nitrogen intermediates after lipopolysaccharide (LPS) stimulation. A significant reduction in nitrite release was observed in primary cultures of macrophages isolated from chloroquine/iron dextran-administered rats in comparison to macrophages lavaged from rats iron-loaded alone. Macrophages isolated from iron-deficient rats showed a significant increase in nitrite after LPS stimulation, whereas nitrite release in the macrophages lavaged from the rats which had also received chloroquine during the iron depletion regime was much lower. These results indicate that the use of agents which decrease the iron content and diminish the oxidative response of the cell to altered iron status may be of therapeutic value in patients with iron loading, particularly of the reticuloendothelial system.

Effects of desferrithiocin and its derivatives on peripheral iron and striatal dopamine and 5-hydroxytryptamine metabolism in the ferrocene-loaded rat.

Iron overload disorders, such as beta-thalassaemia, are currently treated with the iron chelator desferrioxamine (DFO) or 1,2-dimethyl-3-hydroxypyridin-4-one (L1), which is currently under clinical evaluation. However, DFO is inactive orally and needs to be administered by intramuscular infusion, whilst there are concerns over the long-term effectiveness and toxicity of L1. In addition, both DFO and L1 affect brain dopamine (DA) and 5-hydroxytryptamine (5-HT) metabolism. In this study, the 3,5,5-trimethylhexanoyl ferrocene rat model of iron overload was used to compare the iron-chelating capabilities of a novel orally active siderophore, desferrithiocin (DFT) and its desmethyl derivatives DFT-D and DFT-L, to that of DFO, along with their ability to affect brain DA and 5-HT metabolism. Chronic administration of ferrocene produced a 12-fold increase in liver iron levels, as assessed by electrothermal atomic absorption. Subsequent treatment with DFT over a two-week period produced a 37% reduction in liver iron levels, whereas similar treatment with DFT-D and DFT-L produced a more marked reduction in these levels (65% and 59%, respectively) in the ferrocene-treated animals. In contrast, using the same dosing regimen, DFO and L1 only produced a 16% and 18% reduction, respectively, in liver iron levels. Both DFT and its derivatives failed to affect either striatal DA or 5-HT metabolism when assessed by HPLC. In view of the previously described oral bioavailability of DFT, the marked ability of DFT and its derivatives to chelate hepatic iron, and their inability to affect brain DA or 5-HT metabolism, such siderophores appear potentially useful clinical iron chelators.

Cellular pharmacology of deferrioxamine B and derivatives in cultured rat hepatocytes in relation to iron mobilization.

Two radiolabelled derivatives of deferrioxamine B (DF) have been synthesized: methyl-DF and acetyl-DF. Both derivatives are non cytotoxic and stable in cell culture but they are degraded in human plasma and more extensively in rat plasma. Methyl-DF, acetyl-DF and DF mobilize radioiron to the same extent from hepatocytes loaded with 59Fe citrate in the same range of extracellular concentrations. The uptake and release of the 3H-labelled derivatives and their corresponding iron complexes have been measured and appear to represent a passive phenomenon resulting from the gradient of concentration between the cellular compartment and the extracellular medium. The results indicate that only a limited pool of cellular iron is accessible for chelation and that neither the permeability of the cellular membrane, nor the intracellular concentration of the chelators are the limiting factors for iron mobilization. On the basis of the subcellular distribution of the 3H-DF analogues, methylamine inhibition of iron chelation by siderophores in cell cultures and the positive effect of acidic pH and hydrolysis by lysosomal enzymes on in vitro iron mobilization from radiolabelled ferritin, we suggest that iron mobilization by DF and its derivatives occurs in lysosomes where they complex iron released from ferritin under the conjugate actions of acidic pH and lysosomal enzymes.

Brain iron in the ferrocene-loaded rat: its chelation and influence on dopamine metabolism.

After administration of the ferrocene derivative 3,5,5-trimethyl hexanoyl ferrocene to rats for 4 weeks various brain regions including substantia nigra, cerebellum and cerebral cortex showed up to 50% increase in iron content. Subsequent administration of one of the hydroxypyridones CP20, CP24 and CP94, or the siderophore desferrioxamine caused a significant decrease in the iron content of these various brain regions. Each of the hydroxypyridones and the siderophore influenced dopamine metabolism by causing significant variations in both homovanillic acid and dopamine turnover.

Desferrithiocin and desferrioxamine B. Cellular pharmacology and storage iron mobilization.

3H-Desferrithiocin (DFT) has been synthesized from desmethyl desferrithiocin. The uptake and release of this 3H siderophore and of its iron complex have been studied in cultured rat hepatocytes and systematically compared to 14C desferrioxamine B (DFO). At 37 degrees, the uptake of both chelators is strictly proportional to the extracellular concentration and no toxicity is observed up to, at least, 1 mM. Uptake of 3H DFT is rapid and reaches a plateau after ca. 1 hr. The accumulation of 3H DFT attains a maximum three times that of 14C DFO and the plateau is reached much more rapidly. Upon reincubation in a drug-free medium of cells that had accumulated 3H DFT, most of the 3H label is rapidly released in the culture medium. These kinetic parameters suggest that the accumulation of these two chelators results from their diffusion across cellular membranes, as a function of the gradient of concentration between the cellular compartment and the extracellular medium. Differential centrifugation of homogenates from hepatocytes incubated with 3H DFT shows that the bulk of cell associated 3H-label (82%) is found in the cytosol, whereas a small proportion (14.5%) is present in the particulate fraction. Isopycnic centrifugation on sucrose gradients suggests that 3H-label associated with the particulate fraction is localized within mitochondria. In contrast, 14C DFO distributes in almost equal proportions between cytosol and the particulate fraction (MLP). At least part of the 14C-label in MLP is associated with lysosomes. Rat hepatocytes cultivated for long term in synthetic culture medium have been used to study iron mobilization by chelators from 59Fe loaded cells. DFT mobilizes iron more rapidly than DFO. This effect is also observed in vitro with ferritin, where, in addition, DFT is much more efficient than DFO to mobilize iron at acidic pH. These results strongly suggest that different iron mobilization from cultured hepatocytes results from differences in the cellular pharmacology of these two chelators and, in particular, in their rate of uptake, cellular accumulation levels and subcellular localizations. DFT could mobilize iron from cytosol and, possibly, to a small extent from mitochondria, whereas DFO would do so from cytosol and lysosomes.

Decreased release of nitric oxide (NO) by alveolar macrophages after in vivo loading of rats with either iron or ethanol.

Alveolar macrophages were isolated by pulmonary lavage from rats which had been either chronically overloaded with iron by intraperitoneal injections of iron dextran for four weeks, or rendered alcoholic by administration of increasing concentrations of alcohol vapour, also for four weeks. Although the hepatic iron content increased in both groups of animals, only the macrophages isolated from the iron-loaded animals showed a significant increase in iron content (P = < 0.05). Furthermore, in these macrophages there was a significant increase in oxidative tone as demonstrated by a six fold increase in superoxide dismutase activity. In both the iron-loaded and chronically alcoholised macrophages, there was a significant diminution in nitric oxide release after stimulation with lipopolysaccharide and/or interferon-gamma, which impaired the ability of both of these groups of macrophages to inhibit the germination of spores from the fungus Rhizopus, a nitric oxide-dependent process. Such an alteration in nitric oxide release reduces the macrophage's microbicidal activity.

Iron mobilisation and cellular protection by a new synthetic chelator O-Trensox.

We tested a new synthetic, 8-hydroxyquinoline-based, hexadentate iron chelator, O-Trensox and compared it with desferrioxamine B (DFO). Iron mobilisation was evaluated: (i) in vitro by using ferritin and haemosiderin; DFO mobilised iron much more rapidly from ferritin at pH 7.4 than did O-Trensox, whereas at pH 4, ferritin and haemosiderin iron mobilisation was very similar with both chelators; (ii) in vitro by using cultured rat hepatocytes which had been loaded with 55Fe-ferritin; here DFO was slightly more effective after 100 hr than O-Trensox; (iii) in vivo administration i.p. to rats which had been iron-loaded with iron dextran; O-Trensox mobilised 51.5% of hepatic iron over two weeks compared to 48.8% for DFO. We also demonstrated the effect of O-Trensox in decreasing the entry of 55Fe citrate into hepatocyte cultures. The protective effect of O-Trensox against iron toxicity induced in hepatocyte cultures by ferric citrate was shown by decreased release of the enzymes lactate dehydrogenase (LDH), aspartate aminotransferase (AST), and alanine aminotranferase (ALT) from the cultures and, using electron paramagnetic resonance (EPR) measurements, decreased production of lipid radicals. O-Trensox was more effective than DFO in quenching hydroxyl radicals in an acellular system.