Interaction of anions with iron-transferrin-chelate complexes

Preparation and characterization of iron(III) nitrilotriacetate complex in aqueous solutions for quantitative protein binding experiments

Various preparations of iron(III) nitrilotriacetate (FeNTA) solution reported in the literature lack a comprehensive method for accurate determination of FeNTA concentration and often result in unstable solutions. A detailed procedure for the preparation of FeNTA solution is presented that includes the standardization of both components of the chelate. The standardization of the components allowed the accurate determination of the molar absorption coefficients for the calculation of the FeNTA concentration in two different buffers at pH 5.6 and 7.4. The variation of pH in this range or ionic strength in the range from 0 M to 3 M (KCl) has little effect on the value of the molar absorption coefficient. The precise concentrations of all species involved in the equilibria between Fe and NTA were determined in the pH range 2-12 using the Jenkins-Traub algorithm to solve the 5th-order polynomial in Microsoft Excel. In view of the experimental observations and the calculated distribution of species, the stability of FeNTA solutions may be affected by the Fe : NTA ratio and the total concentrations, with dilute solutions and those with an excess of NTA over Fe showing higher stability.

Anion binding properties of the transferrins. Implications for function

BACKGROUND

Since the transferrins have been defined by the highly cooperative binding of Fe(3+) and a carbonate anion to form an Fe-CO(3)-Tf ternary complex, the focus has been on synergistic anion binding. However, there are other types of anion binding with both apotransferrin and diferric transferrin that affect metal binding and release.

SCOPE OF REVIEW

This review covers the binding of anions to the apoprotein, as well as the formation and structure of Fe-anion-transferrin ternary complexes. It also covers interactions between ferric transferrin and non-synergistic anions that appear to be important in vivo.

GENERAL SIGNIFICANCE

The interaction of anions with apotransferrin can alter the effective metal binding constants, which can affect the transport of metal ions in serum. These interactions also play a role in iron release under physiological conditions.

MAJOR CONCLUSIONS

Apotransferrin binds a variety of anions with no special selectivity for carbonate. The selectivity for carbonate as a synergistic anion is associated with the iron binding reaction. Conformational changes in the binding of the synergistic carbonate and competition from non-synergistic anions both play a role in intracellular iron release. Anion competition also occurs in serum and reduces the effective metal binding affinity of Tf. Lastly, anions bind to allosteric sites (KISAB sites) on diferric transferrin and alter the rates of iron release. The KISAB sites have not been well-characterized, but kinetic studies on iron release from mutant transferrins indicate that there are likely to be multiple KISAB sites for each lobe of transferrin. This article is part of a Special Issue entitled Transferrins: Molecular mechanisms of iron transport and disorders.

Kinetics of metal ion exchange between citric acid and serum transferrin

The exchange of Fe(3+), Tb(3+), In(3+), Ga(3+), and Al(3+) between the C-terminal metal-binding site of the serum iron transport protein transferrin and the low-molecular-mass serum chelating agent citrate has been studied at pH 7.4 and 25 degrees C. The removal of Ga(3+), In(3+), and Al(3+) follows simple saturation kinetics with respect to the citrate concentration. In contrast, removal of both Fe(3+) and Tb(3+) shows a combination of saturation and first-order kinetic behavior with respect to the citrate concentration. The saturation component is consistent with a mechanism for metal release in which access to the bound metal is controlled by a rate-limiting conformational change in the protein. The first-order kinetic pathway is very rapid for Tb(3+), and this is attributed to a direct attack of the citrate on the Tb(3+) ion within the closed protein conformation. It is suggested that this pathway is more readily available for Tb(3+) because of the larger coordination number for this cation and the presence of an aquated coordination site in the Tb(3+)-CO(3)-Tf ternary complex. There is relatively little variation in the k(max) values for the saturation pathway for Tb(3+), Ga(3+), Al(3+), and In(3+), but the k(max) value for Fe(3+) is significantly smaller. It is suggested that protein interactions across the interdomain cleft of transferrin largely control the release of the first group of metal ions, while the breaking of stronger metal-protein bonds slows the rate of iron release. The rates of metal binding to apotransferrin are clearly controlled in large part by the hydrolytic tendencies of the free metal ions. For the more amphoteric metal ions Al(3+) and Ga(3+), there is rapid protein binding, and the addition of citrate actually retards this reaction. In contrast, the nonamphoteric In(3+) ion binds very slowly in the absence of citrate, presumably due to the rapid formation of polymeric In-hydroxo complexes upon addition of the unchelated metal ion to the pH 7.4 protein solution. The addition of citrate to the reaction accelerates the binding of In(3+) to apoTf, presumably by forming soluble, mononuclear In-citrate complexes.

Effect of lactoferrin on oxidative stability of corn oil emulsions and liposomes

Interest in using lactoferrin in foods for its antimicrobial activity inspired the present study of its antioxidant activity. Natural bovine lactoferrin inhibited oxidation in buffered corn oil emulsions and lecithin liposome systems at pH 6.6 and 50 degrees C. The antioxidant activity increased with lactoferrin concentration in both phosphate- and Tris-buffered emulsions, but not in both buffered liposome systems. A mixture of 1 microM lactoferrin and 0.5 microM ferrous ions was a significantly better antioxidant than 1 microM lactoferrin alone in Tris-buffered emulsions and in phosphate-buffered liposomes. Lactoferrin was a prooxidant at 1 microM in phosphate-buffered liposomes and at 15 and 20 microM in Tris-buffered liposomes. Copper was a stronger prooxidant than iron in both buffered emulsions. Lactoferrin decreased the prooxidant effect of iron, but not of copper, in emulsions. The antioxidant or prooxidant activities of lactoferrin depended on the lipid system, buffer, its concentration, the presence of metal ions, and oxidation time.

Kinetic studies on the removal of iron and aluminum from recombinant and site-directed mutant N-lobe half transferrins

Kinetic studies have been conducted in pH 7.4 Hepes buffer at 25 degreesC on the removal of Fe(III) and Al(III) from the recombinant N-lobe half molecule of human serum transferrin (Tf/2N) and from the R124A, K206A, and K296A mutants of this protein. The rates of iron removal from Tf/2N by 3-hydroxypyridin-4-one (deferiprone) and nitrilotriacetic acid (NTA) are essentially identical with previous results on N-terminal monoferric transferrin (Tf-FeN). For both Tf/2N and Tf-FeN, iron removal by deferiprone follows simple saturation kinetics, while iron removal by NTA follows simple first-order kinetics. There is some discrepancy between the two proteins with respect to iron removal by PPi, but this may be due to differences in the chloride concentrations among different studies. The addition of Fe(NTA)2 to R124A at ambient bicarbonate concentrations forms the Fe-NTA-Tf ternary complex, but the usual Fe-CO3-Tf complex can be formed by adding ferrous ion in the presence of a larger excess of bicarbonate. This complex releases its iron very rapidly by a mechanism that is first-order with respect to the ligand. This suggests that the first-order component of metal release from transferrin involves the displacement of the synergistic carbonate anion. Since iron removal from K206A and K296A at pH 7.4 is extremely slow, studies have been conducted on the more labile Al3+ complexes of Tf/2N, K206A, and K296A. The removal of Al3+ from Tf/2N by PPi follows the same complex kinetic order with respect to the ligand concentration that is observed for iron removal, while the removal of Al3+ from both K206A and K296A reverts to a simple saturation process. The addition of perchlorate retards the removal of Al3+ from both K206A and K296A, suggesting that these lysine residues are not associated with the allosteric effects of inorganic anions on the rates of metal removal.

Binding of monovalent anions to human serum transferrin

Serum transferrin is the protein whose primary function is to bind iron and transport it through the blood. Apotransferrin has two specific metal-binding sites that bind a variety of metal ions in addition to the ferric ion. The distinguishing feature of the transferrins is that a "synergistic" bicarbonate anion is bound along with the metal ion to form a stable Fe(3+)-CO3-Tf ternary complex. Previous research has shown that apotransferrin will also bind divalent anions such as phosphate and sulfate. Difference UV spectroscopy has now been used to show that a series of monovalent anions bind weakly to apotransferrin. Equilibrium constants for the binding of chloride, perchlorate, bromide, fluoride and Hepes have been calculated. A reaction scheme for the binding of anions is proposed which predicts that the binding of the nonsynergistic anions to apotransferrin will interfere with metal binding by competing directly with the binding of the synergistic bicarbonate anion. Difference UV data are presented which demonstrate this type of competition between nonsynergistic anions and Tb3+. Competition from the nonsynergistic anions follows the order HPO4(2-) > SO4(2-) approximately F- > ClO4- approximately Cl- approximately Br-. Speciation calculations have been performed to determine the concentrations of anion-apotransferrin complexes in Hepes and Tris buffers and in human serum and to estimate the extent to which competition from anions in the buffer will interfere with metal-binding to apotransferrin.

Structures and functionalities of milk proteins

During the last decade, marked progress has been made in the study of the fine details of the structures of milk proteins such as caseins, beta-lactoglobulin, alpha-lactalbumin, and lactotransferrin. Many of the functional properties of the individual milk proteins, as well as the milk protein products, may be described at the molecular level. This article is an attempt to thoroughly review the three-dimensional structures of major milk proteins, and to correlate them with the functional aspects of these proteins as food ingredients.

Periodate modification of human serum transferrin Fe(III)-binding sites. Inhibition of carbonate insertion into Fe(III)- and Cu(II)-chelator-transferrin ternary complexes

Periodate modification of human serum transferrin produces a species that binds Fe(III) weakly at pH 7.4 contrary to previous reports that Fe(III)-binding activity is completely lost. Ternary complexes of periodate-modified transferrin and either Fe(III) with nitrilotriacetate (NTA), oxalate, citrate, or EDTA, or of Cu(II) with oxalate could be formed. Peak wavelength maxima of these spectral bands are identical to those reported for native transferrin in the absence of bicarbonate. No carbonate ternary complexes of periodate-modified transferrin with Fe(III), Al(III), Cu(II), or Zn(II) could be formed. Conditional (Fe(NTA)) binding constants (log K) for C- and N-terminal modified sites are 7.33 and 7.54, respectively. The respective extinction coefficients at 470 nm are decreased 45% compared with the native protein. The electron paramagnetic resonance spectrum of the complex closely resembles that of the Fe(III)-NTA ternary complex formed with native transferrin in the absence of bicarbonate. Anions, including bicarbonate, at high concentrations destabilize formation of this Fe(III)-NTA ternary complex, while Fe(III) chelators readily remove the bound Fe(III). Bicarbonate, sulfate, and pyrophosphate still bind to the modified binding sites in the absence of metal although with slightly lower affinity and with lower molar difference absorptivities. Results are interpreted as an inhibition of a crucial protein conformational change by an intramolecular cross-link, preventing formation of the particularly stable metal-carbonate ternary complex from the less stable metal-chelate ternary complex. The method can be used to produce monosited transferrins.

Site selectivity in the binding of inorganic anions to serum transferrin

Equilibrium constants for the sequential binding of two anions at the specific metal-binding sites of apotransferrin have been measured by difference ultraviolet spectroscopy in 0.1 M N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (Hepes) at pH 7.4 and 25 degrees C. Log K1 values for phosphate, phosphite, sulfate, and arsenate fall in the narrow range of 3.5-4.0, while the log K1 for bicarbonate is 2.73. No binding is observed for nitrate, perchlorate, or borate. A dinegative charge appears to be the most important criterion for anion binding. Equilibrium constants have also been measured for binding of anions to both forms of mono(ferric)transferrin. There appears to be a very small site selectivity (0.2 to 0.4 log units) for phosphate, arsenate, and phosphite that favors binding to the N-terminal site, but there is no detectable selectivity for binding of sulfate or bicarbonate. Comparison of the binding affinities and anion selectivity with literature data on anion-binding to protonated macrocyles and cryptates strongly supports the existence of specific anion-binding sites on the protein. Binding constants were also measured in 0.01 M Hepes. The anionic sulfonate group of the buffer appears to have a small effect on anion binding.

Human platelets mediate iron release from transferrin by adenine nucleotide-dependent and -independent mechanisms

We assessed the ability of platelet sonicates and mediators secreted by unstimulated and thrombin-stimulated platelets to facilitate the release of iron from transferrin. Platelet sonicates and platelet conditioned media potentiated the release of iron from transferrin. The rate of release of iron was dependent on the pH of the reaction and amount of platelet sample added. Conditioned media from thrombin-stimulated platelets was more effective in mediating the release of iron from transferrin than was conditioned media from unstimulated cells. The rate of iron released from transferrin following addition of ATP and ADP in amounts equivalent to that present in platelet conditioned media was significantly less than the rate of iron released following the addition of conditioned media from platelets. Depletion of ATP and ADP in platelet conditioned media by incubation with apyrase only partially inhibited their ability to enhance the rate of iron release from transferrin. These observations indicate that platelets enhance the release of iron from transferrin by adenine nucleotide-dependent and -independent mechanisms. These observations are consistent with the hypothesis that platelets promote oxidant-induced tissue injury at sights of inflammation secondary to their ability to enhance the local release of iron from transferrin.

The influence of inorganic anions on the formation and stability of Fe3+-transferrin-anion complexes

Harris (Biochemistry 24 (1985) 7412) reports that inorganic anions bind to human apotransferrin in such a way as to perturb the ultraviolet spectrum. The locus of binding is thought to involve the specific metal/anion-binding sites since no perturbation is observed with Fe3+-transferrin-CO3(2-). Paradoxically, we were unable to demonstrate the formation of Fe3+-transferrin-inorganic anion complexes despite the presence of high concentrations of SO4(2-), H2PO4-, Cl-, ClO4- or NO3-. Similar results were found for human lactoferrin. Electron paramagnetic resonance spectroscopy and visible spectrophotometry were used to monitor the results. An attempt to form the H2PO4- complex by displacement of glycine from Fe3+-transferrin-glycine resulted only in the disruption of the ternary complex. A series of inorganic anions varied in their ability to release iron from Fe3+-transferrin-CO3(2-) at pH 5.5, the approximate pH of endosomes where iron release takes place within cells. The order of effectiveness was H2P2O7(2-) much greater than H2PO4- greater than SO4(2-) greater than NO3- greater than Cl- greater than ClO4-. The rate of iron removal from Fe3+-transferrin-CO3(2-) at pH 5.5 by a 4-fold excess of pyrophosphate was greatly enhanced by physiological NaCl concentration. Iron removal was complete within 10 min, the approximate time for iron release from Fe3+-transferrin-CO3(2-) in developing erythroid cells. Thus, inorganic anions may have a significant effect on the release of iron under physiological conditions despite the fact that such inorganic anions cannot act as synergistic anions. The results are discussed in relation to a special role for the carboxylate group in allowing ternary complex formation.

The interaction of anions with native and phenylglyoxal-modified human serum transferrin

The interaction of various anions with human serum transferrin was investigated due to the concomitant binding of iron and a synergistic anion to form the transferrin-anion-iron complex. Two tetrahedral oxyanion oxidizing agents, periodate and permanganate, were found to partially inactivate transferrin when used at equimolar ratios of oxidizing agent to protein active sites. Hypochlorite, a strong oxidizing agent with little structural similarity to periodate and permanganate, had little effect on iron-binding activity when used at similar low molar ratios of reagent to transferrin active sites. Transferrin treated with a 3:1 molar ratio of periodate or permanganate to active sites lost 74 or 67% of its iron-binding capacity, respectively. The composition of the buffer affected the extent of transferrin inactivation by periodate and permanganate; for example, the extent of inactivation by periodate was threefold greater in a borate buffer than in a phosphate buffer. Comparative oxidations in buffer systems suggest the following order of affinity of three buffer anions for the apotransferrin metal-binding center: phosphate greater than bicarbonate greater than borate. The interaction of phosphate ions with the iron-transferrin complex was also examined due to the increased susceptibility to periodate inactivation of iron-saturated transferrin in phosphate buffer (M. H. Penner, R. B. Yamasaki, D. T. Osuga, D. R. Babin, C. F. Meares, and R. E. Feeney (1983) Arch. Biochem. Biophys. 225, 740-747). The apparent destabilization of the iron-transferrin complex in phosphate buffer was found to be due to the competitive removal of iron by phosphate from the iron-protein complex. We found that phenylglyoxal-modified Fe-transferrin, with no loss of bound iron, was much more resistant to iron removal by phosphate and other competitive chelators.

Characterization of transferrin metal-binding sites by diffusion-enhanced energy transfer

The distance from the protein surface to ferric or manganic ions in the two specific metal-binding sites of human serum transferrin has been estimated by measuring energy transfer from freely diffusing terbium chelaters in aqueous solution to transferrin-bound metal ions. In addition, both monoferric forms of the protein were studied, as well as the diferric complex formed by using oxalate instead of (bi)carbonate as the auxiliary anion in binding of iron(III) to transferrin. Second-order rate constants for energy transfer between electrically neutral terbium(III)--N-(2-hydroxy-ethyl)ethylenediaminetriacetate and the FeA, FeB, and Fe2 forms of transferrin were 0.9 X 10(5) M-1 S-1, 1.4 X 10(5) M-1 S-1, and 2.6 X 10(5) M-1 S-1, respectively (based on iron concentraton). For the Fe2 species, substitution of oxalate for (bi)carbonate has the effect of decreasing the accessibility of both electrically neutral and negatively charged terbium chelates to the protein-bound iron chromophores. Theoretical considerations of the effect of acceptor location in the protein on energy transfer suggest that the iron chromophores are not on the surface of the protein but are less than 1.7 nm below the surface. The use of diterbium transferrin as energy donor to a small cobalt chelate in solution or to diferric transferrin corroborates these results.