Crystal structures and iron release properties of mutants (K206A and K296A) that abolish the dilysine interaction in the N-lobe of human transferrin.

Human transferrin (Tf) is responsible for the binding and transport of iron in the bloodstream of vertebrates. Delivery of this bound iron to cells occurs by a process of receptor-mediated endocytosis during which Tf releases its iron at the reduced endosomal pH of approximately 5.6. Iron release from Tf involves a large conformational change in which the two domains that enclose the binding site in each lobe move apart. We have examined the role of two lysines, Lys206 and Lys296, that form a hydrogen-bonded pair close to the N-lobe binding site of human Tf and have been proposed to form a pH-sensitive trigger for iron release. We report high-resolution crystal structures for the K206A and K296A mutants of the N-lobe half-molecule of Tf, hTf/2N, and quantitative iron release data on these mutants and the double mutant K206A/K296A. The refined crystal structures (for K206A, R = 19.6% and R(free) = 23.7%; for K296A, R= 21.2% and R(free) = 29.5%) reveal a highly conserved hydrogen bonding network in the dilysine pair region that appears to be maintained even when individual hydrogen bonding groups change. The iron release data show that the mutants retain iron to a pH 1 unit lower than the pH limit of wild type hTf/2N, and release iron much more slowly as a result of the loss of the dilysine interaction. Added chloride ions are shown to accelerate iron release close to the pH at which iron is naturally lost and the closed structure becomes destabilized, and to retard it at higher pH.

The chloride effect is related to anion binding in determining the rate of iron release from the human transferrin N-lobe.

The major function of human transferrin is to deliver iron from the bloodstream to actively dividing cells. Upon iron release, the protein changes its conformation from 'closed' to 'open'. Extensive studies in vitro indicate that iron release from transferrin is very complex and involves many factors, including pH, the chelator used, an anion effect, temperature, receptor binding and intra-lobe interactions. Our earlier work [He, Mason and Woodworth (1997) Biochem. J. 328, 439-445] using the isolated transferrin N-lobe (recombinant N-lobe of human transferrin comprising residues 1-337; hTF/2N) has shown that anions and pH modulate iron release from hTF/2N in an interdependent manner: chloride retards iron release at neutral pH, but accelerates the reaction at acidic pH. The present study supports this idea and further details the nature of the dual effect of chloride: the anion effect on iron release is closely related to the strength of anion binding to the apoprotein. The negative effect seems to originate from competition between chloride and the chelator for an anion-binding site(s) near the metal centre. With decreasing pH, the strength of anion binding to hTF/2N increases linearly, decreasing the contribution of competition with the chelator. In the meantime, the 'open' or 'loose' conformation of hTF/2N, induced by the protonation of critical residues such as the Lys-206/Lys-296 pair at low pH, enables chloride to enter the cleft and bind to exposed side chains, thereby promoting cleft opening and synergistically allowing removal of iron by the chelator, leading to a positive anion effect. Disabling one or more of the primary anion-binding residues, namely Arg-124, Lys-206 and Lys-296, substantially decreases the anion-binding ability of the resulting mutant proteins. In these cases, the competition for the remaining binding residue(s) is increased, leading to a negative chloride effect or, at most, a very small positive effect, even at low pH.

Crystal structures of two mutants (K206Q, H207E) of the N-lobe of human transferrin with increased affinity for iron.

The X-ray crystallographic structures of two mutants (K206Q and H207E) of the N-lobe of human transferrin (hTF/2N) have been determined to high resolution (1.8 and 2.0 A, respectively). Both mutant proteins bind iron with greater affinity than native hTF/2N. The structures of the K206Q and H207E mutants show interactions (both H-bonding and electrostatic) that stabilize the interaction of Lys296 in the closed conformation, thereby stabilizing the iron bound forms.

Mutations at the histidine 249 ligand profoundly alter the spectral and iron-binding properties of human serum transferrin N-lobe.

Human serum transferrin is an iron-binding and -transport protein which carries iron from the blood stream into various cells. Iron is held in two deep clefts located in the N- and C-lobes by coordinating to four amino acid ligands, Asp 63, Tyr 95, Tyr 188, and His 249 (N-lobe numbering), and to two oxygens from carbonate. We have previously reported the effect on the iron-binding properties of the N-lobe following mutation of the ligands Asp 63, Tyr 95, and Tyr 188. Here we report the profound functional changes which result from mutating His 249 to Ala, Glu, or Gln. The results are consistent with studies done in lactoferrin which showed that the histidine ligand is critical for the stability of the iron-binding site [H. Nicholson, B. F. Anderson, T. Bland, S. C. Shewry, J. W. Tweedie, and E. N. Baker (1997) Biochemistry 36, 341-346]. In the mutant H249A, the histidine ligand is disabled, resulting in a dramatic reduction in the kinetic stability of the protein toward loss of iron. The H249E mutant releases iron three times faster than wild-type protein but shows significant changes in both EPR spectra and the binding of anion. This appears to be the net effect of the metal ligand substitution from a neutral histidine residue to a negative glutamate residue and the disruption of the "dilysine trigger" [MacGillivray, R. T. A., Bewley, M. C., Smith, C. A., He, Q.-Y., Mason, A. B., Woodworth, R. C., and Baker, E. N. (2000) Biochemistry 39, 1211-1216]. In the H249Q mutant, Gln 249 appears not to directly contact the iron, given the similarity in the spectroscopic properties and the lability of iron release of this mutant to the H249A mutant. Further evidence for this idea is provided by the preference of both the H249A and H249Q mutants for nitrilotriacetate rather than carbonate in binding iron, probably because NTA is able to provide a third ligation partner. An intermediate species has been identified during the kinetic interconversion between the NTA and carbonate complexes of the H249A mutant. Thus, mutation of the His 249 residue does not abolish iron binding to the transferrin N-lobe but leads to the appearance of novel iron-binding sites of varying structure and stability.

Mutation of the iron ligand His 249 to Glu in the N-lobe of human transferrin abolishes the dilysine "trigger" but does not significantly affect iron release.

Serum transferrin is the major iron transport protein in humans. Its function depends on its ability to bind iron with very high affinity, yet to release this bound iron at the lower intracellular pH. Possible explanations for the release of iron from transferrin at low pH include protonation of a histidine ligand and the existence of a pH-sensitive "trigger" involving a hydrogen-bonded pair of lysines in the N-lobe of transferrin. We have determined the crystal structure of the His249Glu mutant of the N-lobe half-molecule of human transferrin and compared its iron-binding properties with those of the wild-type protein and other mutants. The crystal structure, determined at 2.4 A resolution (R-factor 19.8%, R(free) 29.4%), shows that Glu 249 is directly bound to iron, in place of the His ligand, and that a local movement of Lys 296 has broken the dilysine interaction. Despite the loss of this potentially pH-sensitive interaction, the H249E mutant is only slightly more acid-stable than wild-type and releases iron slightly faster. We conclude that the loss of the dilysine interaction does make the protein more acid stable but that this is counterbalanced by the replacement of a neutral ligand (His) by a negatively charged one (Glu), thus disrupting the electroneutrality of the binding site.

Identification of platination sites on human serum transferrin using (13)C and (15)N NMR spectroscopy.

Reactions between various apo and metal-bound forms of human serum transferrin (80 kDa) and the recombinant N-lobe (40 kDa) with [Pt(en)Cl(2)] or cis-[PtCl(2)(NH(3))(2)] have been investigated in solution via observation of [(1)H,(15)N] NMR resonances of the Pt complexes, [(1)H,(13)C] resonances of the eCH(3) groups of the protein methionine residues, and by chromatographic analysis of single-site methionine mutants. For the whole protein, the preferred Pt binding site appears to be Met256. Additional binding occurs at the other surface-exposed methionine (Met499), which is platinated at a slower rate than Met256. In contrast, binding of similar Pt compounds to the N-lobe of the protein occurs at Met313, rather than Met256. Met313 is buried in the interlobe contact region of intact transferrin. After loss of one chloride ligand from Pt and binding to methionine sulfur of the N-lobe, chelate-ring closure appears to occur with binding to a deprotonated backbone amide nitrogen, and the loss of the other chloride ligand. Such chelate-ring closure was not observed during reactions of the whole protein, even after several days.

Dual role of Lys206-Lys296 interaction in human transferrin N-lobe: iron-release trigger and anion-binding site.

The unique structural feature of the dilysine (Lys206-Lys296) pair in the transferrin N-lobe (hTF/2N) has been postulated to serve a special function in the release of iron from the protein. These two lysines, which are located in opposite domains, hydrogen bond to each other in the iron-containing hTF/2N at neutral pH but are far apart in the apo-form of the protein. It has been proposed that charge repulsion resulting from the protonation of the dilysines at lower pH may be the trigger to open the cleft and facilitate iron release. The fact that the dilysine pair is positively charged and resides in a location close to the metal-binding center has also led to the suggestion that the dilysine pair is an anion-binding site for chelators. The present report provides comprehensive evidence to confirm that the dilysine pair plays this dual role in modulating release of iron. When either of the lysines is mutated to glutamate or glutamine or when both are mutated to glutamate, release of iron is much slower compared to the wild-type protein. This is due to the fact that the driving force for cleft opening is absent in the mutants or is converted to a lock-like interaction (in the case of the K206E and K296E mutants). Direct titration of the apo-proteins with anions as well as anion-dependent iron release studies show that the dilysine pair is part of an active anion-binding site which exists with the Lys296-Tyr188 interaction as a core. At this site, Lys296 serves as the primary anion-binding residue and Tyr188 is the main reporter for electronic spectral change, with smaller contributions from Lys206, Tyr85, and Tyr95. In iron-loaded hTF/2N, anion binding becomes invisible as monitored by UV-vis difference spectra since the spectral reporters Tyr188 and Tyr95 are bound to iron. Our data strongly support the hypothesis that the apo-hTF/2N exists in equilibrium between the open and closed conformations, because only in the closed form is Lys296 in direct contact with Tyr188. The current findings bring together observations, ideas, and experimental data from a large number of previous studies and shed further light on the detailed mechanism of iron release from the transferrin N-lobe. In iron-containing hTF/2N, Lys296 may still function as a target to introduce an anion (or a chelator) near to the iron-binding center. When the pH is lowered, the protonation of carbonate (synergistic anion for metal binding) and then the dilysine pair form the driving force to loosen the cleft, exposing iron; the nearby anion (or chelator) then binds to the iron and releases it from the protein.

X-ray crystallography and mass spectroscopy reveal that the N-lobe of human transferrin expressed in Pichia pastoris is folded correctly but is glycosylated on serine-32.

The ferric form of the N-lobe of human serum transferrin (Fe(III)-hTF/2N) has been expressed at high levels in Pichia pastoris. The Fe(III)-hTF/2N was crystallized in the space group P41212, and X-ray crystallography was used to solve the structure of the recombinant protein at 2.5 A resolution. This represents only the second P. pastoris-derived protein structure determined to date, and allows the comparison of the structures of recombinant Fe(III)-hTF/2N expressed in P. pastoris and mammalian cells with serum-derived transferrin. The polypeptide folding pattern is essentially identical in all of the three proteins. Mass spectroscopic analyses of P. pastoris- hTF/2N and proteolytically derived fragments revealed glycosylation of Ser-32 with a single hexose. This represents the first localization of an O-linked glycan in a P. pastoris-derived protein. Because of its distance from the iron-binding site, glycosylation of Ser-32 should not affect the iron-binding properties of hTF/2N expressed in P. pastoris, making this an excellent expression system for the production of hTF/2N.

Two high-resolution crystal structures of the recombinant N-lobe of human transferrin reveal a structural change implicated in iron release.

The N-lobe of human serum transferrin (hTF/2N) has been expressed in baby hamster kidney cells and crystallized in both orthorhombic (P212121) and tetragonal (P41212) space groups. Both crystal forms diffract to high resolution (1.6 and 1.8 A, respectively) and have been solved by molecular replacement. Subsequent refinement resulted in final models for the structure of hTF/2N that had crystallographic R-factors of 18.1 and 19.7% for the two crystal forms, respectively; these models represent the highest-resolution transferrin structures determined to date. The hTF/2N polypeptide has a folding pattern similar to those of other transferrins, including the presence of a deep cleft that contains the metal-binding site. In contrast to other transferrins, both crystal forms of hTF/2N display disorder at the iron-binding site; model building suggests that this disorder consists of alternative conformations of the synergistically bound carbonate anion, the side chain for Arg-124, and several solvent molecules. Subsequent refinement revealed that conformation A has an occupancy of 0.63-0. 65 and corresponds to the structure of the iron-binding site found in other transferrins. The alternative conformation B has an occupancy of 0.35-0.37; in this structure, the carbonate has rotated 30 degrees relative to the iron and the side chain for Arg-124 has moved to accommodate the new carbonate position. Several water molecules appear to stabilize the carbonate anion in the two conformations. These structures are consistent with the protonation of the carbonate and resulting partial removal of the anion from the metal; these events would occur prior to cleft opening and metal release.

[1H,13C] NMR determination of the order of lobe loading of human transferrin with iron: comparison with other metal ions.

Human serum transferrin (hTF) is a single-chain bilobal glycoprotein (80 kDa) which transports Fe3+ and a variety of other metal ions in blood. Only diferric transferrin, not the apo-protein, binds strongly to transferrin receptors and is taken up by cells via receptor-mediated endocytosis. We show here that 2D [1H,13C] NMR studies of recombinant epsilon-[13C]Met-hTF allow the order of lobe loading with various metal ions, including Fe3+, to be determined. In particular, the resonance for Met-464, a residue in the hydrophobic patch of helix 5, is very sensitive to iron binding in the C-lobe. The selectivity of lobe loading with Fe3+ is compared to loading with Fe2+ (which binds as Fe3+), Al3+, Ga3+ and Bi3+. Similar changes in shifts of the Met residues are observed for these metal ions, suggesting that they induce similar conformational changes in the protein.

Iron release from recombinant N-lobe and single point Asp63 mutants of human transferrin by EDTA.

Transferrins bind ferric ion and deliver the iron to cells. The mechanism of the iron release has been studied kinetically, in vitro, with the aid of single point mutants in which the iron-binding ligand, Asp63 (aspartic acid-63, D63), has been changed to Ser, Asn, Glu and Ala. Iron release from the unmutated N-lobe of human serum transferrin (hTF/2N) by EDTA is influenced by a variety of factors. The rate-determining conformational-change mechanism may be a major pathway for iron release from hTF/2N's having a 'closed' conformation, which leads to a saturation kinetic mode with respect to ligand concentration. The effect of chloride depends on the protein conformation, showing a negative action in the case of tight binding and a positive action when the protein has an 'open' or 'loose' conformation. The negative effect of chloride could originate from the binding competition between chloride and the chelate to the active site for iron release, and the positive effect could derive from the synergistic participation of chloride in iron removal. The 'open' conformation may be induced by decreasing pH: the transitional point appears to be at about pH 6.3 for the wild-type hTF/2N; the 'loose' conformation may be facilitated by mutations at D63, which result in the loss of a key linking component in interdomain interactions of the protein. In the latter case, structural factors dominate over other potential negative effects because the weak interdomain contacts derived from the mutation of D63 cause the binding site to open easily, even at pH 7.4. Therefore chloride exhibits an accelerating action on iron release by EDTA from all the D63 mutants.

The effect of salt and site-directed mutations on the iron(III)-binding site of human serum transferrin as probed by EPR spectroscopy.

The effects of site-directed mutation and salt on the iron(III)-binding site of the recombinant half-molecule of the N-terminal lobe (hTf/2N) of human transferrin was studied by EPR spectroscopy. Changes were observed in the EPR spectra of all variants investigated (D63S, D63C, G65R, K206Q, H207E, H249E, H249Q, K296E and K296Q) compared with that of the wild-type protein. The most pronounced changes in the metal site were caused by replacement of the coordinating residues, Asp-63 and His-249, and the non-coordinating residue Lys-296, which is located in the hinge region of the iron-binding cleft. The EPR spectral changes from replacement of other non-coordinating residues were more subtle, indicating small changes in Fe3+ coordination to the protein. The EPR spectrum of variant G65R suggests that it adopts two distinct conformations in solution, one in which the two domains forming the iron-binding cleft are closed and one in which they are open; in the latter instance Asp-63 is no longer coordinated to the Fe3+. Chloride-binding studies on hTf/2N, K206Q, H207E, K296Q and K296E showed similar binding isotherms, indicating that none of the hinge region residues replaced, i.e. Lys-206, His-207 or Lys-296, are the sites of chloride binding. The results show that the coordination environment of the Fe3+ is sensitive to structural changes from site-directed mutation of both remote and coordinated residues and also to chloride-binding and ionic strength effects.

Expression and initial characterization of five site-directed mutants of the N-terminal half-molecule of human transferrin.

Five site-directed mutants of the N-terminal half-molecule of human serum transferrin have been expressed in baby hamster kidney cells and purified to homogeneity. Expression levels and overall yields varied considerably from the wild-type protein, depending on the mutant in question. The mutants are D63S, D63C, G65R, K206Q, and H207E and are based on mutations observed in a variety of transferrins of known sequence. Their molecular masses, determined by electrospray mass spectrometry, agree with theory, except for the D63C mutant, which appears to be cysteinylated. All mutants bind iron but with varying affinities; qualitatively, in increasing order D63S approximately D63C approximately G65R much less than wild type less than or equal to H207E much less than K206Q. In general, reduction of formal negative charge within the binding cleft shifts the visible spectral maximum of the iron complex toward the blue and reduces the affinity for iron, and increasing the formal negative charge shifts the visible maximum toward the red and increases the affinity for iron. The K206Q mutant is exceptional inasmuch as its visible maximum shows a blue shift, but its affinity for iron is the greatest of all of the mutants studied. All mutants reported, in addition to the wild-type protein, exhibit very similar visible molar extinction coefficients for the iron complex and very similar changes in extinction coefficients at 240 nm on binding Fe(III) or Ga(III). These results suggest that in all cases the bound metal ion is coordinated by two tyrosyl side chains.

Mössbauer spectroscopy of iron-ovotransferrin: a crystal field interpretation.

Mössbauer spectra from frozen solutions of ovotransferrin were recorded in a variety of applied external magnetic fields and at various temperatures in a small applied field. The results were fitted to a simple model for the electronic structure at the iron site. This model requires admixtures of the free ion 6S and 4P states, indicating a weak cubic crystal field. Possible implications of this model regarding the binding site are discussed.

Electrophoretic characterization of human, equine and bovine transferrins.

1. Human, bovine and equine transferrins have been characterized with respect to mol. wt, and behavior on urea-polyacrylamide gels, and isoelectric focussing gels. 2. As shown by SDS-polyacrylamide gel electrophoresis human transferrin has one major polypeptide whereas both bovine and equine transferrins have two polypeptides. 3. The transferrins show multiple banded patterns on urea-polyacrylamide and isoelectric focussing gels, particularly when iron saturated. The various forms are not resolved by neuraminidase treatment.

The synthesis and turnover of ferritin in rat L-6 cells. Rates and response to iron, actinomycin D, and desferrioxamine.

Rates of ferritin accumulation in the L-6 line of rat skeletal myoblasts cultured in the presence of ferric nitrilotriacetate (FeNTA) were measured and found to vary with the extracellular concentration of FeNTA as predicted from dose response experiments. The rate of ferritin accumulation is constant for up to 92 h in these cells after addition of iron, with the exception of the first few hours of synthesis in which the rate is approximately twice that observed at later times. Experiments in which the specific activity of newly synthesized ferritin was calculated implied that the rate of ferritin degradation might be higher in this earlier period as well; pulse-chase experiments confirmed this hypothesis. Ferritin synthesis is thought to be induced by iron in the absence of RNA synthesis. Accordingly, actinomycin D was shown not to inhibit the synthesis of ferritin in response to FeNTA. Rather, a pronounced stimulation of the synthetic rate was observed. Finally, desferrioxamine was shown both to decrease the rate of ferritin synthesis and increase its rate of degradation. Possible mechanisms for these phenomena are discussed.

Physiological levels of binding and iron donation by complementary half-molecules of ovotransferrin to transferrin receptors of chick reticulocytes.

Two fragments, each corresponding to approximately half of the ovotransferrin (OTf) molecule and containing an iron-binding site were produced by digestion with affinity bound trypsin and were purified by isoelectric focusing and gel filtration chromatography. The immunologically distinct "half-molecules" individually have little ability to bind to transferrin receptors on chick embryo red blood cells or to donate iron to them. Combining them, however, leads to both binding and iron donation approaching that found for holo-OTf. Furthermore, similar amounts of radiolabeled iron can be extracted into the putative heme fraction from Fe2OTf and from the various combined half-molecules. These findings conflict with those reported by Keung and Azari ( (1982) J. Biol. Chem. 257, 1184-1188) for subtilisin-derived half-molecules of OTf examined in a similar system. They found that each half-molecule appeared to bind at a level of approximately one-third that of Fe2OTf and that the half-molecules competed with each other for binding sites. In contrast, our equilibrium binding studies, in the presence of 2,4-dinitrophenol to prevent iron removal, led to the determination of 4.79 X 10(4) binding sites/cell for Fe2OTf, 4.44 X 10(4) for the NH2-terminal half-molecules in the presence of excess COOH-terminal half-molecules and 4.17 X 10(4) for COOH-terminal half-molecules in the presence of NH2-terminal half-molecules; apparent binding constants were estimated to be 3.29 X 10(6), 1.19 X 10(6), and 0.67 X 10(6) M-1 for these same samples. Problems associated with equilibrium binding studies in which a narrow range of concentrations of ligand is used and/or iron is being removed are discussed. Labeled combined half-molecules were half as effective as labeled Fe2OTf in competition with unlabeled Fe2OTf. These findings are consistent with the lower apparent binding constant found in the equilibrium binding studies. Equimolar apo-OTf had no effect on binding of either Fe2OTf or the combined half-molecules. It seems apparent from our studies that the NH2- and COOH-terminal half-molecules each contain a recognition region both of which are necessary for binding to the transferrin receptor and iron donation to the chick embryo red blood cell.

Ferritin synthesis in rat L-6 cells: response to iron challenge.

The short term response of the L-6 cell line of rat skeletal myoblasts to elevated extracellular iron concentrations was studied. It was found in all cases that iron as the nitrilotriacetate (NTA) chelate was effective at donating iron to the cells and at stimulating ferritin synthesis. After 48 h in 50 microM ferric NTA, the cellular ferritin levels rose from an undetectable level to 1.11 (+/- 0.07) ng ferritin/microgram cell protein, or 0.1% of total cell protein. Similarly, the total iron in the cells rose under the same conditions from an unmeasurable level to plateau at over 10 fmol iron/cell. In addition, it was found that these cells synthesize ferritin in response to iron in a dose-dependent manner over a range of iron concentrations from 5-1000 microM. A sensitive and specific immunoradiometric assay for rat ferritin was used in these studies to quantitate ferritin in cell lysates.

Exocytosis of pinocytosed fluid in cultured cells: kinetic evidence for rapid turnover and compartmentation.

The uptake and fate of pinocytosed fluid were investigated in monolayers of pulmonary alveolar macrophages and fetal lung fibroblasts using the fluid-phase marker, [14C]sucrose. Initial experiments revealed that cellular accumulation of chromatographically repurified [14C]sucrose was not linear with incubation time. Deviation from linearity was shown to be due to constant exocytosis of accumulating marker. Chromatographic analysis revealed that the cells were unable to metabolize sucrose and were releasing it intact by a process that was temperature-sensitive but not dependent on extracellular calcium and magnesium. A detailed analysis of the kinetics of exocytosis was undertaken by preloading cells with [14C]sucrose for various lengths of time and then monitoring the appearance of radioactivity into isotope-free medium. Results indicated that modeling the process of fluid-phase pinocytosis and subsequent exocytosis required at least two intracellular compartments in series, one compartment being of small size and turning over very rapidly (t1/2 = 5 min in macrophages, 6--8 min in fibroblasts) and the other compartment being apparently larger in size and turning over very slowly (t1/2 = 180 min in macrophages, 430--620 min in fibroblasts). Computer-simulation based on this model confirmed that the kinetics of efflux faithfully reflected the kinetics of influx and that the rate of efflux completely accounted for the deviation from linearity of accumulation kinetics. Moreover, the sizes of the compartments and magnitude of the intercompartment fluxes were such that the majority of fluid internalized in pinocytic vesicles was rapidly returned to the extracellular space via exocytosis. This result provides direct experimental evidence for a process previously thought necessary based solely on morphological and theoretical considerations. Furthermore, the turnover of pinocytosed fluid was so dynamic that accumulation deviated from linearity even within the first few minutes of incubation. We were able to show that the kinetics of exocytosis allowed calculation of the actual pinocytic rate, a rate that was nearly 50% greater than the apparent initial rate obtained from the slope of the uptake curve over the first 10 min.