Recombinant human transferrin: beyond iron binding and transport

Iron is indispensible for life and essential for such processes as oxygen transport, electron transfer and DNA synthesis. Transferrin (Tf) is a ubiquitous protein with a central role in iron transport and metabolism. There is evidence, however, that Tf has many other biological roles in addition to its primary function of facilitating iron transport and metabolism, such as its profound effect on mammalian cell growth and productivity. The multiple functions of Tf can be exploited to develop many novel applications. Indeed, over the past several years, considerable efforts have been directed towards exploring human serum Tf (hTf), especially the use of recombinant native hTf and recombinant Tf fusion proteins, for various applications within biotechnology and medicine. Here, we review some of the remarkable progress that has been made towards the application of hTf in these diverse areas and discuss some of the exciting future prospects for hTf.     

Manganese and iron transport across pulmonary epithelium

Pathways mediating pulmonary metal uptake remain unknown. Because absorption of iron and manganese could involve similar mechanisms, transferrin (Tf) and transferrin receptor (TfR) expression in rat lungs was examined. Tf mRNA was detected in bronchial epithelium, type II alveolar cells, macrophages, and bronchus-associated lymphoid tissue (BALT). Tf protein levels in lung and bronchoalveolar lavage fluid did not change in iron deficiency despite increased plasma levels, suggesting that lung Tf concentrations are regulated by local synthesis in a manner independent of body iron status. Iron oxide exposure upregulated Tf mRNA in bronchial and alveolar epithelium, macrophages, and BALT, but protein was not significantly increased. In contrast, TfR mRNA and protein were both upregulated by iron deficiency. To examine potential interactions with lung Tf, rats were intratracheally instilled with (54)Mn or (59)Fe. Unlike (59)Fe, interactions between (54)Mn and Tf in lung fluid were not detected. Absorption of intratracheally instilled (54)Mn from the lungs to the blood was unimpaired in Belgrade rats homozygous for the functionally defective G185R allele of divalent metal transporter-1, indicating that this transporter is also not involved in pulmonary manganese absorption. Pharmacological studies of (54)Mn uptake by A549 cells suggest that metal uptake by type II alveolar epithelial cells is associated with activities of both L-type Ca(2+) channels and TRPM7, a member of the transient receptor potential melastatin subfamily. These results demonstrate that iron and manganese are absorbed by the pulmonary epithelium through different pathways and reveal the potential role for nonselective calcium channels in lung metal clearance.     

A compartmental model of iron regulation in the mouse

A simple compartmental model is developed for investigating the mechanism of iron homeostasis. In contrast to previous mathematical models of iron metabolism, the liver is included as a key site of iron regulation. Compartments for free iron in blood, diferric transferrin (Tf) in blood, hepatocytes, red blood cells, and macrophages are included, and their roles in iron regulation are explored. The function of hepcidin in regulating iron absorption is modeled through an inverse relationship between hepatocyte transferrin receptor 2 (TfR2) levels and the rate of iron export processes mediated by ferroportin (Fpn). Simulations of anemia and erythropoiesis stimulation support the idea that the iron demands of the erythroid compartment can be communicated through diferric Tf. The iron-responsive element of Fpn is found to be important for stabilizing intracellular iron stores in response to changing iron demands and allowing proper iron regulation through diferric Tf. The contribution of iron dysregulation to the pathogenesis of iron overload disorders is also investigated. It is shown that the characteristics of HFE hemochromatosis can be reproduced by increasing the setpoint of iron absorption in the duodenum to a level where the system cannot downregulate iron absorption to meet the iron excretion rate.     

"Dilysine trigger" in transferrins probed by mutagenesis of lactoferrin: crystal structures of the R210G,R210E,and R210L mutants of human lactoferrin

The mammalian iron-binding proteins lactoferrin (Lf) and transferrin (Tf) bind iron very tightly, but reversibly. Despite homologous structures and essentially identical iron binding sites, Tf begins to release iron at pH 6.0, whereas Lf retains iron to pH approximately 3.5. This difference in iron retention gives the two proteins different biological roles. Two lysine residues, Lys 206 and Lys 296, which form a hydrogen-bonded dilysine pair in human Tf, have been shown to strongly influence iron release from the N-lobe. The equivalent residues in human Lf are Arg 210 and Lys 301, and we have here mutated Arg 210 in the N-lobe half-molecule of human lactoferrin, Lf(N), to probe its role in iron release. The Lf(N) mutants R210G, R210E, and R210L were expressed, purified, and crystallized, and their crystal structures were determined and refined at resolutions of 1.95 A (R210G), 2.2 A (R210E), and 2.0 A (R210L). The overall structures are very similar to that of wild-type Lf(N), but with small differences in domain orientations. In each of the mutants, however, Lys 301 (equivalent to Lys 296 in Tf) changes its conformation to fill the space occupied by Arg 210 Neta2 in wild-type Lf(N), interacting with the two tyrosine ligands Tyr 92 and Tyr 192. By comparison with other Lf and Tf structures, we conclude that Lys 301 (or Lys 296 in Tf) only occupies this site when residue 210 (206 in Tf) is nonpositive (neutral as in R210G and R210L or negative as in R210E). Thus, Lys 206 in the Tf dilysine pair is identified as having a depressed pK(a). Three specific sites are variably occupied by polar groups in the Lf mutants and other Lf and Tf proteins, and when coupled with iron-release data, these give new insights into the factors that most influence iron retention at low pH.     

Neither human hephaestin nor ceruloplasmin forms a stable complex with transferrin

Iron homeostasis is essential for maintaining the physiological requirement for iron while preventing iron overload. Cell toxicity is caused by the generation of hydroxyl-free radicals that result from redox reactions involving Fe(II). Multicopper ferroxidases regulate the oxidation of Fe(II) to Fe(III), circumventing the generation of these harmful by-products. Ceruloplasmin (Cp) is the major multicopper ferroxidase in blood; however, hephaestin (Hp), a membrane-bound Cp homolog, was recently discovered and has been implicated in the export of iron from duodenal enterocytes into blood. In the intracellular milieu, it is likely that iron exists as reduced Fe(II), yet transferrin (Tf), the plasma iron transporter, is only capable of binding oxidized Fe(III). Due to the insoluble and reactive nature of free Fe(III), the oxidation of Fe(II) upon exiting the duodenal enterocyte may require an interaction between a ferroxidase and the iron transporter. As such, it has been suggested that as a means of preventing the release of unbound Fe(III), a direct protein-protein interaction may occur between Tf and Hp during intestinal iron export. In the present study, the putative interaction between Tf and both Cp and a soluble form of recombinant human Hp was investigated. Utilizing native polyacrylamide gel electrophoresis, covalent cross-linking and surface plasmon resonance (SPR), a stable interaction between the two proteins was not detected. We conclude that a stable complex between these ferroxidases and Tf does not occur under the experimental conditions used. We suggest alternative models for loading Tf with Fe(III) during intestinal iron export.     

Indispensability of Iron-bound Chick Transferrin for Chick Myogenesis in Vitro

Myotrophic activity of highly purified chick transferrins (Tfs) to chick primary myogenic cells has been studied in a culture medium containing horse serum. Iron-binding to Tfs is indispensable for the activity. The removal of iron from Tfs gives rise to a complete loss of the activity and it is restored by iron-rebinding depending on the amount of bound iron. This result, combined with other physicochemical and immunological data, strongly, confirms that the myotrophic activity is exerted by the Tfs themselves, not by a contaminating material(s). It has been found that culture medium containing horse Tf which seems inadequate for the study of the biological effects of Tfs is, however, suitable for studies on chick Tfs, since horse Tf is inactive in promoting chick myogenesis. Terminal sialic acid residues are unrelated to myotrophic activity since Tfs with different numbers of residues (0, 1, and 2 moles/Tf molecule) are comprable in their activities. The mechanism of Tf action on cells and contradictions among previous papers as to the requirement of Tf for cell growth have been discussed from the viewpoint of an iron-donor with class-specificity.     

Binding to cellular receptors results in increased iron release from transferrin at mildly acidic pH

In order to better understand the cellular delivery of iron from serum transferrin (Tf), we compared iron release from receptor-bound and free Tf. While free Tf did not release all iron until below pH 4.6, receptor-bound Tf released significantly more iron at mildly acidic pH, with essentially all iron released between pH 5.6 and 6.0. Since Tf is acidified to a minimum pH of 5.4 in K562 cells, this result accounts for the nearly complete extraction of iron from Tf by these cells. Comparison of fluorescence from Tf conjugated with lissamine rhodamine sulfonyl chloride (LRSC-Tf) free in solution and bound to receptor provides further evidence that the Tf receptor modulates low pH-mediated conformational changes in Tf. As pH was decreased from neutrality, the fluorescence of free LRSC-Tf began to increase below pH 6.2; the fluorescence of LRSC-Tf bound to human receptors did not increase until below pH 5.6. Binding to the Tf receptor, while facilitating iron release from Tf, appears to partially inhibit a conformational change that causes the increase in LRSC-Tf fluorescence at low pH. The fluorescence of human LRSC-Tf bound to murine receptors increases at a higher pH, 6.0, indicating that there are differences in conformational stabilization of Tf by receptors of different species. The results suggest that the Tf receptor, in addition to providing a means by which cells may internalize Tf, functions to increase the release of iron from Tf in the endosome.     

Site-specific rate constants for iron acquisition from transferrin by the Aspergillus fumigatus siderophores N′,N′′,N′′′-triacetylfusarinine C and ferricrocin

Aspergillus fumigatus is an opportunistic fungal pathogen that causes life-threatening infections in immunocompromised patients. Despite low levels of free iron, A. fumigatus grows in the presence of human serum in part because it produces high concentrations of siderophores. The most abundant siderophores produced by A. fumigatus are N',N',N'-triacetylfusarinine C (TAF) and ferricrocin, both of which have thermodynamic iron binding constants that theoretically allow them to remove transferrin (Tf)-bound iron. Urea-polyacrylamide gel electrophoresis was used to measure the change in concentration of Tf species incubated with TAF or ferricrocin. The rate of removal of iron from diferric Tf by both siderophores was measured, as were the individual microscopic rates of iron removal from each Tf species (diferric Tf, N-terminal monoferric Tf and C-terminal monoferric Tf). TAF removed iron from all Tf species at a faster rate than ferricrocin. Both siderophores showed a preference for removing C-terminal iron, evidenced by the fact that k(1C) and k(2C) were much larger than k(1N) and k(2N). Cooperativity in iron binding was observed with TAF, as the C-terminal iron was removed by TAF much faster from monoferric than from diferric Tf. With both siderophores, C-terminal monoferric Tf concentrations remained below measurable levels during incubations. This indicates that k(2C) and k(1C) are much larger than k(1N). TAF and ferricrocin both removed Tf-bound iron with second-order rate constants that were comparable to those of the siderophores of several bacterial pathogens, indicating they may play a role in iron uptake in vivo and thereby contribute to the virulence of A. fumigatus.     

Protection of human and murine hepatocytes against Fas-induced death by transferrin and iron

Recent studies in a murine model show that transferrin (Tf) interferes with Fas-mediated hepatocyte death and liver failure by decreasing pro-apoptotic and increasing anti-apoptotic signals. We show here in vitro in murine and human hepatocyte cell lines and in vivo in mice that Fas-induced apoptosis is modulated by exogenous Tf and iron. The results obtained with iron-free Tf (ApoTf), iron-saturated Tf (FeTf), and the iron chelator salicylaldehyde isonicotinoyl hydrazone (SIH) in its iron-free and iron-saturated (FeSIH) forms indicate that apoptosis-modulating effects of Tf are not mediated by iron alone. Both the Tf molecule and iron affect multiple aspects of cell death, and the route of iron delivery to the cell may be critical for the final outcome of cellular Fas signaling. Survival of hepatocytes ‘stressed’ by Fas signals can be manipulated by Tf and iron and may be a target for prophylactic and therapeutic interventions.     

Secreted glyceraldehye-3-phosphate dehydrogenase is a multifunctional autocrine transferrin receptor for cellular iron acquisition

Background

The long held view is that mammalian cells obtain transferrin (Tf) bound iron utilizing specialized membrane anchored receptors. Here we report that, during increased iron demand, cells secrete the glycolytic enzyme glyceraldehyde-3-phosphate dehydrogenase (GAPDH) which enhances cellular uptake of Tf and iron.

Methods

These observations could be mimicked by utilizing purified GAPDH injected into mice as well as when supplemented in culture medium of model cell lines and primary cell types that play a key role in iron metabolism. Transferrin and iron delivery was evaluated by biochemical, biophysical and imaging based assays.

Results

This mode of iron uptake is a saturable, energy dependent pathway, utilizing raft as well as non-raft domains of the cell membrane and also involves the membrane protein CD87 (uPAR). Tf internalized by this mode is also catabolized.

Conclusions

Our research demonstrates that, even in cell types that express the known surface receptor based mechanism for transferrin uptake, more transferrin is delivered by this route which represents a hidden dimension of iron homeostasis.

General significance

Iron is an essential trace metal for practically all living organisms however its acquisition presents major challenges. The current paradigm is that living organisms have developed well orchestrated and evolved mechanisms involving iron carrier molecules and their specific receptors to regulate its absorption, transport, storage and mobilization. Our research uncovers a hidden and primitive pathway of bulk iron trafficking involving a secreted receptor that is a multifunctional glycolytic enzyme that has implications in pathological conditions such as infectious diseases and cancer.     

Structural and functional studies on the stalk of the transferrin receptor

Transferrin (Tf) is an iron carrier protein that consists of two lobes, the N- and C-lobes, which can each bind a Fe³⁺ ion. Tf binds to its receptor (TfR), which mediates iron delivery to cells through an endocytotic pathway. Receptor binding facilitates iron release from the Tf C-lobe, but impedes iron release from the N-lobe. An atomic model of the Tf-TfR complex based on single particle electron microscopy (EM) indicated that receptor binding is indeed likely to hinder opening of the N-lobe, thus interfering with its iron release. The atomic model also suggested that the TfR stalks could form additional contacts with the Tf N-lobes, thus potentially further slowing down its iron release. Here, we show that the TfR stalks are unlikely to make strong interactions with the Tf N-lobes and that the stalks have no effect on iron release from the N-lobes of receptor-bound Tf.     

The C2 variant of human serum transferrin retains the iron binding properties of the native protein

The tryptic digests of blood samples obtained from transferrin C1 and C2 (TfC 1 and TfC2 hereafter) genotypes were analysed by Liquid Chromatography coupled to Electrospray Mass Spectrometry (LC/ESI--MS/MS). The analytical results confirmed the single base change in exon 15 of the Tf gene. The solution behaviour and the iron binding properties of the two Tf variants were studied by UV-visible spectrophotometry and by circular dichroism. It appears that TfC2 globally manifests the same spectral features as the native protein. The local conformation of the two iron binding sites is conserved in the two Tf variants as evidenced by the visible absorption and CD spectra. Also, the iron binding capacities and their pH-dependent profiles are essentially the same. Overall, our investigation points out that the single amino acid substitution in TfC2 (Pro 570 Ser) does not affect the general conformation of the protein nor the local structure of the iron binding sites. The implications of these results for the etiopathogenesis of Alzheimer's disease are discussed.     

Transferrin response in normal and iron-deficient mice heterozygotic for hypotransferrinemia; effects on iron and manganese accumulation

Hypotransferrinemia is a genetic defect in mice resultingin 1% of normal plasma transferrin (Tf) concen-trations;heterozygotes for thismutation (+/hpx) have low circulating Tf concentrations. These mice providea unique opportunity toexamine the developmental pattern and response of Tf to iron-deficient diets, andfurthermore,to address the controversial role of Tf in Mn transport. Twenty-three weanling +/hpx miceandforty-five wild-type BALB/cJ mice were either killed at weaning or fed diets containing either13 or 72 mgkg Fe, and killed after four or eight weeks. Plasma Tfconcentrations were lower in +/hpx mice, plasmaTf nearly doubled and liver Tf was only 50% of normalin response to iron deficiency. Brain iron concen-trationdid not correlate significantlywith either plasma Tf or TIBC. However, iron accumulation into braincontinued with irondeficiency whereas most other organs had less iron. These results imply that eitherthereis a selected targeting of iron to the brain by plasma Tf or there is an alternative irondelivery system tothe brain. Furthermore, we observed no differences in tissuedistribution of Mn despite the differences incirculating Tf concentrationsand body iron stores; this suggests that there are non-Tf dependent mecha-nismsfor Mntransport.     

An iron-dependent and transferrin-mediated cellular uptake pathway for plutonium

Plutonium is a toxic synthetic element with no natural biological function, but it is strongly retained by humans when ingested. Using small-angle X-ray scattering, receptor binding assays and synchrotron X-ray fluorescence microscopy, we find that rat adrenal gland (PC12) cells can acquire plutonium in vitro through the major iron acquisition pathway--receptor-mediated endocytosis of the iron transport protein serum transferrin; however, only one form of the plutonium-transferrin complex is active. Low-resolution solution models of plutonium-loaded transferrins derived from small-angle scattering show that only transferrin with plutonium bound in the protein's C-terminal lobe (C-lobe) and iron bound in the N-terminal lobe (N-lobe) (Pu(C)Fe(N)Tf) adopts the proper conformation for recognition by the transferrin receptor protein. Although the metal-binding site in each lobe contains the same donors in the same configuration and both lobes are similar, the differences between transferrin's two lobes act to restrict, but not eliminate, cellular Pu uptake.     

Structural variations within the transferrin binding site on transferrin-binding protein B, TbpB

Pathogenic bacteria acquire the essential element iron through specialized uptake pathways that are necessary in the iron-limiting environments of the host. Members of the Gram-negative Neisseriaceae and Pasteurellaceae families have adapted to acquire iron from the host iron binding glycoprotein, transferrin (Tf), through a receptor complex comprised of transferring-binding protein (Tbp) A and B. Because of the critical role they play in the host, these surface-exposed proteins are invariably present in clinical isolates and thus are considered prime vaccine targets. The specific interactions between TbpB and Tf are essential and ultimately might be exploited to create a broad-spectrum vaccine. In this study, we report the structure of TbpBs from two porcine pathogens, Actinobacillus pleuropneumoniae and suis. Paradoxically, despite a common Tf target, these swine related TbpBs show substantial sequence variation in their Tf-binding site. The TbpB structures, supported by docking simulations, surface plasmon resonance and hydrogen/deuterium exchange experiments with wild-type and mutant TbpBs, explain why there are structurally conserved elements within TbpB homologs despite major sequence variation that are required for binding Tf.     

Role of transferrin in iron uptake by the brain: a comparative study

Summary The role of specific transferrin (Tf) and Tf receptor interaction on brain capillary endothelial cells in iron transport from the plasma to the brain was investigated by using Tf from several species of animals labeled with ⁵⁹Fe and ¹²⁵I, and 15-day and adult rats. The rate of iron transfer was much greater in the 15-day rats. It was greatest with Tf from the mammals, rat, rabbit and human, but much lower with chicken ovotransferrin and quokka (a marsupial), toad, lizard, crocodile, and fish Tf. The uptake of Tf by the brain showed a similar pattern, except for a very high uptake of ovotransferrin (ovo Tf). Iron uptake by the femurs (a source of bone marrow) was also high with Tf from the mammalian species and low with the other types of Tf, but showed little change with aging of the animals. It is concluded that iron transport into the brain is dependent on the function of Tf receptors, probably on capillary endothelial cells, and that these receptors show the same type of species specificity as the receptors on immature erythroid cells. Also, the decrease in iron uptake by the brain as rats age from 15 days to adulthood is specific for the brain and is not a general effect of the aging process.Abbreviations Tf transferrin - ovo Tf ovotransferrin     

Comparative studies of the binding and growth-supportive ability of mammalian transferrins in human cells

The ability of human-derived cells in culture to bind, remove iron from, and grow in the presence of transferrins (Tf) isolated from the sera of species commonly included in tissue culture medium was investigated. Kinetic studies on HeLa cells reveal apparent first-order association rate constants of 0.43 min-1 for human Tf and 0.15 min-1 for equine Tf. Labeled chicken ovo-Tf and fetal bovine Tf were not recognized by the HeLa cells. Competition experiments with HeLa cells that use either isolated Tf or parent serum confirm these findings. Equilibrium binding experiments performed on HeLa cells at 37 degrees C in the presence of 2,4-dinitrophenol to prevent iron removal indicate 1 X 10(6) Tf bound/cell with a dissociation constant (K'D) of 28 nM for human Tf and 182 nM for equine Tf. Equilibrium binding performed at 0 degrees C to prevent endocytosis reveals 4.1-6.7 X 10(5) Tf binding sites/cell with a K'D of 8.3 nM for human Tf and 41.5 nM for equine Tf. Parallel experiments in normal human diploid fibroblast-like MRC-5 cells indicate expression of 0.82-2.78 X 10(5) Tf binding sites/cell with a K'D of 8.2 nM for human and 39.1 nM for equine Tf. Thus, the results of equilibrium binding studies of a more differentiated cell type are consistent with those found for HeLa cells. Fetal bovine Tf was found to compete weakly with labeled human Tf for human receptor on HeLa cells in a soluble receptor assay, with an approximately 500-fold excess needed to reduce binding to half maximal. Iron uptake experiments show an iron donating hierarchy where human greater than horse greater than calf, suggesting that the rate of iron uptake depends on the affinity of receptor for transferrin. Growth experiments involving HeLa cells in chemically defined serum-free medium demonstrate that bovine Tf will support growth as well as human Tf, but at concentrations much higher than are required of human Tf.     

Crystal structure and iron-binding properties of the R210K mutant of the N-lobe of human lactoferrin: implications for iron release from transferrins

Lactoferrin (Lf) and serum transferrin (Tf) combine high-affinity iron binding with an ability to release this iron at reduced pH. Lf, however, retains iron to significantly lower pH than Tf, giving the two proteins distinct functional roles. In this paper, we compared the iron-release profiles for human Lf, Tf, and their N-lobe half-molecules Lf(N) and Tf(N) and showed that half of the difference in iron retention at low pH ( approximately 1.3 pH units) results from interlobe interactions in Lf. To probe factors intrinsic to the N-lobes, we further examined the specific role of two basic residues that are proposed to form a pH-sensitive dilysine trigger for iron release in the N-lobe of Tf [Dewan, J. C., Mikami, B., Hirose, M., and Sacchettini, J. C. (1993) Biochemistry 32, 11963-11968] by mutating Arg 210 to Lys in the N-lobe half-molecule Lf(N). The R210K mutant was expressed, purified, and crystallized, and its crystal structure was determined and refined at 2.0-A resolution to a final R factor (R(free)) of 19.8% (25.0%). The structure showed that Lys 210 and Lys 301 in R210K do not form a dilysine interaction like that between Lys 206 and Lys 296 in human Tf. The R210K mutant retained iron to lower pH than Tf(N), consistent with the absence of the dilysine interaction but released iron at approximately 0.7 pH units higher than Lf(N). We conclude that (i) the ability of Lf to retain iron to significantly lower pH than Tf is due equally to interlobe interactions and to the absence in Lfs of an interaction analogous to the dilysine pair in Tfs, even when two lysines are present at the corresponding sequence positions, and (ii) an appropriately positioned basic residue (Arg 210 in human Lf) modulates iron release by inhibiting protonation of the N-lobe iron ligands, specifically His 253.     

The molecular mechanism for receptor-stimulated iron release from the plasma iron transport protein transferrin

Human transferrin receptor 1 (TfR) binds iron-loaded transferrin (Fe-Tf) and transports it to acidic endosomes where iron is released in a TfR-facilitated process. Consistent with our hypothesis that TfR binding stimulates iron release from Fe-Tf at acidic pH by stabilizing the apo-Tf conformation, a TfR mutant (W641A/F760A-TfR) that binds Fe-Tf, but not apo-Tf, cannot stimulate iron release from Fe-Tf, and less iron is released from Fe-Tf inside cells expressing W641A/F760A-TfR than cells expressing wild-type TfR (wtTfR). Electron paramagnetic resonance spectroscopy shows that binding at acidic pH to wtTfR, but not W641A/F760A-TfR, changes the Tf iron binding site > or =30 A from the TfR W641/F760 patch. Mutation of Tf histidine residues predicted to interact with the W641/F760 patch eliminates TfR-dependent acceleration of iron release. Identification of TfR and Tf residues critical for TfR-facilitated iron release, yet distant from a Tf iron binding site, demonstrates that TfR transmits long-range conformational changes and stabilizes the conformation of apo-Tf to accelerate iron release from Fe-Tf.     

Preparation of iron-deficient tissue culture medium by deferoxamine-sepharose treatment and application to the differential actions of apotransferrin and diferric transferrin.

We have shown that triiodothyronine-dependent GH1 rat pituitary cell growth in serum-free defined culture required apotransferrin (apoTf) (D. A. Sirbasku, et al., Biochemistry 30, 295-304, 7466-7477, 1991). These studies were done in "low-Fe" medium without Fe(III)/Fe(II) salts. Nonetheless, significant concentrations of iron may have been contributed by other components, making this medium unsuitable for study of the differential effects of apoTf and diferric transferrin (2Fe.Tf). Measuring residual iron in culture medium has been troublesome because the most sensitive method (i.e., atomic absorption) detected levels only in excess of 10 ng/ml and did not distinguish between the forms of iron present. To estimate the Fe(III) available to bind to apoTf, we developed a more sensitive and specific method. Urea-polyacrylamide gel electrophoresis (PAGE) separates apoTf, the two monoferric transferrins, and 2Fe.Tf. [125I]apoTf was incubated with medium, or components, and the formation of [125I]-2Fe.Tf was monitored by urea-PAGE/autoradiography. By this method, the concentration of Fe(III) in low-Fe medium was estimated at 8.4 to 20 ng/ml and the sources were identified. We next sought to remove the Fe(III). Standard chelators were ineffective or cytotoxic. In contrast, an affinity method with deferoxamine-Sepharose depleted greater than or equal to 90% of the Fe(III). In this medium, apoTf and 2Fe.Tf showed differential effects with GH1 cells and with MCF-7, MTW9/PL2, an MDCK cells. With the methods described here, the effects of apoTf and 2Fe.Tf on growth can be studied separately.