Periodate inactivation of ovotransferrin and human serum transferrin

Azari and Phillips (Azari, P., and Phillips, J. L. 1970 Arch. Biochem. Biophys. 138, 32-38) reported that periodate treatment of iron-free ovotransferrin causes a rapid loss of iron-binding activity and an oxidation of 3 to 5 tyrosines and 1 tryptophan. Rapid inactivation and loss of tyrosine in ovotransferrin has been confirmed, and the work extended to human serum transferrin and effects of denaturing concentrations of urea. Extensive (> 80%) inactivations of both ovotransferrin and human serum transferrin were observed when approximately 4 tyrosines were destroyed. Amino acid analysis and 360-MHz 1H NMR spectra confirmed that tyrosines are the only residues rapidly oxidized; the correlation of tyrosine loss with the loss of iron-binding activity suggests strongly that the tyrosines involved are those that function as ligands to metal ions bound to the protein. NMR spectra also showed that periodate oxidation causes local changes of structure in ovotransferrin (presumably at the metal-binding sites) but does not grossly alter the conformation. The addition of 5 to 8 M urea greatly retarded the inactivation and losses of tyrosine.     

Aluminum exchange between citrate and human serum transferrin and interaction with transferrin receptor 1

The kinetics and thermodynamics of Al(III) exchange between aluminum citrate (AlL) and human serum transferrin were investigated in the 7.2-8.9 pH range. The C-site of human serum apotransferrin in interaction with bicarbonate removes Al(III) from Al citrate with an exchange equilibrium constant K1 = (2.0 +/- 0.6) x 10(-2); a direct second-order rate constant k1 = 45 +/- 3 M(-1) x s(-1); and a reverse second-order rate constant k(-1) = (2.3 +/- 0.5) x 10(3) M(-1) x s(-1). The newly formed aluminum-protein complex loses a single proton with proton dissociation constant K1a = (15 +/- 3) nM to yield a first kinetic intermediate. This intermediate then undergoes a modification in its conformation followed by two proton losses; first-order rate constant k2 = (4.20 +/- 0.02) x 10(-2) s(-1) to produce a second kinetic intermediate, which in turn undergoes a last slow modification in the conformation to yield the aluminum-loaded transferrin in its final state. This last process rate-controls Al(III) uptake by the N-site of the protein and is independent of the experimental parameters with a constant reciprocal relaxation time tau3(-1) = (6 +/- 1) x 10(-5) x s(-1). The affinities involved in aluminum uptake by serum transferrins are about 10 orders of magnitude lower than those involved in the uptake of iron. The interactions of iron-loaded transferrins with transferrin receptor 1 occur with average dissociation constants of 3 +/- 1 and 5 +/- 1 nM for the only C-site iron-loaded and of 6.0 +/- 0.6 and 7 +/- 0.5 nM for the iron-saturated ST in the absence or presence of CHAPS, respectively. No interaction is detected between receptor 1 and aluminum-saturated or mixed C-site iron-loaded/N-site aluminum-loaded transferrin under the same conditions. The fact that aluminum can be solubilized by serum transferrin in biological fluids does not necessarily imply that its transfer from the blood stream to cytoplasm follows the receptor-mediated pathway of iron transport by transferrins.     

In vitro interaction between ceruloplasmin and human serum transferrin

The thermodynamics of the interactions of serum apotransferrin (T) and holotransferrin (TFe(2)) with ceruloplasmin (Cp), as well as those of human lactoferrin (Lf), were assessed by fluorescence emission spectroscopy. Cp interacts with two Lf molecules. The first interaction depends on pH and μ, whereas the second does not. Dissociation constants were as follows: K(11Lf) = 1.5 ± 0.2 μM, and K(12Lf) = 11 ± 2 μM. Two slightly different interactions of T or TFe(2) with Cp are detected for the first time. They are both independent of pH and μ and occur with 1:1 stoichiometry: K(1T) = 19 ± 7 μM, and K(1TFe2) = 12 ± 4 μM. These results can improve our understanding of the probable process of the transfer of iron from Cp to T in iron and copper transport and homeostasis.     

Comparative oxidations of tyrosines and methionines in transferrins: human serum transferrin, human lactotransferrin, and chicken ovotransferrin

Periodate treatments of apo human serum transferrin (HST), and apo chicken ovotransferrin (COT) were previously reported to cause a rapid loss of Fe+3 binding capacity, with a loss of 3 to 5 tyrosine residues [P. AZARI AND J. L. PHILLIPS (1970) Arch. Biochem. Biophys. 138, 32-38; K. F. GEOGHEGAN, J. L. DALLAS, AND R. E. FEENEY (1980) J. Biol. Chem. 255, 11429-11434]. The effects of periodate and hydrogen peroxide on human lactotransferrin (HLT), HST, and COT have been compared. All three apotransferrins were rapidly inactivated and lost approximately 4 to 5 tyrosine residues by 5 mM periodate treatment; their iron complexes had little or no inactivation and losses of approximately 1 to 2 tyrosine residues. All three iron transferrins were highly resistant to inactivation by 5 mM periodate in bicarbonate, with or without the addition of phosphate, while in phosphate (with ambient carbonate) Fe2HLT was highly resistant, Fe2COT slightly less resistant, and Fe2HST much less resistant. Similar oxidations of methionines to the sulfoxides were found in both the apo and iron forms. After 150 min of 5 mM periodate treatment HST lost approximately 3 (apo 3.1, iron 2.8) of 9, HLT approximately 3 (apo 2.6, iron 2.9) of 6, and COT approximately 7 (apo 7.2, iron 7.2) of 11 methionines per mole of protein. In the presence of 8 M urea HST had essentially all of its methionine residues oxidized by periodate, but only lost part of its activity on renaturation. Treatment of all apo transferrins with 300 mM hydrogen peroxide resulted in little or no losses (less than 10%) in activity. HST lost approximately one-third of its methionines and no tyrosines during the 300 mM hydrogen peroxide treatment. Therefore the essentiality of tyrosines for all three transferrins was confirmed and the nonessentiality of methionines was demonstrated.     

Cleavage of human serum transferrin with N-bromosuccinimide.

Human serum transferrin was fragmented by N-bromosuccinimide and reduction-alkylation. It was observed that there were at least two each of tryptophanyl-serine and tryptophanyl-aspartic acid, and one each of tryptophanyl-alanine and tryptophanyl-glutamic acid bonds. The size of fragments detected by polyacrylamide gel electrophoresis ranged from 8,000 to 70,000 daltons. Several of the fragments were isolated in a homogeneous form with respect to molecular weight, but were shown to be mixtures of at least five molecular species each by end group analysis.     

The effect of serum and experimental variables on the transferrin and reticulocyte interaction.

The transfer of iron from transferrin to the developing erythrocyte is a research area of high interest and considerable controversy. We have found that the results of transferrin-reticulocyte incubation studies are quite sensitive to the experimental procedures that are utilized. Reticulocytosis has been induced in rabbits by phelbotomy and phenylhydrazine injections. While the latter gives a higher reticulocyte count, the cells appear to exhibit an altered transferrin-membrane interaction. Transferrin has been iodinated by published methods utilizing chloramine-T and molecular iodine. The iodotransferrin products exhibit the same iron donation ability, however, evidence was found that the chloramine-T treatment leads to a nonspecific binding of transferrin to the reticulocyte. The means of saturating transferrin with 59Fe is also of prime importance. Fe(NH4)2(SO4)2 and especially FeCl3 were found to yield nonspecifically bound iron when added to transferrin or serum. This artifact was reflected in an altered transferrin-reticulocyte interaction. Using what we believe to be optimal conditions, the effect of serum on the transferrin-reticulocyte system was re-examined. The results clearly indicated an enhancement of iron uptake by reticulocytes in the presence of serum, as well as an accelerated incorporation of iron by the cytoplasmic fraction.     

Alternative structural state of transferrin. The crystallographic analysis of iron-loaded but domain-opened ovotransferrin N-lobe

Transferrins bind Fe3+ very tightly in a closed interdomain cleft by the coordination of four protein ligands (Asp60, Tyr92, Tyr191, and His250 in ovotransferrin N-lobe) and of a synergistic anion, physiologically bidentate CO32-. Upon Fe3+ uptake, transferrins undergo a large scale conformational transition: the apo structure with an opening of the interdomain cleft is transformed into the closed holo structure, implying initial Fe3+ binding in the open form. To solve the Fe3+-loaded, domain-opened structure, an ovotransferrin N-lobe crystal that had been grown as the apo form was soaked with Fe3+-nitrilotriacetate, and its structure was solved at 2.1 A resolution. The Fe3+-soaked form showed almost exactly the same overall open structure as the iron-free apo form. The electron density map unequivocally proved the presence of an iron atom with the coordination by the two protein ligands of Tyr92-OH and Tyr191-OH. Other Fe3+ coordination sites are occupied by a nitrilotriacetate anion, which is stabilized through the hydrogen bonds with the peptide NH groups of Ser122, Ala123, and Gly124 and a side chain group of Thr117. There is, however, no clear interaction between the nitrilotriacetate anion and the synergistic anion binding site, Arg121.     

The interaction of prostaglandins with human serum lipoproteins

Using high density and low density lipoproteins (HDL and LDL) labeled with fluorescent analogues of phosphatidylcholine or sphingomyelin it was found that low amounts (10^–12 M) of prostaglandins E₁ and F₂ induced different structural rearrangements of the lipoprotein surface, whereas prostaglandins E₂ and F₁ had no effect. The effects of prostaglandin E₁ on HDL were largely paralled by those of this prostaglandin on synthetic recombinants prepared from pure apolipoprotein A₁, phospholipids and cholesterol and were demonstrated to be caused by prostaglandin-apolipoprotein interaction. The interaction resembled that of a ligand with a specific receptor protein because it was specific, reversible, concentration and temperature dependent and saturable. However the retaining capacity of HDL or LDL for prostaglandin E₁ as determined by equilibrium dialysis was very low and a single prostaglandin E₁ molecule was able to induce structural changes in large numbers of discrete lipoprotein particles. To explain this remarkable fact a non-equilibrium model of ligand-receptor interaction is proposed. According to that model in open systems characterized by weak ligand-receptor binding, high diffusion rate of the ligand and long relaxation times which exceed the interval between two successive receptor occupations, the ligand-induced changes will accumulate, resulting in transformation of the system into a new state which may be far away from equilibrium. It is emphasized that the low mobility of lipids constituting the environment of the receptor protein plays a critcal role in this type of signal amplification.It was further demonstrated that the PGE₁-induced changes of the lipoprotein surface resulted in an enhancement of LDL-to-HDL transfer of cholesterol esters and phosphatidylcholine especially in the presence of serum lipid transfer proteins. The acceleration of the interlipoprotein transfer caused by prostaglandin E₁ in turn increases the rate of cholesterol esterification in serum. It is suggested that in such a way prostaglandin E₁ may influence the homeostasis of cholesterol.Abbreviations LDL low density lioproteins - HDL high density lipoproteins - PG prostaglandin - ASM anthrylvinyl-labeled sphingomyelin (N-12-(9-anthryl)-11-trans-dodecanoylsphingosin-1-phosphocholine - APC anthrylvinylphosphatidylcholine (1-radyl-2-[(9-anthryl)-11-transdodecanoyl)-sn-glycerophosphocholine - NAP-SM nitroazidophenyl labeled sphingomyelin (N-[N-(2-nitro-4azidophenyl)-12-aminododecanoyl]-sphingosin-1-phosphocholine) - NAP-PC adizophenyl labeled phosphatidylcholine (1-radyl-2-[N-(2-nitro-4azidophenyl)-12-aminododecanoyl]-sn-glycero-3-phosphocholine - DPPC dipalmitoylphosphatidylcholine - P fluorescence polarization - E parameter of tryptophanyl to ASM resonance energy transfer - LEP lipid-exchange protein     

Production of human serum transferrin in Escherichia coli.

Transferrin (Tf) crystals diffract to only medium resolution. The mediocre quality of the crystals may be due to two factors: (1) the genetic variations naturally present in the primary sequence of Tf, and (2) the glycosylation of the protein. To control genetic variations and glycosylation of samples of Tf, it would be desirable to express the Tf gene from a recombinant clone. Additionally, expression of Tf from a clone would allow for manipulation of the structure of Tf. The cDNA encoding Tf has been cloned into the pL-based expression vector, pRE1, and the T7-based expression vectors, pRSETA and pET11A. The Tf expression plasmids, pTF-SSn and pTF-ESn, based on the T7 expression vectors, efficiently produce a 76-kDa protein that is approximately the same size as deglycosylated Tf, cross reacts with anti-Tf antibodies, and matches the deduced N-terminal amino acid sequence. Expression of Tf in Escherichia coli will allow the production of genetically pure, unglycosylated protein.     

Interaction of VO ion with human serum transferrin and albumin

The complexation of VO²⁺ ion with the high molecular mass components of the blood serum, human serum transferrin (hTf) and albumin (HSA), has been re-examined using EPR spectroscopy. In the case of transferrin, the results confirm those previously obtained, showing that VO²⁺ ion occupies three different binding sites, A, B₁ and B₂, distinguishable in the X-band anisotropic spectrum recorded in D₂O. With albumin the results show that a dinuclear complex (VO)₂^dHSA is formed in equimolar aqueous solutions or with an excess of protein; in the presence of an excess of VO²⁺, the multinuclear complex (VO)_x^mHSA is the prevalent species, where x = 5-6 indicates the equivalents of metal ion coordinated by HSA. The structure of the dinuclear species is discussed and the donor atoms involved in the metal coordination are proposed on the basis of the measured EPR parameters. Two different binding modes of albumin can be distinguished varying the pH, with only one species being present at the physiological value. The results show that the previously named “strong” site is not the N-terminal copper binding site, and some hypothesis on the metal coordination is discussed, with the ⁵¹V A_z values for the proposed donor sets obtained by DFT (density functional theory) calculations. Finally, preliminary results obtained in the ternary system VO²⁺/hTf/HSA are shown in order to determine the different binding strength of the two proteins. Due to the low VO²⁺ concentration used, the recording of the EPR spectra through the repeated acquisition of the weak signals is essential to obtain a good signal to noise ratio in these systems.     

Composition of pH-sensitive triad in C-lobe of human serum transferrin. Comparison to sequences of ovotransferrin and lactoferrin provides insight into functional differences in iron release

The transferrins (TF) are a family of bilobal glycoproteins that tightly bind ferric iron. Each of the homologous N- and C-lobes contains a single iron-binding site situated in a deep cleft. Human serum transferrin (hTF) serves as the iron transport protein in the blood; circulating transferrin binds to receptors on the cell surface, and the complex is internalized by endocytosis. Within the cell, a reduction in pH leads to iron release from hTF in a receptor-dependent process resulting in a large conformational change in each lobe. In the hTF N-lobe, two critical lysines facilitate this pH-dependent conformational change allowing entry of a chelator to capture the iron. In the C-lobe, the lysine pair is replaced by a triad of residues: Lys534, Arg632, and Asp634. Previous studies show that mutation of any of these triad residues to alanine results in significant retardation of iron release at both pH 7.4 and pH 5.6. In the present work, the role of the three residues is probed further by conversion to the residues observed at the equivalent positions in ovotransferrin (Q-K-L) and human lactoferrin (K-N-N) as well as a triad with an interchanged lysine and arginine (K534R/R632K). As expected, all of the constructs bind iron and associate with the receptor with nearly the same K(D) as the wild-type monoferric hTF control. However, interesting differences in the effect of the substitutions on the iron release rate in the presence and absence of the receptor at pH 5.6 are observed. Additionally, titration with KCl indicates that position 632 must have a positively charged residue to elicit a robust rate acceleration as a function of increasing salt. On the basis of these observations, a model for iron release from the hTF C-lobe is proposed. These studies provide insight into the importance of charge and geometry of the amino acids at these positions as a partial explanation for differences in behavior of individual TF family members, human serum transferrin, ovotransferrin, and lactoferrin. The studies collectively highlight important features common to both the N- and C-lobes of TF and the critical role of the receptor in iron release.