Ontogenetic changes in protein dephosphorylating activity of the cytosolic compartment of rat erythrocytes

Total casein phosphatase activity of erythrocytes from one-month-old rats was separated by DEAE-cellulose chromatography into three peaks--E1, E2 and E3--and only into two peaks--E1 and E3--when the erythrocyte donors were six- and 12-month-old rats. The activity of E1 (Mr 330 K) decreased continuously in erythrocytes during the first year of postnatal life. E2 (Mr 230 K) also decreased and completely disappeared from the cells of 12-month-old rats. E3 (Mr 180 K) was the dominant molecular form in the cytosol of erythrocytes during the first year of life. It decreased only up to six months of life. In this form E3 seems to be cooperative with respect to the substrate and to inhibitor molecules. The decrease of its kinetic parameters (Vmax and K0.55) was also found during postnatal ontogenesis. E3 isolated from erythrocytes of older rats (6 and 12 months) was more susceptible to inhibitory effect of pyrophosphate and to the change of ionic strength of eluting buffer than the enzyme from one-month-old rats. 0.2 mol.1(-1) NaCl lowered Mr of E3 phosphatase from 180 K to 128 K only in older rats.     

Electrophoretic characterization and subcellular distribution of hexokinase isoenzymes in red blood cells of rabbits

The isoenzyme pattern of hexokinase in rabbit red cells (erythrocytes, fetal erythrocytes and reticulocytes) were determined by means of agarose gel and disc electrophoresis. One duplicated hexokinase (4a and 4b according to the IUPAC-nomenclature) was detected in rabbit erythrocytes as also described for human erythrocytes. Besides the isoenzymes 4a and 4b reticulocytes also contain hexokinase 2 and 3 like rabbit and rat liver. The high KM glucose phosphorylating enzyme, hexokinase 1 could be demonstrated only under specific conditions in the reticulocytes during the initial stage of the anemia. After the fractionation of reticulocyte homogenates the total hexokinase activity was recovered in the mitochondria and cytosol to nearly equal amounts as revealed by the distribution of markers. Hexokinase 2 and 3 were detectable in reticulocytes and in isolated mitochondria only after the addition of certain dissociating agents. In contrast to the tightly bound mitochondrial hexokinases 2 and 3 the type 4a and 4b are more loosely bound and exhibit a bilocal distribution between mitochondria and cytosol of reticulocytes.     

A cellular activator of catecholamine-sensitive adenylate cyclase in rat reticulocytes and erythrocytes: changes during reticulocyte development and effects on the beta receptor

Rat reticulocytes contain an isoproterenol-sensitive adenylate cyclase activity which is lost with maturation to erythrocytes despite no change in the density of β-adrenergic receptors. To explore this observation, a cytosol factor, previously shown to be important in the expression of catecholamine-sensitive adenylate cyclase in the reticulocyte, was compared to a cytosol factor obtained in a similar manner from mature erythrocytes. The cytosol factor from reticulocytes augmented isoproterenol-responsive adenylate cyclase activity in reticulocyte and erythrocyte membranes half-maximally at 0.7 ± 0.1 (SEM) and 1.1 ± 0.3 μg/ml, respectively. These concentrations of reticulocyte-derived cytosol factor were significantly lower (P < 0.01) than those concentrations of the factor from erythrocytes necessary to augment isoproterenol-responsive adenylate cyclase activity in reticulocyte (9.7 ± 2.3) and erythrocyte (7.5 ± 1.0) membranes. Cytosol factor from reticulocytes also caused greater total isoproterenol responsiveness than that from erythrocytes both in reticulocyte (784 ± 107 vs 525 ± 65 pmol/mg protein) and in erythrocyte membranes (54 ± 6 vs 36 ± 3); P < 0.05. Neither reticulocyte nor erythrocyte cytosol factor affected the concentration at which isoproterenol half-maximally stimulated adenylate cyclase in either set of membranes. However, the cytosol factor from reticulocytes markedly decreased the binding affinity of isoproterenol for β receptors in reticulocytes from 0.8 ± 0.2 to 6.9 ± 1.4 μm; P < 0.001. This reticulocyte factor had no significant effect on the binding affinity of isoproterenol for erythrocyte membranes. Erythrocyte factor did not change the binding affinity for isoproterenol in either reticulocyte or erythrocyte membranes.     

Levels of active ubiquitin carrier proteins decline during erythroid maturation

Energy-dependent proteolysis is lost during maturation of rabbit reticulocytes to erythrocytes (Speiser, S., and Etlinger, J.D. (1982) J. Biol Chem. 257, 14122-14127), but nothing is known about the fates of individual components in the multienzyme ATP- and ubiquitin (Ub)-dependent proteolytic pathway during this process. Rabbit reticulocytes contain five low molecular weight carrier proteins (E2s) that form labile Ub adducts in the presence of Ub-activating enzyme (E1) (Pickart, C. M. and Rose, I. A. (1985) J. Biol. Chem. 260, 1573-1581). A method to estimate levels of active E2s in erythroid cells has been developed involving: 1) stepwise anion exchange fractionation of a soluble lysate; 2) addition of purified E1, MgATP, and radioiodinated Ub to the fractions followed by gel electrophoresis of the resulting E2-Ub adducts; and 3) quantitative densitometry of autoradiographs. Levels of active E2s are much lower in (rabbit) erythrocytes than in reticulocytes. Mean -fold decreases are: E235K, 6 x; E2(25K), 11 x; E2(20K), 18 x; E2(17K), not detected in erythrocytes; E2(14K), 12 x. The large decreases in levels of E2(20K) and E2(14K) are consistent with known functions of these proteins in DNA repair and Ub-dependent proteolysis, respectively. Decreases in levels of the other E2s, whose biological roles are presently unknown, suggest diminished requirements, if any, for them in erythrocyte metabolism. The analysis revealed two previously undescribed carrier proteins, one of which has a high molecular weight. Additional catalytic properties of E2(35K) and E2(14K) are reported.     

Cobalt-dependent protein phosphatases from human cord blood erythrocytes. I. Submolecular structure and regulation of activity of E3 casein phosphatase

We have identified three phosphoprotein phosphatases in the cytosol of human cord blood erythrocytes by sequential anion-exchange chromatography and gel filtration. The most abundant was E3 protein phosphatase. After rechromatography on a column of Ultrogel AcA-44 the enzyme had a molecular weight of 95,000 daltons. According to the data obtained by SDS/PAGE, the 95,000-dalton form was composed of non-identical subunits with a molecular mass of 23,000 and 16,000 daltons. Since ethanol decreased the molecular mass of the 95,000-dalton enzyme to 25,000 daltons, we suggest that the protein of 23,000-25,000 daltons represents the catalytic subunit. The decrease in the molecular weight is followed by a 2-fold increase in the Vmax value and by a change in kinetics: the negatively cooperative 95,000-dalton enzyme (h = 0.45) transforms into Michaelis-Menten kinetics (h = 1.0) in the 25,000-dalton form. Both molecular forms, 95,000 and 25,000 daltons, only dephosphorylated casein but not phosvitine and histones. Both forms were activated by CoCl2 and inhibited by organic, and most potently, by inorganic pyrophosphates to approximately the same degree. As opposed to the inorganic pyrophosphate, which affects the catalytic properties of the enzyme molecule, CoCl2 did not affect the catalytic properties of the enzymes, but it probably did affect the rate of 'E-S' complex formation. CoCl2 protected the 95,000-dalton enzyme from pyrophosphate inhibition. The data indicate that CoCl2 and pyrophosphate may take part in the regulation of the activity of both forms of E3 phosphatase.     

Cobalt-dependent protein phosphatases from human cord blood erythrocytes. II. Further characterization of E2 casein phosphatase

Of three casein phosphatases isolated from the cytosol of human cord blood erythrocytes two were cobalt-dependent, E2 and E3. In the presence of CoCl2, E2 activity was the most prominent. In addition to casein, E2 dephosphorylated phosvitin and p-nitrophenyl phosphate (p-NPP) with pH optima at 6.8-7.2 for proteins and 9.0 for p-NPP. The native enzyme had a molecular weight of 104,000 daltons after AcA-44 Ultrogel filtration. According to SDS/PAGE it consisted of two subunits, 78,000 and 15,000 daltons. The 104,000-dalton form exhibited Michaelis-Menten kinetics and had the greatest affinity for casein between protein substrates tested. Ethanol denaturated the enzyme by 80%. Optimal activation of E2 phosphatase was achieved with 5 mmol/l CoCl2 which did not affect the catalytic properties of the enzyme but did affect the rate of 'E-S' complex formation. Inorganic pyrophosphate was not inhibitory for the 104,000-dalton enzyme. Judging by all these properties the natural substrate for E2 casein phosphatase could be P-pyruvate kinase.     

Soluble and membrane-bound polyphosphoinositide phosphohydrolases in mammalian erythrocytes

Phosphatases and phosphodiesterases that hydrolyse polyphosphoinositides are described in both membrane and cytosol fractions of human, pig, rat, rabbit, and sheep erythrocytes using exogenous substrates. With suitably optimized assay conditions, Ca2+-dependent phosphatidylinositol bisphosphate (PIP2) phosphodiesterase activity was found in the hemoglobin-free cytosol fraction, as well as the membrane. Membrane activity is completely dependent upon Triton X-100 and salt and inhibited by cetyltrimethylammonium bromide (CTAB), while the soluble activity requires CTAB and is inhibited by Triton. A low Ca2+-dependent PIP2 phosphatase activity, not present in other tissues, was also detected. The cation-independent phosphatidylinositol phosphate (PIP) phosphatase is localized in the membrane in most species, while the diesterase and the PIP2 phosphatases (both Mg2+ and Ca2+ dependent) are localized in the cytosol. Rat and rabbit erythrocytes are atypical in having a substantial proportion of their Mg2+-dependent PIP2 phosphatase activities in the membrane. All activities are lowest in sheep erythrocytes, except the PIP phosphatase, most of which is soluble in this species. Ca2+-dependent PIP2 phosphatase activity is not correlated with the activity or subcellular distribution of any of the other hydrolases and seems to be a separate enzyme. All the phosphoinositide hydrolase activities, particularly the diesterase, are orders of magnitude lower in erythrocytes than in other tissues. Both soluble and membrane diesterase activities are lost as erythrocytes age. Soluble polyphosphoinositide diesterase does not seem to be active with membrane-bound substrate, since pig and sheep erythrocytes that have negligible membrane activity do not respond to Ca2+ loading, yet have substantial diesterase activity in the cytosol. This supports the view that the diesterase is not physiologically functional in normal erythrocytes.     

The antioxidant status during maturation of reticulocytes to erythrocytes in type 2 diabetics

Free radical-induced lipid peroxidation has been associated with numerous disease processes including diabetes mellitus. The extent of lipid peroxidation (LPO) and antioxidant defense system [i.e., levels of glutathione (GSH), glucose-6-phosphate dehydrogenase (G6PDH), glutathione reductase (GR), glutathione peroxidase (GPx), glutathione-S-transferase (GST), and catalase (CAT)] were evaluated in reticulocytes and erythrocytes of type 2 diabetic males and age-matched controls. Type 2 diabetics have shown increased lipid peroxidation and decreased levels of GSH, GR, GPx, G6PDH, and GST both in reticulocytes and erythrocytes compared to controls, indicating the presence of oxidative stress and defective antioxidant systems in these patients. CAT activity is found to be enhanced in both the reticulocytes and erythrocytes of diabetics, with a greater percentage enhancement in reticulocytes. The extent of increase in lipid peroxidation is greater in erythrocytes compared to reticulocytes in these patients. Furthermore, the maturation of reticulocytes to erythrocytes resulted in decreased GSH and decreased activities of all antioxidant enzymes (except CAT) both in normals and type 2 diabetes individuals, indicating decreased scavenging capacity as reticulocytes mature to erythrocytes. These maturational alterations are further intensified in type 2 diabetics. The present study reveals that the alterations in lipid peroxidation and antioxidant system lean toward early senescence of erythrocytes in type 2 diabetic patients.     

Modifications of aldolase during in vivo aging of rabbit red cells.

Crude hemolysates, partially purified aldolase and aldolase purified to homogeneity from reticulocytes and mature erythrocytes, were incubated with a specific antiserum raised against crystalline rabbit muscle aldolase. We show that the same aldolasic activity corresponds to a greater amount of antigen in older than in younger cells, in crude hemolysates as well as in the above mentioned preparations; that is to say, old-cell aldolase contains cross-reacting material (CRM). Properties of purified enzyme from reticulocytes and mature erythrocytes were compared to those of muscle crystalline aldolase: -- the molecular specific activity of purified aldolase from erythrocytes is lower than with crystalline muscle aldolase, i.e. CRM is maintained throughout the purification steps. -- the specific activity of red cell aldolase towards both substrates (FDP and F1P) is lower than that of crystalline muscle aldolase. However, the ratio of activity towards the two substrates FDP/F1P is decreased in erythrocytes and reticulocytes. -- no other difference was found: Michaelis constant towards FDP, thermodenaturation constant and C terminal extremities are identical as are the molecular weights.     

Rabbit red blood cell hexokinase:intracellular distribution during reticulocytes maturation

Summary The intracellular localization and isozyme distribution of hexokinase were studied during rabbit reticulocyte maturation and aging. In reticulocytes 50% of the enzyme was particulate while in the mature erythrocytes all the hexokinase activity was soluble. The bound enzyme co-sediments with mitochondria and by column chromatography it was found to be hexokinase Ia. The cytosol of reticulocytes contains hexokinase Ia (38%) and hexokinase Ib (62%) while the mature erythrocytes contain only hexokinase Ia. The amount of bound hexokinase decreases very quickly during cell maturation and aging as was shown by following in vivo reticulocyte maturation or by analysis of hexokinase compartmentation in cells of different ages, obtained by density gradient ultracentrifugations. A role for this intracellular distribution of hexokinase is suggested.     

Synthesis and intracellular distribution of cathepsins E and D in differentiating murine Friend erythroleukemia cells.

The synthesis, accumulation, and cellular distribution of cathepsins E and D during the dimethyl sulfoxide (DMSO)-induced differentiation of Friend erythroleukemia cells were investigated. The cellular levels of cathepsins E and D rapidly increased within 1 day of DMSO induction and then sharply decreased over the next 7 days. Since the cells during 1 day of differentiation were morphologically the same as uninduced cells, the results suggest the importance of these enzymes in more cellular proteolysis for the following committed differentiation. While cathepsin D was present mostly in the sedimentable fraction of cells throughout the differentiation period, the distribution of cathepsin E varied to the stage of differentiation. The ratio of the soluble/sedimentable cathepsin E content was 1.1, 1.4, 0.9, and 0.7 in cells after 0, 1, 4, and 7 days of DMSO treatment, respectively. The maturation of reticulocytes to erythrocytes was accompanied by complete loss of the soluble cathepsin E and of all of the cellular cathepsin D. Immunoblotting analyses revealed that both uninduced and induced cells contained two forms of cathepsin E; a high molecular weight form (82 kDa) which was mainly associated with the sedimentable fraction and a low molecular weight form (74 kDa) which was found largely in the soluble fraction. The distribution of these two forms was not significantly changed throughout the differentiation period, but the 74-kDa protein was completely eliminated with maturation of reticulocytes to erythrocytes. Cathepsin D also appeared in two forms in both uninduced and induced cells; a minor (46 kDa) and a major (42 kDa) form which appear to have a precursor-product relationship.     

Analysis of iron-binding components in the low molecular weight fraction of rat reticulocyte cytosol

Rat reticulocytes were incubated with rat 125I-Tf-59Fe under conditions inhibiting heme synthesis. Cytosol, prepared from the reticulocytes, was separated and analysed by gel filtration and Amicon Ultrafiltration. An iron-containing low molecular weight fraction derived from the cytosol was further analysed by HPLC size-exclusion chromatography and HPLC reversed phase chromatography. Conditions inhibiting heme synthesis and uncoupling the oxidative phosphorylations lead to a large increase in the Fe-containing low molecular weight fraction in the cytosol. The components in the low molecular weight fraction have an apparent molecular weight of 5500 Dalton as determined with HPLC size-exclusion chromatography. The low molecular weight fraction contained several iron chelating components like glycin, 1/2 cystine and citrate, but no specific iron-binding proteins, nucleotides or pyrophosphate.     

Alterations of red cell membranes from phenylhydrazine-treated rabbits

Reticulocytosis was induced in rabbits by two methods: phlebotomy and injection of phenylhydrazine. Normal erythrocytes, reticulocytes from bed rabbits, reticulocytes from phenylhydrazine-treated rabbits, and erythrocytes treated in vitro with phenylhydrazine were compared with respect to their plasma membrane labeling by galactose oxidase and NaB³H₄, and lactoperoxidase-catalyzed incorporation of ¹²⁵I. Normal erythrocyte membranes and membranes from reticulocytes of bled rabbits showed almost identical labeling patterns, the presence of 2–3 glycoproteins with moderate to low mobilities on dodecyl sulfate acrylamide gel electrophoresis. Labeling in the absence of enzyme was negligible. In contrast, the reticulocytes from phenylhydrazine-treated rabbits exhibited a large incorporation of tritium without prior treatment with galactose oxidase. Even after prereduction with unlabeled NaBH₄ to remove this nonspecific labeling, the labeled glycoprotein components found in normal erythrocytes were not detectable. Normal erythrocytes treated in vitro with phenylhydrazine, washed, and labeled with galactose oxidase had labeling patterns, including high nonspecific incorporation of ³H, similar to those observed with in vivo phenylhydrazine treatment.Solubilization of membranes with lithium diiososalicylate followed by partitioning with phenol showed that the same glycoproteins were presented in normal or phenylhydrazine membranes, although only the former extract was labeled by galactose oxidase. Individual carbohydrates from the membranes were analyzed by gas-liquid chromatography and, in the case of hexosamines, on the amino acid analyzer. The results of these analyses indicated a slight decline in galactose content with phenylhydrazine treatment. Reticulocyte membranes from bled rabbits also showed a decrease in galactose content, although it was less pronounced.Most of the label incorporated by nonspecific borohydride labeling of membranes from phenylhydrazine-treated animals was found associated with protein. The modified amino acids from labeled proteins are similar to those formed in reactions of oxidized lipids and proteins in model systems.     

Alteration in protein kinase C activity and subcellular distribution in sickle erythrocytes

In agreement with previous data, membrane protein phosphorylation was found to be altered in intact sickle cells (SS) relative to intact normal erythrocytes (AA). Similar changes were observed in their isolated membranes. The involvement of protein kinase C (PKC) in this process was investigated. The membrane PKC content in SS cells, measured by [3H]phorbol ester binding, was about 6-times higher than in AA cells. In addition, the activity of the enzyme, measured by histone phosphorylation was also found to be increased in SS cell membranes but decreased in their cytosol compared to the activity in AA cell membranes and cytosol. The increase in membrane PKC activity was observed mostly in the light fraction of SS cells, fractionated by density gradient, whereas the decrease in cytosolic activity was only observed in the dense fraction. PKC activity, measured in cells from the blood of reticulocyte-rich patients, exhibited an increase in both membranes and cytosol, thus explaining some of the effects observed in the SS cell light fraction, which is enriched in reticulocytes. The increase in PKC activity in the membranes of SS cells is partly explained by their young age but the loss of PKC activity in their cytosol, particularly in that of the dense fraction, seems to be specific to SS erythrocytes. The relative decrease in membrane PKC activity between the dense and the light fractions of SS cells might be related to oxidative inactivation of the enzyme.     

Transferrin in the reticulocyte cytosol

Radioactive iodine-labeled iron-saturated human transferrin was shown to enter the cytosol of rabbit reticulocytes but not erythrocytes, and to be combined therein with a small “carrier” material not identical to the membrane transferrin receptor.     

Eperythrozoon ovis: the difference in carbohydrate metabolism between infected and uninfected sheep erythrocytes.

The glucose utilization lactic and pyruvic acid production and oxygen uptake of normal and Eperythrozoon ovis infected sheep erythrocytes were measured under aerobic conditions. Infected cells showed marded increases in both glucose utilization and acid production as compared with controls. Uninfected erythocyte samples which included a percentage of reticulocytes comparable to that found in E. ovis infection showed no apparent difference in glucose utilization and lactic acid production form the normal control erythrocytes, although considerable increases in the oxygen uptake were recorded.     

Membrane transport of non-transferrin-bound iron by reticulocytes

The transport of non-transferrin-bound iron into rabbit reticulocytes was investigated by incubating the cells in 0.27 M sucrose with iron labelled with 59Fe. In most experiments the iron was maintained in the reduced state, Fe(II), with mercaptoethanol. The iron was taken up by cytosolic, haem and stromal fractions of the cells in greater amounts than transferrin-iron. The uptake was saturable, with a Km value of approx. 0.2 microM and was competitively inhibited by Co2+, Mn2+, Ni2+ and Zn2+. It ceased when the reticulocytes matured into erythrocytes. The uptake was pH and temperature sensitive, the pH optimum being 6.5 and the activation energy for iron transport into the cytosol being approx. 80 kJ/mol. Ferric iron and Fe(II) prepared in the absence of reducing agents could also be transported into the cytosol. Sodium chloride inhibited Fe(II) uptake in a non-competitive manner. Similar degrees of inhibition was found with other salts, suggesting that this effect was due to the ionic strength of the solution. Iron chelators inhibited Fe(II) uptake by the reticulocytes, but varied in their ability to release 59Fe from the cells after it had been taken up. Several lines of evidence showed that the uptake of Fe(II) was not being mediated by transferrin. It is concluded that the reticulocyte can transport non-transferrin-bound iron into the cytosol by a carrier-mediated process and the question is raised whether the same carrier is utilized by transferrin-iron after its release from the protein.     

Intracellular distribution of ribonuclease activity during erythroid cell development.

Five ribonuclease activities, separable by polyacrylamide gel electrophoresis, have been detected in erythroid bone marrow cells from anaemic rabbits. Their intracellular distribution has been investigated and compared with that of the ribonucleases in reticulocytes. Both the acid and alkaline ribonuclease activities of reticulocytes are much lower (30--50 fold) than those of bone marrow erythroid cells. The most marked decrease in enzyme activity occurs in the fractions containing ribosomes and mitochondria plus lysosomes. In these subcellular organelles there was also a qualitative change in the ribonuclease electrophoretic pattern, whereas the cytosol enzymes of marrow erythroid cells and reticulocytes remained largely unchanged. Several ribonucleases released from reticulocyte membranes with urea were similar to those present in the lysosomal plus mitochondrial fraction, as shown by detection of enzyme activity after polyacrylamide gel electrophoresis. The decline in ribonuclease activity was found to begin in the orthochromatic cells, which have a highly condensed nucleus and are no longer active in DNA and RNA synthesis, and to coincide with a decrease in acid phosphatase activity and loss of lysosomes.     

Phosphorylase phosphatase activities of rat liver in streptozotocin-diabetes

Protein phosphatase-1 and 2A, accounting for all the hepatic activity regulating phosphorylase, were assayed in streptozotocin-induced (8 weeks) diabetic Wistar rats. Cytosolic protein phosphatase-1 and 2A were distinguished by chromatography on heparin-Sepharose and by inhibition with inhibitor-2. Approx. 25-35% increases in type-1 phosphorylase phosphatase activity measured in cytosols were registered in diabetic rats when compared with control and 24 h fasting animals. The enrichment of protein phosphatase-1 in the cytosol of streptozotocin-treated rat livers could not be attributed to the reduced glycogen content with the onset of diabetes, since this elevated level of type-1 phosphatase was not observed in fasting rats with low glycogen content. The translocation of type-1 phosphatase from the particulate fraction into the cytosol was also recorded in trypsin-treated samples of diabetic rat livers. The apparent molecular weight of type-1 phosphatase in the cytosol of control and fasted rats was 160,000 as judged by gel filtration. The type-1 phosphatase activity that was released from the particulate fraction by streptozotocin-induced diabetes identified a further enzyme species (Mr 110,000) in the cytosol. Our data imply that the higher levels of cytosolic protein phosphatase-1 in diabetic rat liver could be a consequence of the dissociation of the catalytic subunit of protein phosphatase-1 and the glycogen-binding subunit in rat livers.     

Iron transport mechanisms in reticulocytes and mature erythrocytes

The mechanism of iron transport into erythroid cells was investigated using rabbit reticulocytes and mature erythrocytes incubated with ⁵⁹Fe-labelled Fe(II) in isotonic sucrose or in solutions in which the sucrose was replaced with varying amounts of isotonic NaCl or KCl. Iron uptake was inhibited at all concentrations of NaCl, in a concentration-dependent manner, but with KCl inhibition occurred only at concentrations up to 10 mM. Higher KCl concentrations stimulated iron uptake to the cytosol of the cells, but inhibited its incorporation into heme. This effect became more marked as the iron concentration was raised. It was found that KCl inhibits iron incorporation into heme and stimulates iron uptake by mature erythrocytes, as well as by reticulocytes. It is concluded that erythroid cells can take up nontransferrin-bound Fe(II) by two mechanisms. One is a high-affinity mechanism that is limited to reticulocytes, saturates at a low iron concentration, and is inhibited by metabolic inhibitors. The other is a low-affinity process that is found in both reticulocytes and erythrocytes, becomes more prominent at higher iron concentrations, and is stimulated by KCl, as well as RbCl, LiCl, CsCl, and choline Cl. The KCl stimulation is inhibited by amiloride, but not by metabolic inhibitors, and its operation is not dependent on changes in cell volume or membrane potential, but it does require the presence of a permeant extracellular anion. Iron uptake by this process appears to occur by facilitated transport and is possibly assoicated with exchange of Na⁺. A further aspect of this study was a comparison of iron uptake by reticulocytes from Fe(II)-sucrose and Fe(II)-ascorbate using a variety of incubation conditions. No major differences were observed. © 1995 Wiley-Liss, Inc.