Aerobactin-mediated utilization of transferrin iron

Aerobactin and enterobactin, hydroxamate- and catechol-type siderophores, respectively, were found capable of removing iron (III) from transferrin in buffered solution. Although under these conditions aerobactin displaced the iron much more slowly than did enterobactin, the rate for the former could be accelerated by addition of pyrophosphate as mediator. Transfer of iron (III) from transferrin to aerobactin appeared to proceed via a ternary complex. Cells of Escherichia coli BN 3040 NalR iuc containing transport systems for both enterobactin and aerobactin, the genetic determinants for the latter specified on a ColV-type plasmid, took up iron from [55Fe]transferrin in minimal medium. In this case aerobactin was effective at a much lower concentration, although enterobactin still displayed superior ability to transfer the iron. In serum, however, the rate measured with aerobactin exceeded that found with enterobactin. The results indicate that aerobactin, in spite of its relatively unimpressive affinity for iron (III) as a siderophore, is nonetheless equipped with structural features or properties that enhance its ability to remove the metal ion from transferrin, especially when receptor-bearing cells of E. coli are present to act as a thermodynamic sink for the iron. These attributes of the aerobactin system of iron assimilation may account for its status as a virulence determinant in hospital isolates of E. coli.     

Effect of iron transferrin on nucleic acid synthesis in phytohemagglutinin-stimulated human lymphocytes

In serum-free cultures of phytohemagglutinin-stimulated human lymphocytes, iron transferrin causes enhanced uptake of both tritiated thymidine and tritiated uridine over that seen with only phytohemagglutinin. This effect is specific for the iron transferrin complex, no enhancement produced by either free iron(III) or apotransferrin. Iron bound to transferrin is quantitatively taken up by stimulated lymphocyte cultures, while under similar conditions only 10% of transferrin-bound zinc is incorporated. The relative specificity of action of iron and zinc on nucleic acid synthesis is discussed.     

Quantification of non-transferrin-bound iron in the presence of unsaturated transferrin.

Non-transferrin-bound iron (NTBI) has been reported to be associated with several clinical states such as thalassemia, hemochromatosis, and in patients receiving chemotherapy. We have investigated a number of ligands as potential alternatives to nitrilotriacetic acid (NTA) to capture NTBI without chelating transferrin- or ferritin-bound iron in plasma. We have established, however, that NTA is the optimal ligand to chelate the different forms of NTBI present in sera and can be adopted for utilization in the NTBI assay. NTA (80 mM) removes all forms of NTBI, while only mobilizing a small fraction of the iron bound to both transferrin and ferritin. We have compared three different detection systems for the quantification of NTA-chelated NTBI: the established HPLC-based method, a simple colorimetric method, and a method based on inductive conductiometric plasma spectroscopy. The sensitivity and reproductibility of the colorimetric method were acceptable compared with the other two methods and would be more convenient as a routine laboratory screening assay for NTBI. However, the limitations of this method are such that it can only be utilized in situations where desferrioxamine is not used and when transferrin saturation levels are close to 100%. Only the HPLC-based method is applicable for patients receiving (desferrioxamine) chelation therapy. In some diseases such as hemochromatosis, transferrin may be incompletely saturated. In such cases, to avoid in vitro donation of iron onto the vacant sites of transferrin, sodium-tris-carbonatocobaltate(III) can be added to block the free iron binding sites on transferrin. If this step is not taken, there may be an underestimation of NTBI values.     

Unfolding of iron and copper complexes of human lactoferrin and transferrin

1. Human lactoferrin and transferrin are capable of binding two iron or copper ions into specific binding sites in the presence of bicarbonate. 2. Urea and several alkyl ureas have been effective in unfolding these metal-protein complexes. 3.Biphasic transitions are observed for the unfolding of each of the metal complexes of these proteins as determined by direct visible spectroscopy suggesting the release of iron(III) and Cu(II) ions from both of these metal-binding proteins during the unfolding process. 4. Greater stabilization and increased resistance to protein unfolding is observed for all iron(III) complexes compared to Cu(II) complexes of lactoferrin and transferrin as determined by isothermal unfolding and thermal denaturation. 5. Relative stabilization of the different metal-protein complexes investigated within this study were determined to be as follows: Lf-Fe(III) greater than Lf-Cu(II); Tf-Fe(III) greater than Tf-Cu(II), and Lf-Fe(III) greater than Tf-Fe(III); Lf-Cu(II) greater than Tf-Cu(II).     

Resonance raman spectroscopy of iron (3)--ovotransferrin and iron (3)--human serum transferrin

We report the resonance Raman spectra in the frequency range 300–1800 cm^?1 of Fe (III)-ovotransferrin and Fe (III)-human serum transferrin in aqueous solution at about 10^?4M protein concentration. This is the first observation of resonance Raman scattering ascribable to amino acid ligand vibrational modes of a nonheme iron protein. The resonance Raman spectra of the transferrins are similar except that the resonance band near 1270 cm^?1 is shifted to a higher frequency for Fe(III)-human serum transferrin than that for Fe(III)-ovotransferrin. The resonance Raman bands observed near 1170, 1270, 1500 and 1600 cm^?1 may reflect resonance enhancement of p-hydroxy-phenyl frequencies of tyrosine residues and/or imidazolium frequencies of histidine residues.     

Fast biological iron chelators: kinetics of iron removal from human diferric transferrin by multidentate hydroxypyridonates

For decades, desferrioxamine B (Desferal) has been the therapeutic iron chelator of choice for iron-overload treatment, despite numerous problems associated with its use. Consequently, there is a continuous search for new iron chelating agents with improved properties, particularly oral activity. We have studied new potential therapeutic iron sequestering agents: multidentate ligands containing the hydroxypyridonate (HOPO) moiety. The ligands TRENCAM-3,2-HOPO, TRPN-3,2-HOPO, TREN-Me-3,2-HOPO, TREN-1,2,3-HOPO, 5LIO-3,2-HOPO, and BU-O-3,4-HOPO have been examined for their ability to remove iron from human diferric transferrin. The iron removal ability of the HOPO ligands is compared with that of the hydroxamate desferrioxamine B, the catecholates TRENCAM and enterobactin, as well as the bidentate hydroxypyridonate deferiprone, a proposed therapeutic substitute for Desferal. All the tested HOPO ligands efficiently remove iron from diferric transferrin at millimolar concentrations, with a hyperbolic dependence on ligand concentration. At high ligand concentrations, the fastest rates are found with the tetra- and bidentate hydroxypyridonates 5LIO-3,2-HOPO and deferiprone, and the slowest rates with the catecholate ligands. At low concentrations, closer to therapeutic dosage, hexadentate ligands which possess high pM values have the fastest rates of iron removal. TRENCAM-3,2-HOPO and TREN-Me-3,2-HOPO are the most efficient at lower doses and are regarded as having high potential as therapeutic agents. The kinetics of removal of Ga(III) from transferrin [in place of the redox active Fe(III)] were performed with TRENCAM and TREN-Me-3,2-HOPO to determine that there is no catalytic reduction step involved in iron removal.     

EPR study of serum ceruloplasmin and iron transferrin in myocardial infarction.

Electron paramagnetic resonance (EPR) analysis of frozen serum from myocardial infarction patients has been conducted. Signal at g=4.3 was found definitively attributable to iron(III)-transferrin complex. Imcrease of serum ceruloplasmin as compared to normal was confirmed, with a concomitant decrease of iron-transferrin content. A mechanism for such correlated variation is hypothesized.     

Interaction between 6-hydroxydopamine and transferrin: "Let my iron go"

The dopamine analogue 6-hydroxydopamine (6-OHDA) is selectively toxic to catecholaminergic neurons. Because of its selectivity for neuroblastic cells in the sympathetic nervous system lineage, 6-OHDA has been suggested as a chemotherapeutic agent for targeted treatment of patients with neuroblastoma. We tested the hypothesis that the toxicity of 6-OHDA is caused by its interaction with serum ferric transferrin (Fe-TF) resulting in release of iron. We further hypothesized that this iron, through its redox-cycling by 6-OHDA, triggers generation of reactive oxygen species. 6-OHDA-induced release of iron from Fe-TF was demonstrated by: (1) low-temperature EPR spectroscopic evidence for decay of the characteristic Fe-TF signal (g = 4.3) and appearance of the high-spin signal from iron chelated by 6-OHDA oxidation products; (2) spectrophotometric detection of complexing of iron with the Fe(2+) chelator ferrozine; (3) redox-cycling of ascorbate yielding EPR-detectable ascorbate radicals; and (4) generation of hydroxyl radicals as evidenced by EPR spectroscopy of their adduct with a spin trap, 5, 5'-dimethylpyrroline oxide (DMPO) (DMPO-OH). Our low-temperature EPR studies showed that in human plasma, 6-OHDA caused iron release only under nitrogen gas but not under air or oxygen. The absence of a 6-OHDA effect in plasma under aerobic conditions was most likely due to its ferroxidase activity [with consequent reuptake of Fe(III) by apoTF] and catalytic oxidation of 6-OHDA by ceruloplasmin. Modeling of these plasma activities by a stable nitroxide radical, 2,2,6, 6-tetramethyl-1-piperidinyloxy (TEMPOL), resulted in protection of plasma Fe-TF against iron release under nitrogen. Parenteral administration of 6-OHDA to mice resulted in iron release from Fe-TF as evidenced by transformation of the Fe-TF low-temperature EPR signal that was indistinguishable from that seen in in vitro models. In addition, administration of the iron chelator deferoxamine (DFO) to mice prior to administration of toxic doses of 6-OHDA resulted in a decrease in activity impairment of mice as compared to that seen with 6-OHDA alone. These findings underscore the physiological and pharmacological relevance of 6-OHDA-mediated iron release from Fe-TF and suggest that iron chelators (DFO) may be used for prevention of 6-OHDA toxicity.     

Concentration of transferrin iron and transferrin saturation in blood

Transferrin iron, transferrin protein concentrations, and transferrin saturation have been determined for the first time in the whole blood. Microsamples were taken from healthy adults and patients with occupational secondary haemochromatosis using quantitative electron spin resonance technique. At elevated transferrin saturation, transferrin saturation values determined in the plasma and serum samples were shown to be less than respective values determined in the whole blood of the same patients. At increased transferrin iron concentration the difference between experimental and reference data sets determined in the blood and plasma was statistically significant in contrast to data sets determined in serum. Therefore, the analysis of the blood microsamples ensured an adequate estimation of transferrin iron concentration, especially at high transferrin saturation. A new index--transferrin iron concentration in the formed blood elements--was introduced. The values of the index were determined in the groups of healthy adults, patients with secondary occupational hemochromatosis and healthy newborns.     

Studies of the binding of different iron donors to human serum transferrin and isolation of iron-binding fragments from the N- and C-terminal regions of the protein.

1. Trypsin digestion of human serum transferrin partially saturated with iron(III)-nitrilotriacetate at pH 5.5 or pH 8.5 produces a carbohydrate-containing iron-binding fragment of mol.wt. 43000. 2. When iron(III) citrate, FeCl3, iron (III) ascorabate and (NH4)2SO4,FeSO4 are used as iron donors to saturate the protein partially, at pH8.5, proteolytic digestion yields a fragment of mol.wt. 36000 that lacks carbohydrate. 3. The two fragments differ in their antigenic structures, amino acid compositions and peptide 'maps'. 4. The fragment with mol.wt. 36000 was assigned to the N-terminal region of the protein and the other to the C-terminal region. 5. The distribution of iron in human serum transferrin partially saturated with various iron donors was examined by electrophoresis in urea/polyacrylamide gels and the two possible monoferric forms were unequivocally identified. 6. The site designated A on human serum transferrin [Harris (1977) Biochemistry 16, 560--564] was assigned to the C-terminal region of the protein and the B site to the N-terminal region. 7. The distribution of iron on transferrin in human plasma was determined.     

Arsenic species that cause release of iron from ferritin and generation of activated oxygen

The in vitro effects of four different species of arsenic (arsenate, arsenite, monomethylarsonic acid, and dimethylarsinic acid) in mobilizing iron from horse spleen ferritin under aerobic and anaerobic conditions were investigated. Dimethylarsinic acid (DMA(V)) and dimethylarsinous acid (DMA(III)) significantly released iron from horse spleen ferritin either with or without the presence of ascorbic acid, a strong synergistic agent. Ascorbic acid-mediated iron release was time-dependent as well as both DMA(III) and ferritin concentration-dependent. Iron release from ferritin by DMA(III)) alone or with ascorbic acid was not significantly inhibited by superoxide dismutase (150 or 300 units/ml). However, the iron release was greater under anaerobic conditions (nitrogen gas), which indicates direct chemical reduction of iron from ferritin by DMA(III), with or without ascorbic acid. Both DMA(V) and DMA(III)) released iron from both horse spleen and human liver ferritin. Further, the release of ferritin iron by DMA(III)) with ascorbic acid catalyzed bleomycin-dependent degradation of calf thymus DNA. These results indicate that exogenous methylated arsenic species and endogenous ascorbic acid can cause (a) the release of iron from ferritin, (b) the iron-dependent formation of reactive oxygen species, and (c) DNA damage. This reactive oxygen species pathway could be a mechanism of action of arsenic carcinogenesis in man.     

Factors affecting the adenosine triphosphate induced release of iron from transferrin.

The release of iron from transferrin was investigated by incubating the diferric protein in the presence of potential iron-releasing agents. The effective chemical group appears to be pyrophosphate, which is present in blood cells as nucleoside di- and triphosphates, notably adenosine triphosphate (ATP). An alternative structure with comparable activity is represented by 2,3-diphosphoglycerate. Neither 1 mM adenosine monophosphate (AMP) nor 1 mM orthophosphate released iron from transferrin. The ATP-induced iron-releasing activity was dependent on weak acidic conditions and was sensitive to temperature and sodium chloride concentration. The rate of iron release rapidly increased as transferrin was titrated with HCl from pH 6.8 to 6.1 in the presence of 1 mM ATP and 160 mM NaCl at 20 degrees C. Iron release from transferrin without ATP was observed below pH 5.5. Ascorbate (10(-4) M) reduced Fe(III), but only after iron release from transferrin by a physiological concentration of ATP. A proposal for the mechanism of iron release from transferrin by ATP and the utilization of reduced iron by erythroid cells is described.     

The effect of acid pH and citrate on the release and exchange of iron on rat transferrin.

The effect of acid pH and citrate on the exchange of iron between binding sites of rat transferrin has been studied. In the absence of citrate, diferric transferrin shows stepwise loss of iron atoms with the first atom of iron released at approximately pH 5.2. Citrate at physiologic concentrations (1.10(-3) M) or greater allows random iron removal at pH 6.5 or less. Iron dissociation from monoferric transferrin at acid pH, with or without citrate, is a random process. At pH 7.4, randomization of iron on transferrin takes from 3 to 6 h in the presence of millimolar concentrations of citrate. We conclude that at acid pH and in the presence of citrate concentrations likely to occur in vivo in the rat there is little scrambling of iron bound to transferrin.     

Hydroxylated phytosiderophore species possess an enhanced chelate stability and affinity for iron(III)

Graminaceous plant species acquire soil iron by the release of phytosiderophores and subsequent uptake of iron(III)-phytosiderophore complexes. As plant species differ in their ability for phytosiderophore hydroxylation prior to release, an electrophoretic method was set up to determine whether hydroxylation affects the net charge of iron(III)-phytosiderophore complexes, and thus chelate stability. At pH 7.0, non-hydroxylated (deoxymugineic acid) and hydroxylated (mugineic acid; epi-hydroxymugineic acid) phytosiderophores form single negatively charged iron(III) complexes, in contrast to iron(III)-nicotianamine. As the degree of phytosiderophore hydroxylation increases, the corresponding iron(III) complex was found to be less readily protonated. Measured pKa values of the amino groups and calculated free iron(III) concentrations in presence of a 10-fold chelator excess were also found to decrease with increasing degree of hydroxylation, confirming that phytosiderophore hydroxylation protects against acid-induced protonation of the iron(III)-phytosiderophore complex. These effects are almost certainly associated with intramolecular hydrogen bonding between the hydroxyl and amino functions. We conclude that introduction of hydroxyl groups into the phytosiderophore skeleton increases iron(III)-chelate stability in acid environments such as those found in the rhizosphere or the root apoplasm and may contribute to an enhanced iron acquisition.     

The effect of acid pH and citrate on the release and exchange of iron on rat transferrin

The effect of acid pH and citrate on the exchange of iron between binding sites of rat transferrin has been studied. In the absence of citrate, diferric transferrin shows stepwise loss of iron atoms with the first atom of iron released at approximately pH 5.2. Citrate at physiologic concentrations (1 · 10^?3 M) or greater allows random iron removal at pH 6.5 or less. Iron dissociation from monoferric transferrin at acid pH, with or without citrate, is a random process. At pH 7.4, randomization of iron on transferrin takes from 3 to 6 h in the presence of millimolar concentrations of citrate. We conclude that at acid pH and in the presence of citrate concentrations likely to occur in vivo in the rat there is little scrambling of iron bound to transferrin.     

Allosteric effects of sulfonate anions on the rates of iron release from serum transferrin

Serum transferrin is the protein that transports ferric ion through the bloodstream and is thus a potential target for iron chelation therapy. However, the release of iron from transferrin to low-molecular-weight chelating agents is usually quite slow. Thus a better understanding of the mechanism for iron release is important to assist in the design of more effective agents for iron removal. This paper describes the effect of sulfonate anions on the rates of iron removal from C-terminal monoferric transferrin by acetohydroxamic acid, deferiprone, nitrilotriacetic acid (NTA), and diethylenetriaminepentaacetic acid at 25 °C in 0.1 M N-(2-hydroxyethyl)piperazine-N′-(2-ethanesulfonic acid) (Hepes) buffer at pH 7.4. These ligands remove iron via a combination of pathways that show saturation and first order dependence on the ligand concentration. The kinetic effects of the anions methanesulfonate, methylenedisulfonate, and ethylenedisulfonate were evaluated. All these anions increase the overall rates of iron release, presumably by binding to an allosteric anion binding site on the protein. The two disulfonates produce a larger acceleration in iron release than the monosulfonate. More detailed studies using methylenedisulfonate show that this anion accelerates the rate of iron release via the saturation pathway. The addition of methylenedisulfonate results in the appearance of a large saturation pathway for iron release by NTA, which otherwise removes iron by a simple first-order process. The sulfonate group was selected for these studies because it represents an anionic functional group that can be covalently linked to a therapeutic ligand to accelerate iron release in vivo. The current studies indicate that the binding of the sulfonates to the allosteric site on the protein is quite weak, so that one would not expect a significant acceleration in iron release at clinically relevant ligand concentrations.     

Exchange of iron by gallium in siderophores

Siderophores are iron transport compounds produced by numerous microorganisms and which strongly chelate Fe(III), but not Fe(II). Other trivalent metals, such as Al(III), Cr(III), or Ga(III), are not capable of significantly displacing iron from siderophores. However, I demonstrate here that Ga(III) can effectively displace iron under reducing conditions. With ascorbate as reductant and ferrozine as Fe(II) trapping agent, the kinetics of reductive displacement of iron by Ga(III) were followed spectroscopically by the increase of absorbance at 562 nm due to formation of the Fe(II)-ferrozine complex. No significant reduction of siderophore occurred in the absence of Ga(III). With excess Ga(III), the displacement was quantitative and very rapid. The rate of metal exchange was pseudo first order with respect to Ga(III) concentration and highly pH dependent, suggesting that siderophore ligands are displaced from the iron in a concerted mechanism by Ga(III) and protonation to expose the Fe(III) to reduction by ascorbate. Reaction rates were dependent upon the structure of the siderophore, being greatest for ferric rhodotorulic acid and slowest for ferrichrome A at pH 5.4. The pH profile for ferric rhodotorulic acid was unusual in that it showed a maximum at pH 6.5, while all other siderophores examined showed an increase in rate as pH was lowered from 7.0. The physiological significance of this reaction to the clinical use of gallium is discussed.     

OsYSL16 plays a role in the allocation of iron

Graminaceous plants acquire iron by secreting mugineic acid family phytosiderophores into the rhizosphere and taking up complexes of iron and phytosiderophores through YSL (yellow stripe 1-like) transporters. Rice OsYSL15 is a transporter of the iron(III)-2'-deoxymugineic acid complex. OsYSL16 has 85?% similarity to both OsYSL15 and the iron(II)-nicotianamine transporter OsYSL2. In the present study, we show that OsYSL16 functionally complemented a yeast mutant defective in iron uptake when grown on medium containing iron(III)-deoxymugineic acid, but not when grown on medium containing iron(II)-nicotianamine. OsYSL16-knockdown seedlings were smaller than wild-type seedlings when only iron(III)chloride was supplied as an iron source. The iron concentration in shoots of OsYSL16-knockdown plants was similar to that of the wild type; however, they showed more severe chlorosis than wild-type plants under iron-deficient conditions. Furthermore, OsYSL16-knockdown plants accumulated more iron in the vascular bundles of the leaves. Expression of the OsYSL16 promoter fused to the β-glucuronidase gene showed that OsYSL16 is expressed in the root epidermis and vascular bundles of whole plants. The expression was typically observed around the xylem. In the vascular bundles of unelongated nodes, it was detected in the xylem of old leaves and the phloem of new leaves. Graminaceous plants translocate iron from the roots to old leaves mainly via the xylem and to new leaves mainly via the phloem. Our results suggest that OsYSL16 plays a role in the allocation of iron(III)-deoxymugineic acid via the vascular bundles.     

Evolution of the transferrin family: conservation of residues associated with iron and anion binding

The transferrin family spans both vertebrates and invertebrates. It includes serum transferrin, ovotransferrin, lactoferrin, melanotransferrin, inhibitor of carbonic anhydrase, saxiphilin, the major yolk protein in sea urchins, the crayfish protein, pacifastin, and a protein from green algae. Most (but not all) contain two domains of around 340 residues, thought to have evolved from an ancient duplication event. For serum transferrin, ovotransferrin and lactoferrin each of the duplicated lobes binds one atom of Fe (III) and one carbonate anion. With a few notable exceptions each iron atom is coordinated to four conserved amino acid residues: an aspartic acid, two tyrosines, and a histidine, while anion binding is associated with an arginine and a threonine in close proximity. These six residues in each lobe were examined for their evolutionary conservation in the homologous N- and C-lobes of 82 complete transferrin sequences from 61 different species. Of the ligands in the N-lobe, the histidine ligand shows the most variability in sequence. Also, of note, four of the twelve insect transferrins have glutamic acid substituted for aspartic acid in the N-lobe (as seen in the bacterial ferric binding proteins). In addition, there is a wide spread substitution of lysine for the anion binding arginine in the N-lobe in many organisms including all of the fish, the sea squirt and many of the unusual family members i.e., saxiphilin and the green alga protein. It is hoped that this short analysis will provide the impetus to establish the true function of some of the TF family members that clearly lack the ability to bind iron in one or both lobes and additionally clarify the evolutionary history of this important family of proteins.