Search

Iron metabolism in animals is altered by haemolytic anaemia induced by phenylhydrazine (PHZ). In common with a number of other modulators of iron metabolism, the mode and the mechanisms of this response are yet to be determined. However, recent studies have shown increased expression of the ferrous transporter DMT1 in the duodenum and other tissues of mice administered PHZ. We examined the expression of the ferric reductase Dcytb, DMT1 and some other genes involved in Fe metabolism in tissues of mice dosed with PHZ. The expression of iron-related genes in the duodenum, liver, and spleen of the mice were evaluated using Northern blot analyses, RT-PCR and immunocytochemistry. Dcytb, and DMT1 mRNA and protein increased markedly in the duodenum of mice given PHZ. The efflux protein Ireg1 also increased in the duodenum of the treated mice. These changes correlated with a decrease in hepatic hepcidin expression. Dcytb, DMT1, Ireg1 and transferrin receptor 1 mRNA expression in the spleen and liver of mice treated with PHZ responded to the enhanced iron demand associated with the resulting stimulation of erythropoiesis. Enhanced iron absorption observed in PHZ-treated animals is facilitated by the up-regulation of the genes involved in iron transport and recycling. The probable association of the erythroid and the store regulators of iron homeostasis and absorption in the mice is discussed.

As a first step in the development of a reporter system for gene expression in halophilic archaea, a beta-galactosidase was purified 140-fold from Haloferax alicantei (previously phenon K, strain Aa2.2). An overproducing mutant was first isolated by UV mutagenesis and screening on agar plates containing X-Gal substrate. Cytoplasmic extracts of the mutant contained 25-fold higher enzyme levels than the parent. Purification of the active enzyme was greatly facilitated by the ability of sorbitol to stabilise enzyme activity in the absence of salt, which allowed conventional purification methods (e.g., ion-exchange chromatography) to be utilised. The enzyme was optimally active at 4 M NaCl and was estimated to be 180 +/- 20 kDa in size, consisting of two monomers (each 78 +/- 3 kDa). It cleaves several different beta-galactoside substrates such as ONP-Gal, X-Gal and lactulose, but not lactose, and also has beta-D-fucosidase activity. No beta-glucosidase, beta-arabinosidase or beta-xylosidase activity could be detected. The amino-acid sequence at the N-terminus and of four proteolytic products has been determined.

Cytosolic aconitase activity was assayed in duodenal mucosa from mice subjected to a variety of manipulations known to modulate duodenal iron status and duodenal iron absorption. No changes in cytosolic aconitase activity were observed 1 h after oral FeSO4 dosing or intramuscular desferrioxamine treatment. Three days of hypoxic exposure and two weeks treatment with intramuscular iron dextran also had no effect on cytosolic aconitase. Three weeks growth on an iron deficient diet significantly reduced cytosolic aconitase activity. In no situation was there any evidence for significant amounts of inactive aconitase which could be activated in vitro with FeSO4/cysteine. These data suggest that duodenal cytosolic aconitase is not sensitive to acute changes in mucosal iron levels and is generally much less sensitive to body iron status than is duodenal iron absorption. There is evidence that chronic iron depletion reduces cytosolic aconitase to a relatively small degree but generally activity is maintained, consistent with an important metabolic role for the enzyme.

The effects of solvent, pH and temperature on the 1H-NMR spectra of recombinant murine interleukin-6 (IL-6) are described. Assignments made from two-dimensional homonuclear spectra are presented for resonances of the fifteen aromatic amino-acid side chains. A time-dependent loss of intensity was observed for all resonances in the spectrum of IL-6, probably as a result of aggregation. This aggregation is markedly temperature-dependent. The pKa values of the four histidine residues in murine IL-6 has been measured; one has a value of 5.5, approx. one pH unit less than the value exhibited by the other three. Analysis of the NOESY spectra has allowed a preliminary characterisation of the nature of interactions among the aromatic side chains within the protein fold. 1H and 15N resonances of residues Thr-4 to Val-21 are assigned from three-dimensional 1H-15N correlated spectroscopy, and evidence is presented for these residues comprising a mobile N-terminal tail with little ordered structure. An N-terminal mutant lacking the first 22 residues of the murine IL-6 sequence and known to possess full biological activity was also examined and shown to have essentially retained the tertiary fold of the native molecule.

Serum from homozygous hypotransferrinaemic mice (a mixed group of males and females, aged 6-8 wk) was found to contain low levels of iron (mean 0.9 +/- 0.5 microM (SEM, n = 4), as assayed by conventional serum iron assays. Similarly, low levels of non-transferrin-bound iron were determined with a nitrilotriacetate chelation assay (1.3 +/- 0.4 microM, n = 4) (Singh, S., Hider, R.C. and Porter, J.B. (1990) Analytical Biochemistry 186, 320-323). Mononuclear Fe (citrate) was undectable by electron paramagnetic resonance spectroscopy (EPR). Significantly larger quantities of iron (16 +/- 5 microM, n = 8) were detected by the bleomycin assay (Gutteridge, J.M.C., Rowley, D.A. and Halliwell, B. (1981) Biochemical Journal 199, 263-265), while non-haem iron assay or atomic absorption spectrophotometry revealed up to 96 microM iron. Haemoglobin iron was detectable at approximately 10 microM by spectrophotometry. Ferri-haem was undetectable by EPR spectroscopy. Serum ferritin levels of 641 +/- 128 micrograms/l (n = 14) in hypotransferrinaemic mice (wild-types 44 +/- 6 micrograms/l, n = 14) were observed and these cannot account for the non-transferrin-bound iron. Hypotransferrinaemic mouse serum therefore contains large quantities of non-transferrin-bound iron which is unreactive in some assays used to detect such iron in human iron overload. Fractionation by Sephadex G200 chromatography revealed three distinct species with apparent molecular weights of > or = 150 kDa, 40-80 kDa and 1-5 kDa. The iron may be distinguished from known extracellular iron proteins and haem-proteins by its availability to hot acid extractions.

59Fe uptake rates by mouse duodenal fragments incubated in vitro were markedly reduced by non-permeable reagents, ferricyanide (oxidising agent) and ferrozine (Fe2+ chelator), in the medium; ferrocyanide had no effect. Reduction of Fe3+, as reflected by an increase in ferrozine-(Fe2+)-chelatable iron, was observed in the presence of the tissue fragments. The generation of Fe2+ occurred linearly with time, was independent of the medium ferrozine concentration, and was not due to release of reducing factors from the duodenal fragments. Fe(3+)-reducing activity was mainly present on the mucosal surface and was localised primarily to the proximal region of the small intestine. Changes in Fe3+ reduction rates closely parallelled the changes in duodenal 59Fe uptake, when metabolic inhibitors or modulators of membrane potential were included in the medium. The enhancement in duodenal mucosal 59Fe uptake in chronic hypoxic and iron-deficient mice parallelled the changes in the tissue reduction of medium Fe3+. Moreover, the rates of reduction were quantitatively similar to rates of uptake. These observations indicate that a sequential reduction and uptake process operates for Fe3+ uptake in mouse duodenum.

The basolateral membrane of mouse duodenal enterocytes can be selectively labelled in vitro with 59Fe by incubating intact enterocytes with 59Fe(III)-nitrilotriacetate at 0-4 degrees C. It has been proposed that this labelling represents binding to a site important in the transfer of intracellular Fe to the portal plasma (Snape, S., Simpson, R.J. and Peters, T.J. (1990) Cell Biochem. Funct. 8, 107-115). Studies presented here show binding to intact enterocytes in vitro was complete within 1 h and was proportional to enterocyte protein concentration. Binding to enterocytes isolated from both normal and chronically hypoxic mice showed a hyperbolic dependence on medium Fe(III) concentration, consistent with a single class of binding sites. Neither apparent binding constant nor maximal binding were increased by hypoxic exposure of mice, suggesting that the increased in vivo labelling of this site in hypoxia is not due to an increase in affinity or capacity of this site for iron. Release of iron from intact enterocytes, labelled at 0-4 degrees C, was measured at 37 degrees C and 0-4 degrees C. Release of 59Fe was extensive and more rapid at 37 degrees C with highest release to mouse serum. Iron released to serum was found to be bound to transferrin. Prior dialysis of serum against buffer led to complete failure of enterocytes to release iron. Reconstituting serum by adding back the dialysate restores release to levels seen in fresh serum, suggesting that low molecular weight serum components, notably bicarbonate, mediate iron transfer from the basolateral membrane to serum transferrin. The properties of the basolateral membrane iron binding site described here are consistent with a role in the iron transfer process.

Mouse intestinal brush-border membrane vesicles take up iron from media containing 59Fe3 +-nitrilotriacetic acid. The iron uptake by the vesicles represents accumulation of iron which relates to an osmotically active space. Uptake is linearly related to vesicle protein concentration and is inhibited by low incubation temperature and low medium free Fe3+ concentrations. Experiments with the lipid soluble iron ligand 8-hydroxyquinoline and with Triton X-100 imply that the uptake is rate limited by membrane transport.

Initial rates of mucosal uptake of Fe3+ from luminal Fe3+-nitrilotriacetate solutions by tied segments of mouse intestine in vivo have been measured. Duodenal uptake showed an approximately hyperbolic dependence of uptake on Fe3+ complex concentration (Km(app) 66 microM, Vmax 6.2 pmol/min per mg intestine) with little dependence on nitrilotriacetate:Fe3+ ratio or on added Ca2+. Duodenal uptake was greatly stimulated by hypoxic treatment of mice. Uptake rates by distal ileum were lower than by duodenum and more sensitive to added Ca2+. These results show that isolated duodenal brush-border membrane Fe3+ transport characteristics (Simpson, R.J. and Peters, T.J. (1984) Biochim. Biophys. Acta 772, 220-226) are inadequate to explain duodenal Fe3+ uptake in vivo. However, ileal uptake can be explained by the properties of isolated ileal brush-border membrane (Simpson, R.J., Raja, K.B. and Peters, T.J. (1985) Biochim. Biophys. Acta 814, 8-12).

A successful method for the analytical subcellular fractionation of mouse duodenal mucosa organelles was established. 59Fe(III)-nitrilotriacetate (pH 7.2) was injected into tied-off duodenal segments in vivo and, after 2-20 min, mucosal homogenates were subjected to subcellular fractionation. Radioactivity was recovered in the cytosolic fractions and in the gradient at a density of 1.18-1.20 g/ml. Enhanced iron absorption was achieved by placing the animals in a hypobaric chamber for 3 days. These animals had a higher proportion of particulate 59Fe compared to controls. Homogenisation in sucrose medium containing the selective plasma membrane perturbant digitonin shifted the particulate iron fraction to a higher density region of the gradient indicating a localisation of the iron binding site to the plasma membrane region of the mucosal cells. No significant radioactive iron was observed in the brush-border region of the gradient. Transferrin immunoreactivity was found only in the cytosolic region of the gradient and was not associated with any organelle.

Addition of iron-binding proteins (human serum transferrin, mouse serum transferrin, human lactoferrin) to the luminal fluid in tied-off segments of mouse intestine in vivo led to reduced 59Fe3+ absorption from 59Fe3+-nitrilotriacetate when compared to 59Fe3+-nitrilotriacetate alone. Assay of transferrin in luminal fluid from tied segments revealed only trace amounts of immunoreactivity. The levels of luminal transferrin are unaltered in chronic hypoxia where iron absorption is significantly enhanced. Studies in vitro revealed that NH4Cl, dansylcadavarine, para-chloromercuribenzoate and trinitrobenzenesulphonate have no effect on initial 59Fe3+ uptake rates from 59Fe3+-nitrilotriacetate, while N-ethylmaleimide (1 mM) caused a 40% inhibition. In vivo 59Fe3+ uptake was unaffected by preincubation of tied-off segments with colchicine (5 mM) for up to 2 h. These results suggest that receptor-mediated endocytosis of transferrin is not a significant mechanism in the uptake of luminal Fe3+ by mouse duodenum.

Ovine insulin-like growth factors I and II (oIGF-I and oIGF-II) have been purified from adult sheep serum. oIGF-II-like receptor-binding activity and IGF-I-like immunoactivity were enriched on SP-Sephadex C-25, then purified using HPLC in the presence of a variety of counter ions. IGF-I- and IGF-II-like activities were separated using HPLC in the presence of 0.2% tetrabutylammonium phosphate at pH 7.0. The final recovery of oIGF-I was 82.6 micrograms from 3.2 litres of adult sheep serum (a yield of 17.6%), and the recovery of oIGF-II was 388 micrograms (a yield of 13.3%). Both IGF preparations were considered to be homogeneous as judged by single sharp peaks during analytical HPLC, and unique N-terminal amino acid sequences. Purified ovine IGFs had molecular weights similar to that of other IGFs (approximately 7000), and the first 30 N-terminal amino acids of both peptides were identical to their human counterparts. The isoelectric points of oIGF-I (pI approximately 8.2) and oIGF-II (pI approximately 6.8) were similar to those of human (h) IGFs (hIGF-I pI approximately 8.2; hIGF-II pI approximately 6.5), and the overall amino acid content of the ovine IGFs was also similar to that of IGFs from other species. oIGF-II preparations from fetal sheep and from adult sheep appeared to be identical. The isolation procedure represents one of general utility that can be easily modified to facilitate the isolation of recombinant IGFs from culture fluid.

Initial Fe3+ uptake rates by mouse intestinal fragments were determined in vitro. Uptake was dependent primarily on the Fe3+-nitrilotriacetate complex concentration. Addition of Ca2+ and Mg2+ to the incubation medium had only small effects on the Fe3+ uptake rate. Duodenal fragments from hypoxic animals showed enhanced uptake of Fe3+; this increase was more pronounced with a divalent cation-free medium. Ca2+ markedly diminished the Fe3+ uptake by mucosa from hypoxic mice; Mg2+ had no appreciable effect. Distal ileal fragments exhibited lower uptake rates compared to the duodenum, but were more sensitive to the effects of added Ca2+. The ileal fragments did not show an adaptive response of Fe3+ uptake to hypoxia. These results suggest the existence of more than one pathway for mucosal Fe3+ uptake. One pathway, sensitive to Ca2+ and not stimulated by hypoxia, may be present in the duodenum and ileum. A second pathway, inhibited by Ca2+ and exhibiting an adaptive response to hypoxia, occurs only in the duodenum. This latter pathway is more sensitive to the effects of metabolic inhibitors.

Recent studies of Fe2+ uptake by mouse proximal intestine brush-border membrane vesicles revealed low-affinity, NaCl-sensitive and high-affinity, NaCl-insensitive, components of uptake (Simpson, R.J. and Peters, T.J. (1985) Biochim. Biophys. Acta 814, 381-388). In this study, the former component is demonstrated to show a strong pH dependence with an optimum of pH 6.8-6.9. Studies at pH 6.5, where the low affinity component is inhibited by more than 25-fold compared with pH 7.2, suggest that the pH-sensitive component represents transport across the brush-border membrane followed by intravesicular binding. Cholate extracts of brush-border membrane vesicles contain pH- and NaCl-sensitive Fe2+ binding moieties which may be involved in the transfer of Fe2+ across the intestinal brush-border membrane and subsequent binding inside the vesicles. Fe2+ uptake by brush-border membrane vesicles from the duodenum of hypoxic mice is higher than uptake by vesicles from control-fed animals, suggesting the existence of a regulable brush-border membrane Fe2+ carrier.

Fe2+ uptake by mouse intestinal brush-border membrane vesicles consists of two components: a rapid, high affinity (Kd less than 1 microM), low capacity binding (less than 2 nmol/mg protein), presumably to the outside of the vesicles, and a second, large capacity component with an initial rate showing a hyperbolic dependence on medium Fe2+ (Km 35-90 microM). The latter, predominant process is relatively independent of medium ascorbate: Fe2+ ratio, is inhibited by Co2+ and Mn2+ but varies greatly from one membrane preparation to another. This component is strongly inhibited by large extravesicular NaCl and KCl concentrations and may represent transport into the vesicles. No significant change in uptake could be observed in vesicles prepared from hypoxic mice.

The NAD-dependent conversion of malate to lactate in human erythrocytes was studied by spin echo proton NMR. A pathway involving the decarboxylation of oxaloacetate catalysed by haemoglobin is proposed to account for the observed reaction. NADP-dependent reaction was negligible. The rate of the reaction was measured in intact erythrocytes under controlled conditions. This rate correlates with that obtained with lysates at 30 microM free NAD and that obtained with purified human erythrocyte enzymes at about 15 microM NAD. The total extractable NAD in the intact cells was 70-90 microM. Experiments with cells containing elevated NAD levels could be explained by a significant fraction of the NAD being weakly bound (Kd about 1 mM) to haemoglobin.

Concentrated human erythrocyte lysates (greater than 99% initial haematocrit) were subjected to high centrifugal fields. This caused the membrane fraction to separate from the cytoplasmic portion, due to the lower density of the former. Enzyme distribution data indicated that glyceraldehyde-phosphate dehydrogenase was predominantly in the cytoplasmic fraction.

In vivo kinetics of mucosal uptake of luminal 59Fe2+ by tied segments of normal mouse duodenum are characterised by a Km of approx. 100 microM and a Vmax of approx. 9 pmol/min per mg wet weight of intestine. These values were determined at pH 7.25 in the presence of excess sodium ascorbate. Studies with luminal Fe2+ concentrations of 100 microM reveal: uptake is relatively independent of ascorbate: Fe ratio and luminal pH and uptake is potently inhibited by 1 mM Co2+ or Mn2+ and large luminal NaCl concentrations but not by Ca2+. 3 days of hypoxia (0.5 atmospheres) yields no significant increase in subsequent total mucosal uptake by in vivo tied segments while uptake is significantly reduced by semi-starvation. Quantitative comparison of in vivo mucosal uptake with subsequent determination of isolated brush-border membrane 59Fe2+ transport in individual mice reveals a positive correlation (P less than 0.01) between the two parameters. These results, in conjunction with studies of isolated mouse duodenal brush-border membrane (Simpson, R.J. and Peters, T.J. (1985) Biochim. Biophys. Acta, 814, 381-388 and (1986) Biochim. Biophys. Acta 856, 109-114) suggest that the Fe2+ transport properties of isolated brush-border membrane are quantitatively adequate to explain in vivo mucosal uptake in normal and hypoxic mice at Fe2+ concentrations up to 100 microM.

Rabbit duodenal brush-border membrane contains chloroform/methanol (2:1, v/v) extractable Fe-binding lipids (27.2 +/- 6.7 nmol/mg protein, mean +/- S.E. (n = 5)). Thin-layer chromatography in two solvent systems reveals that the major Fe-binding component(s) co-migrate with free fatty acids. Fe-binding by pure lipids reveals that phosphatidic acid, phosphatidylserine, oleic and stearic acids all show apparent Fe-binding in filtration assays, although oleic acid shows the highest apparent binding (5-10-fold) on a molar basis. The free fatty acid content of brush-border membrane vesicles is sufficient to account for the chloroform/methanol extractable Fe-binding observed in vesicle preparations. The pH dependence of Fe-binding by oleic acid is similar to that reported for the detergent extractable Fe-binding lipid which has been implicated in transport of Fe from Fe/ascorbate solutions by rabbit duodenal brush-border membrane vesicles (Simpson, R.J. and Peters, T.J. (1986) Biochim. Biophys. Acta 859, 227-236).

Fe2+ uptake by brush-border membrane vesicles from rabbit duodenum has been investigated and found to show similar qualitative properties to those previously demonstrated with mouse proximal intestine brush-border membrane vesicles (Simpson, R.J. and Peters, T.J. (1986) Biochim. Biophys. Acta 856, 109-114). In particular, a relatively low affinity (Km(app) approx. 83 microM), NaCl and pH sensitive transport component is present. The disruption of 59Fe2+-laden vesicles with sodium cholate, followed by gel filtration or centrifugal analysis reveals that cholate insoluble material (Mr greater than 10(6)) is the major destination for 59Fe2+ taken up by intact vesicles. Analysis of cholate extracts for Fe2+ binding ability reveals a single high-capacity (49.8 +/- 15.6 nmol/mg vesicle protein (S.E., n = 3)), high-affinity (Kd(app) less than 5 microM) binding component with an Mr equivalent to approx. 10(4) on gel filtration in the presence of cholate. This binding component is extracted into chloroform/methanol (2:1, v/v) is relatively heat and protease resistant and thus appears to be a lipid.

Iron uptake from Fe/ascorbate by mouse brush-border membrane vesicles is not greatly inhibited by prior treatment with a variety of protein-modification reagents or heat. Non-esterified fatty acid levels in mouse proximal small intestine brush-border membrane vesicles show a close positive correlation with initial Fe uptake rates. Loading of rabbit duodenal brush-border membrane vesicles with oleic acid increases Fe uptake. Depletion of mouse brush-border membrane vesicle fatty acids by incubation with bovine serum albumin reduces Fe uptake. Iron uptake by vesicles from Fe/ascorbate is enhanced in an O2-free atmosphere. Iron uptake from Fe/ascorbate and Fe3+-nitrilotriacetate (Fe3+-NTA) were closely correlated. Incorporation of oleic acid into phosphatidylcholine/cholesterol (4:1) liposomes leads to greatly increased permeability to Yb3+, Tb3+, Fe2+/Fe3+ and Co2+. Ca2+ and Mg2+ are also transported by oleic acid-containing liposomes, but at much lower rates than transition and lanthanide metal ions. Fe3+ transport by various non-esterified fatty acids was highest with unsaturated acids. The maximal transport rate by saturated fatty acids was noted with chain length C14-16. It is suggested that Fe transport can be mediated by formation of Fe3+ (fatty acid)3 complexes.

Studies of 59Fe3+ uptake by brush-border membrane vesicles prepared from mouse duodenum have indicated that uptake represents transport across the brush-border membrane which is rate-limited by the membrane-transfer step (Simpson, R.J. and Peters, T.J. (1984) Biochim. Biophys. Acta 772, 220-226). Further studies presented here reveal that the uptake rate represents the net influx rate for Fe3+ and is independent of Na+ in the medium and of the method of vesicle preparation. Uptake by brush-border membrane vesicles prepared from mouse distal ileum also represents predominantly transport and is higher than that observed with duodenal brush-border membrane vesicles. Studies of the initial uptake rate by vesicles prepared from normal and hypoxic mouse intestine demonstrated an increase in Fe3+ transport in duodenal vesicles only.

The kinetics of inhibition of human erythrocyte glyceraldehyde-3-phosphate dehydrogenase by iodoacetate were studied in the intact cell and in vitro. The kinetics were determined using 1H-NMR to follow solvent exchange of 1H and 2H at the C-2 position of lactate. The exchange occurs via a series of enzyme-catalysed reactions, including that catalysed by glyceraldehyde-3-phosphate dehydrogenase. A direct assay with quenching of the inhibition was also used to check the results. Iodoacetate was shown to act as an active site-directed inhibitor of the dehydrogenase. The enzyme inhibition patterns, which are characterised by a binding step and a kinetic step, are similar in situ and in vitro. Membrane binding, however, was found to alter the inhibition pattern for the enzyme in vitro.

Fatty acids can form lipid-soluble complexes with Fe2+. Incorporation of fatty acids into phosphatidylcholine/cholesterol liposomes renders them permeable to Fe2+. Of several fatty acids tested, the most effective Fe2+ carriers were linoleic and oleic acids followed, in decreasing order of efficacy, by linolenic, myristic, arachidonic and palmitic acids. The initial Fe2+ transport rate for oleic acid depends on free Fe2+ in the medium which in turn shows a strong pH dependence above pH 7.0. The overall pH dependence of Fe2+ transport for several fatty acids shows an optimum near 6.9. Fe2+ transport catalysed by oleic acid can be inhibited by high NaCl concentrations but not 1 mM Co2+. It is suggested that free fatty acids may act as mediators of Fe2+ transport across biological membranes, particularly isolated intestinal brush-border membrane.

Initial rates of 59Fe3+ uptake by mouse duodenal fragments (in vitro) and tied-off duodenal segments (in vivo) have been characterised for control and hypoxic animals. 59Fe3+ uptake by duodenal fragments was rapid, selective and dependent on medium Fe3+-nitrilotriacetate concentration. Most of the 59Fe3+ uptake (70-75%) occurred via the mucosal route and was dependent on the metabolic state of the tissue. Mucosal uptake showed an adaptive increase following exposure of animals to 3 days hypoxia; the enhancement was due to a 2-3-fold increase in Vmax app, without any significant changes in the Km app. Studies of upper small intestine transit times showed a mean residence time of 4-5 min for 59Fe-labelled mouse chow, emphasising the importance of initial uptake measurements. Time courses for in vivo total mucosal uptake exhibited linearity over a wide variety of absorption rates after correction for the permeation by intact metal-chelate complex. The corrected uptake showed a hyperbolic dependence on medium Fe3+-nitrilotriacetate concentration. Kinetic studies revealed a 2-3-fold increase in total mucosal uptake in hypoxia. Mucosa-to-carcass transfer of 59Fe was also markedly increased by chronic hypoxia. The in vitro system exhibits similar qualitative and quantitative kinetics for Fe3+ transport via the mucosal membrane to those obtained in vivo. The results observed in vitro are thus valid and provide a convenient method for further studies on Fe3+ transport in animals and in man.

Intestinal iron uptake by mouse duodenal fragments is inhibited in the absence of oxygen and glucose from the incubation medium and by a variety of metabolic inhibitors. The mechanism of energy coupling to iron uptake is, however, unclear. In vitro experiments using duodenal fragments showed Fe3+ uptake to be markedly inhibited, in a reversible fashion, by the replacement of incubation medium Na+ by K+. Addition of phloridzin to the medium failed to affect iron uptake, suggesting that the above effect was not a consequence of reduced glucose uptake. Substitution of Na+ by Rb+ also potently reduced duodenal iron uptake. Replacement of medium NaCl by either mannitol or choline chloride had no significant effect on Fe3+ uptake, thus excluding the possibility of the Fe3+ uptake process being Na+-dependent. Similar observations were made with duodenal fragments from animals with enhanced Fe3+ absorption, due to chronic hypoxia. Valinomycin (1-5 microM) increased the uptake of both glucose and Fe3+. Higher concentrations (22.5 microM) of the ionophore were inhibitory. In vivo studies (tied-off segments) using Rb+-containing medium confirmed the inhibitory effects of univalent cations on Fe3+ absorption. Enhanced absorption of Fe3+ was also demonstrable in vivo, with low concentrations of valinomycin and nigericin added to the luminal medium. These observations suggest that the Fe3+ uptake process may be dependent on the brush-border membrane potential.