Uptake and release of transferrin and iron by mitogen-stimulated human lymphocytes

Phytohaemagglutinin stimulation of human peripheral blood lymphocytes resulted in the expression of transferrin receptors and the uptake of iron into the cells. As assessed from the resistance of the ¹²⁵I label to pronase, transferrin was rapidly bound and internalized at 37°C, while at 4°C 85% of the ¹²⁵I label remained on the cell surface and was degraded by the pronase. Over 94% of the ¹²⁵I label associated with and subsequently released from the cells was acid precipitable, indicating that transferrin was not degraded during its uptake and release. After reincubation for 1 h in fresh medium 70% of the cell associated ¹²⁵I-transferrin was released. In contrast, less than 30% of the ⁵⁹Fe was released, showing that iron was removed from transferrin and retained by the cells.
Concentration dependent binding of ¹²⁵I-transferrin estimated at 37°C occurred with an apparent K_a of 5·7±1·1 × 10⁷l mol⁻¹ (mean ± SD, n = 4) indicating little variation between cells from different individuals, although the number of transferrin molecules associated with the cells varied greatly from 6·2 × 10⁴/cell to 1·4 × 10⁵/cell. The rate of iron uptake from ⁵⁹Fe and ¹²⁵I labelled transferrin at 37°C by the cells from different subjects was also very variable, with a range between 0·46 and 2·27 pg Fe/min/10⁶ cells ( n = 6). However, iron uptake did not correlate with the amount of transferrin bound. This suggests that transferrin uptake and the release of iron from the transferrin to the interior of the cell are controlled independently.     

In vitro and in vivo studies of iron delivery by human monoferric transferrins

S ummary . According to the Fletcher-Huehns hypothesis there exists a functional difference between the two iron-binding sites of transferrin. In this study we present the results of an evaluation of this hypothesis in vitro and in vivo with human pure monoferric transferrins obtained by preparative isoelectric focusing in granulated gels. The uptake of iron from monoferric transferrins TfFe_c and Fe_NTf by erythroid bone marrow cells, hepatocytes and stimulated T-lymphocytes in vitro was equal, even when both monoferric transferrins were present together in the incubation medium. Ferrokinetic studies in vivo , performed with both pure monoferric transferrins, showed that transferrin TfFe_c, as well as transferrin Fe_NTf, mainly deliver their iron to the erythron. As red cell ⁵⁹Fe utilization, red cell iron turnover and other ferrokinetic parameters, obtained from this study, were identical too it is evident that both iron-binding sites of transferrin are functionally homogeneous in vivo , with respect to iron delivery.     

Evidcnce for the Functional Heterogeneity of the Two Sites of Transferrin in Vitro

A recently developed crossed immunoelectrophoretic method for displaying and quantitating the four possible molecular species of transferrin has been utilized to assess the relative effectiveness of each site of rabbit and human diferric transferrin in providing iron to rabbit reticulocytes. The site which appears to reside in the N-terminal half of the rabbit protein was found to be at least 5 times more effective than its counterpart. However, both sites may serve as iron donors in monoferric as well as diferric rabbit transferrins. It is also possible that iron may be removed from rabbit transferrin in pairwise as well as sequential fashion. In human diferric transferrin, the site in the C-terminal domain functions as the better iron donor for rabbit reticulocytes.     

The behavior of transferrin iron in the rat

The behavior of rat transferrin has been investigated employing acrylamide gel electrophoresis and isoelectric focusing. In vitro trace labeling with iron chelates at 30 min was 93%-98% effective, whereas binding by simple ferric salts was reduced to 71%-76%. Complete and specific binding of 59FeSO4 by the iron binding sites of transferrin was demonstrated after in vitro or in vivo addition of ferrous ammonium sulfate in pH 2 saline up to the point of iron saturation. In vitro the radioriron transferrin complex in plasma was stable and its iron had a negligible exchange with other transferrin binding sites over several hours. The distribution of radioiron added in vitro or through absorption was shown to be random between the binding sites of slow and fast transferrin molecule. Iron distribution among body tissues was similar for mono- and diferric transferrin iron and was not affected by the site distribution of iron on the transferrin molecule. The only important aspect of transferrin iron binding was the more rapid tissue uptake of iron in the diferric form was compared to monoferric transferrin. Additional in vivo effects on internal iron exchange were produced by changes in the iron balance of the animal. In the iron loaded animal, monoferric transferrin injected into the plasma was rapidly loaded by iron from tissue and thereby converted to diferric transferrin. Injection of diferric transferrin in the iron deficient animal was associated with a rapid disappearance from circulation of the original complex and a subsequent appearance of monoferric transferrin as a result of iron returning from tissues. These observations support the concept that plasma iron behaves as a single pool except that diferric iron exchange occurs at a more rapid rate than dose monoferric iron exchange.     

Competitive advantage of diferric transferrin in delivering iron to reticulocytes.

Radioiron- and radioiodine-labeled forms of human diferric and monoferric transferrin and apotransferrin, isolated by preparative isoelectric focusing, were used to define transferrin-iron uptake by human reticulocytes. In mixtures of human diferric and monoferric transferrin, the diferric molecule had a constant 7-fold advantage in delivering iron to reticulocytes, as compared with the 2-fold advantage when single solutions of mono- and diferric transferrins were compared. This was shown to be due to competitive interaction in iron delivery, probably at a common membrane-receptor binding site for transferrin. Apotransferrin did not interfere with the iron-donating process and its limited cellular uptake was inhibited in noncompetitive fashion by diferric transferrin.     

Uptake and release of iron from human transferrin.

Purified fractions of human apotransferrin, monoferric transferrins with iron on the acid-labile binding site and on the acid-stable binding site, and diferric transferrin have been prepared. The iron loading and unloading behavior of these preparations has been examined by isoelectric focusing. Iron release from the two monoferric transferrin preparations to human reticulocytes was of similar magnitude. In a mixture containing equal amounts of diferic and monoferric iron, approximately 4 times the amount of iron delivered by the monoferric species was delivered by the diferric species. Iron loading of transferrin in vitro showed a random distribution between monoferric and diferric transferrin. Among the monoferric transferrins, loading of the acid-labile binding sites was greater than that of the acid-stable binding sites. In vivo iron distribution in normal subjects, as evaluated by in vitro-added 50Fe, gave similar results. Absorption of a large dose of orally administered iron in iron-deficient subjects resulted in a somewhat greater amount of diferric transferrin at low saturation and a somewhat smaller amount of diferric transferrin at higher saturations than would have been anticipated by random loading. These data would indicate that in the human, iron loading of transferrin may be considered essentially random. Unloading from the two monoferric transferrin species is of similar magnitude but far less than that delivered by diferric transferrin.     

Transferrin Receptors in Developing Murine Erythroid Cells

S ummary . Techniques of cell separation were used to isolate murine erythroid cells at different stages of maturation. The number of transferrin receptors in these cell populations was assayed by measuring binding of ¹²⁵I-labelled transferrin. Nearly 23 times as many receptors were found in the least mature cells, chiefly pronormoblasts, as in reticulocytes. Iron transport, determined by measurement of the rate of ⁵⁹Fe uptake from ⁵⁹Fe-labelled transferrin, was proportional to the number of receptors at all stages of differentiation. Electron microscope radioautographic studies of the interaction of ¹²⁵I-labelled transferrin with erythroid precursor cells demonstrated that 15–35% of cell associated transferrin was intracellular in erythroid precursors.     

Ferritin and iron uptake by reticulocytes

The uptake of liver ferritin labelled with ¹²⁵I or ⁵⁹Fe by guinea-pig reticulocytes has been studied to investigate the characteristics of the uptake process, compare it with transferrin uptake and determine whether ferritin-iron is utilized by the cells in haem synthesis. The results confirmed that guinea-pig reticulocytes, but not mature erythrocytes, take up liver ferritin by a saturable, time- and temperature-dependent process. Up to 70% of the iron taken up by the cells was utilized in haem synthesis and competed directly with iron derived from transferrin. Scatchard analysis of the binding parameters indicated that 30–130 × 10³ ferritin molecules were bound per cell to high affinity specific membrane receptors ( K _a: 1·77 × 10⁷ M⁻¹). In contrast, rat took up much less ferritin than guinea-pig reticulocytes and the process was entirely non-specific. Release experiments with guinea-pig reticulocytes at 37°C showed that a maximum of about 70% of the cell-associated ¹²⁵I-ferritin was released from the cells of which up to 15% was trichloroacetic acid-soluble.
We suggest that ferritin uptake by guinea-pig reticulocytes involves receptor-mediated endocytosis. The endocytotic vesicle fuses with a lysosome, iron is removed from the protein and enters a cytosolic pool in which it competes directly with transferrin-derived iron to provide iron for mitochondrial haem synthesis. Some of the ferritin is catabolized and the rest is returned to the extracellular medium during membrane recycling.     

Release of Iron by Resident and Stimulated Mouse Peritoneal Macrophages following Ingestion and Degradation of Transferrin—Antitransferrin Immune Complexes

S ummary . A method of loading macrophages from normal and inflammatory mouse peritoneal exudates with ⁵⁹Fe using ⁵⁹Fe, ¹²⁵I-transferrin—antitransferrin immune complexes is described and the subsequent release of iron and degraded transferrin to the incubation medium has been studied. Release of iron occurred more rapidly from resident macrophages than from thioglycollate broth-induced (stimulated) macrophages, but degradation of the ¹²⁵I-transferrin in the immune complexes was faster in stimulated cells. A small percentage of the iron released was in the form of ferritin. Desferrioxamine (1 mM) increased the release of iron from both stimulated and resident macrophages, the effect being proportionally greater in the stimulated cells. Ascorbic acid (1 mM) had no effect on the release of iron, nor did the addition of apotransferrin (1 mg/ml) to the culture medium. These results support the concept of a blockade of iron release by reticuloendothelial cells in states of inflammation, and suggest that it may be a primary cause of the anaemia of chronic disease.     

Plasma radioiron kinetics in man: explanation for the effect of plasma iron concentration.

The plasma iron turnover was measured in 19 normal subjects. A correlation was found between plasma iron concentration and plasma iron turnover. In addition to the turnover of 55Fe at normal plasma iron concentration (predominantly monoferric transferrin), a second turnover in which the labeled plasma was saturated with iron (to produce predominantly diferric transferrin) was studied with 50Fe. It was demonstrated that diferric transferrin had a greater rate of iron turnover but that the distribution between erythroid and non-erythroid tissues was unchanged. It was concluded that plasma iron turnover is dependent on the monoferric/diferric transferrin ratio in the plasma but that the internal distribution of iron is unaffected.     

The Role of Endocytosis in Transferrin Uptake by Reticulocytes and Bone Marrow Cells

S ummary . The aims of the work described in this paper were to investigate the process by which transferrin enters immature erythroid cells, to determine whether entry of transferrin into the cells is required for iron utilization by the cells and to see whether substances in the medium surrounding the cells enter the cells along with transferrin. The kxperiments were performed using rabbit transferrin, rabbit reticulocytes and rabbit bone marrow cell suspensions. The process of transferrin uptake was studied by electron microscope techniques using autoradiography of ¹²⁵I-labelled transferrin or transferrin conjugated to ferritin or horseradish per-oxidase. The results confirmed that transferrin enters reticulocytes and showed that the protein also enters nucleated erythroid cells of the bone marrow but very little is taken up by platelets, cells of the white cell series in the bone marrow or peripheral blood or mature erythrocytes. The two stages of transferrin uptake by erythroid cells, initial absorption and progressive uptake, were shown to be represented morphologically by binding of transferrin to the outer cell membrane, probably to specific receptors, and entry into the cells by endocytosis, respectively. Reticulocytes were unable to utilize ⁵⁹Fe for haem synthesis when the ⁵⁹Fe was on transferrin which was bound to agarose, Enzacryl AA or Latex beads so that it would not enter the cells. The uptake of transferrin by reticulocytes was not accompanied by uptake of ¹⁴C-polyethylene glycol dissolved in the medium.     

Hepatocyte iron kinetics in the rat explored with an iron chelator

S ummary . The hepatocyte metabolism of ⁵⁹Fe-labelled ferritin, haemoglobin-haptoglobin and transferrin has been examined in rats. All three forms of ⁵⁹Fe became transiently available to desferrioxamine (DF) at the time they would otherwise have entered storage or alternative pathways of iron metabolism. However, differences in both the patterns of spontaneous ⁵⁹Fe reutilization by normal and iron deficient rats and the partition of chelate iron excretion between bile and urine, suggested that iron in transit within hepatocytes did not behave as a single common pool. Ferritin ⁵⁹Fe, entering a pool of non-radioactive iron the size of which is determined by liver iron stores, was chelated predominantly into the bile. Transferrin ⁵⁹Fe was distinguished by a greater reflux to the erythron in iron deficient rats, and by excretion of a larger proportion of ⁵⁹Fe chelated by DF in the urine. Haemoglobin-haptoglobin ⁵⁹Fe followed a metabolic pathway which was relatively independent of both the iron stores and DF. If the heterogeneous behaviour of rat hepatocyte transit iron has a parallel in man, alterations in the size of similar chelatable iron pools could explain the dependence of DF-induced urine and faecal iron excretion on both liver iron stores and the level of erythropoiesis.     

New Kinetic Role for Serum Ferritin in Iron Metabolism

S ummary . The role of the iron in serum ferritin was investigated with regard to iron kinetics. Serum (or plasma) was fractionated on a sucrose gradient by ultra-centrifugation to separate iron in ferritin from other iron-containing components in serum. After injection of heat-treated [⁵⁹Fe]RBC in rats, detectable labelling of the serum ferritin fraction was brief, between 20 and 40 min. The labelling of other iron components in the serum was maximal between 1 and 2 hr. No labelling of the serum ferritin fraction was detected after intravenous administration of [⁵⁹Fe]-transferrin in man and in the rat, or after [⁵⁹Fe]haemoglobin-haptoglobin or oral [⁵⁹Fe²⁺]citrate in the rat. These results indicate that iron in serum ferritin originates in part from the degradation of haemoglobin in senescent red cells and that this labelled ferritin is cleared from the serum rapidly.
In normal rats, [⁵⁹Fe]ferritin, purified from rat liver and injected intravenously, disappeared from the serum at an exponential rate with a half-disappearance time (HDT) of 4 min. After 1 hr 97% of the label had accumulated in the liver. After a delay of 24 hr, increasing amounts of label appeared in the RBC to an estimated maximum of 70–80% at 5 days. Conditions which increased or decreased rates of synthesis of haemoglobin altered the percentage of [⁵⁹Fe]ferritin incorporated into haemoglobin, but these conditions did not change the rate of removal of [⁵⁹Fe]ferritin from the serum by the liver. These results indicate that the iron in serum ferritin is cleared from the serum much more rapidly than transferrin-bound iron. The data are consistent with the passage through serum ferritin of a major portion of the iron that is released from the degradation of red cell haemoglobin in the RE system.     

Transferrin and Iron Uptake by Rabbit Bone Marrow Cells in Vitro

S ummary . The interaction between rabbit transferrin and rabbit bone marrow cells and reticulocytes was investigated using transferrin labelled with ¹²⁵I and ⁵⁹Fe. The pattern of transferrin and iron uptake by bone marrow was found to be similar to that by reticulocytes and occurred in four stages, viz (1) adsorption of transferrin, (2) progressive uptake of transferrin, (3) release of the iron to the cell, (4) release of transferrin from the cell.
Total transferrin uptake per bone marrow erythroid precursor cell was eight times that per reticulocyte, while the rate of iron uptake was at least twice as great with marrow cells. Transferrin uptake and release and iron uptake were temperature dependent. Transferrin and iron uptake were inhibited by metabolic inhbitors, particularly those affecting oxidative metabolism. The activation energies for the association and dissociation reaction of transferrin and the activation energy for iron uptake by bone marrow cells were higher than those reported for reticulocytes. A further difference between the two cell types was that at a constant iron concentration, iron uptake by bone marrow cells decreased with decreasing degrees of transferrin saturation with iron. Small increases in transferrin uptake by bone marrow cells were also obtained with increasing transferrin saturation. The data support the view that both iron and transferrin uptake are dependent upon cellular metabolism.     

Interaction of human diferric transferrin with reticulocytes.

Methods have been devised for preparing human transferrin with a different isotope of iron selectively labeling each of the two iron binding sites and for determining the distribution of radioiron among transferrin molecules. When diferric human transferrin was exposed to human or animal reticulocytes, there was an equal contribution of radioiron from the acid-stable and acid-labile sites. In this delivery, both atoms of iron were removed simultaneously from the diferric transferrin molecule, converting it to apotransferrin. At similar iron concentrations the amount of iron delivered by diferric transferrin was twice that delivered by monoferric transferrin.     

Biliary Iron Excretion in Rats following Pyridoxal Isonicotinoyl Hydrazone

S ummary Biliary excretion of iron after administration of pyridoxal isonicotinoyl hydrazone (PIH), a recently identified effective iron-chelating agent, was investigated in rats. PIH administered both intraperitoneally and orally was shown to increase significantly ⁵⁹Fe excretion into bile of rats which had previously been injected with ⁵⁹Fe-transferrin to label hepatic parenchymal cells. ⁵⁹Fe-PIH appears in bile as early as 15 min after chelator administration and the peak of ⁵⁹Fe-radioactivity in bile is seen 1–5 h following intraperitoneal PIH injection. PIH, administered intraperitoneally, 125–250 mg/kg, increased 24 h biliary radioiron excretion about 35 times and in addition increased urinary and faecal iron excretion. When PIH was given immediately before ⁵⁹Fe-transferrin, 24 h cumulative biliary ⁵⁹Fe excretion was even higher. PIH was also demonstrated to increase biliary excretion of radioiron released from ⁵⁹Fe-haemoglobin catabolysed in reticuloendothelial cells. The effect of PIH was confirmed by estimation of biliary iron concentration using the method of atomic absorption spectrophotometry. Repeated PIH administration to rats decreased ⁵⁹Fe radioactivity in liver and kidney and increased urinary and faecal iron excretion.     

Some Factors Affecting the Release of Iron from Reticuloendothelial Cells

S ummary . Factors modifying the release of iron from reticuloendothelial cells were studied in rats by injecting heat-denatured erythrocytes containing [⁵⁹Fe]haemoglobin. The cells were rapidly taken up by the liver and spleen, and a proportion of the ⁵⁹Fe was released into the plasma, the maximum rate being between 1 and 4 hr after injection. The remaining ⁵⁹Fe was incorporated into storage compounds. A 10-fold variation in the load of denatured erythrocytes produced a proportional change in the amount of iron released, the percentage remaining constant. Percentage release of ⁵⁹Fe was enhanced in venesected rats and diminished in hypertransfused rats. Release was inhibited by injecting either unlabelled denatured erythrocytes or iron bound with nitrilotriacetic acid (NTA-iron) before the ⁵⁹Fe-labelled cells, the maximum effect being obtained if the interval between the two injections was 3-9 hr. Release was also inhibited by injecting NTA-iron 30 min after the denatured labelled erythrocytes. Inhibition was always preceded by a rise in the serum-iron concentration, and was associated with an increase in the percentage of ⁵⁹Fe incorporated into ferritin. It is postulated that the shortage of free transferrin binding sites for iron delays the entry of liberated haemoglobin iron into the plasma, and consequently there is enlargement of a 'pre-release' iron pool. Other workers have shown that iron induces the synthesis of ferritin; the presence of a stimulated mechanism for ferritin synthesis within the reticuloendothelial cells would result in the diversion of an increased percentage of erythrocyte iron into storage compounds.     

Functional heterogeneity of transferrin-bound iron: iron uptake by cell suspensions from bone marrow and liver and by cell cultures of fibroblasts and lymphoblasts.

According to the hypothesis of Fletcher and Huehns, functional differences exist between both iron-binding sites of transferrin. The site designated A should mainly be involved in the delivery of iron to erythroid cells, whereas site B should donate its iron preferentially to cells involved in the absorption and storage of iron. In the present study this hypothesis could be confirmed by in vitro experiments with various cell types. Iron transferrin preincubated with rat bone marrow cells donates less iron to rat bone marrow cells, Chinese hamster fibroblasts, human fibroblasts and human lymphoblasts than freshly prepared iron transferrin equal in iron and transferrin concentraion. Rat liver parenchymal cells, however, take up more iron from preincubated than from freshly prepared iron transferrin. Obviously, site A not only donates iron preferentially to erythroid cells but also to (rapidly) dividing nonerythroid cells in culture. From experiments with iron transferrin mixtures in which radioiron was present at low or high iron saturation, it could be concluded that rat bone marrow cells take up iron equally well from monoferric as from diferric transferrin. The observed functional heterogeneity could, therefore, not be ascribed to differences between monoferric and diferric transferrin.     

Molecular aspects of the binding of absorbed iron to transferrin

S ummary . To study the molecular aspects of the binding of absorbed iron to plasma transferrin, ⁵⁹Fe with high specific activity was administered via intragastric tube to iron-deficient rabbits. The distribution of the absorbed ⁵⁹Fe among the molecular forms of iron-transferrin was analysed using urea-polyacrylamide gel electrophoresis. Absorbed iron was bound to circulating transferrin one atom at a time. In four out of five animals, absorbed iron was predominantly bound to the site in the N-terminal domain of the protein. Thus, the two sites of transferrin may differ in their ability to load absorbed iron.     

Transferrin—Reticulocyte Binding: Evidence for the Functional Importance of Transferrin Conformation

Iron-free and ⁵⁹Fe-labelled human and rabbit ¹²⁵I-transferrins have been chromatographed on DEAE cellulose at pH 7.9 and have also been incubated with rabbit reticulocytes in order to determine transferrin and iron uptake.
Chromatography of human apo- and iron-transferrin indicated clear differences in their surface properties which are considered to be conformation-dependent. These differences were not evident in rabbit transferrin samples.
Human and rabbit iron-transferrins and rabbit apo-transferrin all associated rapidly with rabbit reticulocytes; human apo-transferrin was only weakly bound. Correlation of data from both experiments suggested that reticulocyte binding of human apo-transferrin is limited by its molecular confirmation. The incubation studies confirmed results of other workers and indicated that the data from these experiments with human and rabbit transferrin are not in conflict.
Compared with rabbit iron-transferrin, only half as much human iron-transferrin was bound to rabbit reticulocytes in a comparable experiment. Iron transfer from bound human and rabbit transferrin took place at equal rate. It is argued that fewer receptors on rabbit reticulocytes are available for human transferrin because, when bound, the size and shape of a molecule limits further binding of transferrin molecules at adjacent sites.