Role of Ceruloplasmin in Cellular Iron Uptake

See allHide authors and affiliations

Science  30 Jan 1998:
Vol. 279, Issue 5351, pp. 714-717
DOI: 10.1126/science.279.5351.714


Individuals with hereditary ceruloplasmin (Cp) deficiency have profound iron accumulation in most tissues, which suggests that Cp is important for normal release of cellular iron. Here, in contrast to expectations, Cp was shown to increase iron uptake by HepG2 cells, increasing the apparent affinity for the substrate by three times. Consistent with its role in iron uptake, Cp synthesis was regulated by iron supply and was increased four- to fivefold after iron depletion. Unlike other iron controllers that are posttranscriptionally regulated, Cp synthesis was transcriptionally regulated. Thus, iron-deficient cells could increase Cp synthesis to maintain intracellular iron homeostasis, so that defects would lead to global accumulation of iron in tissues.

Iron is essential for many different biological processes, often functioning as a protein-bound redox element. However, iron in excess of cellular needs is extremely toxic, and its levels are precisely regulated (1). Defective regulation due to hereditary hemochromatosis or secondary iron overload leads to hepatic iron excess and injury, most likely due to iron-stimulated free radical reactions (2). Alterations of iron pools are implicated in neurodegenerative disease, aging, microbial infection, atherosclerosis, and cancer (3). The balance required to maintain appropriate intracellular iron concentration has led to the utilization of multiple mechanisms that regulate, primarily at the posttranscriptional level, the synthesis of iron transport and storage proteins (4).

Ceruloplasmin (Cp) is a 132-kD monomeric copper oxidase implicated in iron metabolism because of its catalytic oxidation of Fe2+to Fe3+ (ferroxidase activity) (5). Also, copper-deficient swine develop anemia that is overcome by Cp injection (6-8). Cp ferroxidase activity accelerates iron incorporation into apo-transferrin (9, 10), which may cause a negative concentration gradient of iron and cellular iron release. Patients with aceruloplasminemia have hemosiderosis characterized by low serum iron, high serum ferritin, massive iron deposition in multiple organs, and neurological deficit and diabetes (11-13), further implicating Cp in release of iron from tissue (14-16). In addition, multicopper oxidases homologous to Cp [fet3 in Saccharomyces cerevisiae (14, 17, 18) and fio1 in S. pombe (19)] facilitate high-affinity iron uptake by yeast. Thus, the yeast oxidases and Cp may promote iron fluxes but in opposite orientations with respect to the cell surface (15). The present studies were initiated to determine the role of Cp in cellular iron efflux, but activation of iron uptake was found.

The effect of Cp on iron transport was examined in HepG2 cells, a human hepatocellular carcinoma line exhibiting transferrin-independent and -dependent iron uptake (20). Iron release was measured from HepG2 cells preloaded with iron by incubation with55Fe-nitrilotriacetic acid (NTA). Cp did not significantly alter iron release by either iron-sufficient cells or by cells made iron-deficient by incubation with the Fe2+chelator bathophenanthroline disulfonate (BP) (Fig.1A). In view of these negative results and the known role of the Cp homolog fet3 in iron uptake by S. cerevisiae, we tested whether Cp stimulated iron uptake rather than release. Cp did not alter iron uptake in untreated, iron-sufficient HepG2 cells, but it approximately doubled uptake by iron-deficient cells (Fig. 1B). A half-maximal increase in iron uptake was observed at 5 μg of Cp per milliliter and a maximal increase was observed at 10 to 30 μg/ml, which are concentrations well below the normal range of 210 to 450 μg/ml in healthy adult human plasma. This observation may account for the apparently normal iron metabolism in patients with Wilson's disease who have lower than normal levels of plasma Cp and for iron deficiency in patients with levels less than 5% of normal (21). The increase in iron uptake by Cp was rapid, was not accompanied by an observable lag, and was linear for at least 40 min (22). Cp decreased the apparent Michaelis constant of iron uptake from about 1.2 μM to about 0.4 μM iron without altering maximal uptake (Fig. 1B, insert).

Figure 1

Effect of Cp on iron flux in HepG2 cells. (A) Iron release (30) from confluent HepG2 cultures maintained in iron-sufficient Dulbecco's modified Eagle's medium–Ham's F12 (DME/F12) medium (□) or iron-depleted by treatment with 1 mM BP for 12 hours (○). (B) 55Fe-NTA uptake (31) by HepG2 cells made iron-deficient by incubation for 16 hours with 1 mM BP (▪) or maintained in iron-sufficient DME/F12 medium (□). Inset: Iron-deficient HepG2 cells were incubated with 0.01 to 10 μM55Fe-NTA and with Cp (30 μg/ml) (•) or medium alone (○). Likewise, iron-sufficient cells were incubated with (+) or without (×) Cp. (C) HepG2 cells were pretreated with cycloheximide (CHX; 10 μg/ml) for 1 hour and then made iron-deficient by BP treatment for 12 hours. Iron uptake was measured in the presence of purified human Cp (30 μg/ml). (D) Cp or transferrin (Tf) loaded with 55Fe (32) was preincubated with rabbit anti-human IgG (1:1000; Accurate, Westbury, New York; black bars) or control IgG (1:1000; Accurate; dense striped bars) for 45 min and was added to cells pretreated with the same antibodies.55Fe-NTA uptake by iron-deficient HepG2 cells incubated with medium (control) or with Cp (100 μg/ml), and55Fe-transferrin (30 μg/ml) uptake, were measured after 15 min. (E) Iron-deficient HepG2 cells were preincubated for 60 min in medium alone (light striped bars), with monoclonal antibody (mAb) to the transferrin receptor (TfR) ?(2 μg/ml; Zymed, South San Francisco, California; black bars), or with an irrelevant antibody (2 μg/ml; dense striped bars). Iron uptake was measured after 15 min. (F) Iron-deficient (right) or iron-sufficient (left) HepG2 cells were allowed to condition their medium for 4 hours (open striped bars). In some wells, the conditioned medium was replaced after 4 hours with fresh medium before addition of 55Fe (dense striped bars). In other wells, rabbit polyclonal anti–human Cp IgG (1:500; black bars) or a control IgG (1:500; light striped bars) was added to the conditioned medium 1 hour before the completion of the 4-hour incubation and55Fe uptake was measured.

The inability of iron-sufficient cells to accumulate iron, even in the presence of Cp, suggested that a cellular component induced (or activated) by iron deficiency was required. Pretreatment of cells with cycloheximide prevented stimulation by Cp, which showed that a newly synthesized cellular protein or proteins are needed (Fig. 1C). One possible mechanism is that transferrin secreted by HepG2 cells (23) is iron-loaded through Cp ferroxidase activity (10) and carries iron into the cell via the transferrin receptor, which is induced by iron deficiency (4). Three experiments were done to show that Cp-mediated iron uptake did not use the transferrin pathway. Antibody to transferrin did not decrease Cp-mediated iron uptake by HepG2 cells; in a positive control, the antibody inhibited uptake of exogenous 55Fe-transferrin (Fig. 1D). Similar results were obtained with a monoclonal antibody that blocked transferrin receptor endocytosis (24) (Fig.1E). Finally, Cp enhanced iron uptake by iron-deprived K562 cells (22) that do not produce transferrin (25). Thus, the observed Cp-mediated iron uptake was transferrin-independent. Alternatively, the need for a newly synthesized cellular component is consistent with use of an inducible iron transporter analogous to ftr1p of S. cerevisiae (26).

Because HepG2 cells secrete Cp constitutively at a high rate (27), the role of endogenously produced Cp in iron uptake was examined. HepG2 cells were allowed to condition their medium for 4 hours to permit extracellular accumulation of Cp. Iron uptake by iron-depleted cells was two times more than uptake by iron-sufficient cells (Fig. 1F). Removal of the conditioned medium before measurement of iron uptake inhibited about 60 to 70% of the increase in iron-deficient cells, which suggests a role for a stable secreted factor. Addition of polyclonal rabbit anti-human Cp immunoglobulin G (IgG) (but not purified IgG) to the conditioned medium decreased iron uptake almost to the level seen when the conditioned medium was removed, which shows that cell-derived Cp was responsible for most of the stimulatory activity secreted by iron-deficient HepG2 cells. Thus, iron uptake by HepG2 cells was mediated by cell-derived Cp in cooperation with a cell component induced by iron deficiency.

We investigated whether the synthesis of Cp, like that of other iron controllers (4), was regulated by cellular iron content. Iron depletion of HepG2 and Hep3B (22) cells by treatment with ferrous (BP, Fig. 2A) or ferric [desferrioxamine (DF), Fig.2B] ion chelators increased Cp secretion four- to fivefold, according to immunoblot analysis using antibody to human Cp. Incubation of HepG2 cells with the copper chelator bathocuproine was completely ineffective (Fig. 2C), showing ion specificity. Going in the other direction, incubation of HepG2 cells with FeCl3 decreased Cp synthesis by up to 50% as compared with that of control cells (Fig. 2D). The direction of these regulatory processes is consistent with a role for Cp in iron uptake and maintenance of cellular iron homeostasis. Based on the mechanisms known to regulate iron-dependent synthesis of the transferrin receptor and ferritin, Cp synthesis may be expected to be regulated by analogous posttranscriptional mechanisms (4). After a 4-hour lag period, treatment of HepG2 cells with iron chelators increased the steady-state level of the two major Cp transcripts two- and fourfold after 8 and 16 hours, respectively (Fig.3A). The lag suggested that new synthesis of cellular proteins was required; this was verified by the ability of the protein synthesis inhibitor cycloheximide to completely block Cp mRNA induction by iron chelators (Fig. 3B). The possibility that iron depletion resulted in increased Cp mRNA stability, analogous to the regulation of the transferrin receptor (4), was examined. Cp mRNA accumulation was induced by iron chelation, actinomycin D was added to block new transcription, and the time course of mRNA decay was determined. Cp mRNA was extremely stable, exhibiting a decay of less than 15% during 10 hours (Fig. 4, A and D). Neither BP (Fig. 4, B and D) nor DF (Fig. 4, C and D) significantly enhanced Cp mRNA stability, which suggests that the effect was transcriptional. A nuclear run-on assay indicated at least a fivefold increase in the rate of Cp transcription by both chelators (Fig. 4E).

Figure 2

Regulation of Cp synthesis by iron depletion. Confluent HepG2 cells in six-well tissue culture clusters were treated with (A) BP, (B) DF, (C) bathocuproinedi sulfonate (BC), and (D) FeCl3for 16 hours in serum-free DMEM/F12 medium. Aliquots of conditioned media [50 μl for (A) and (B), 100 μl for (C) and (D)] were subjected to 7% SDS–polyacrylamide gel electrophoresis and transferred onto Immobilon-P membranes (Millipore, Bedford, Massachusetts). The membranes were incubated with rabbit anti–human Cp IgG (1:20,000; Accurate) as primary antibody, then with peroxidase-conjugated secondary antibody (1:10,000; Boehringer Mannheim, Indianapolis, Indiana), and developed with the use of ECL (Amersham). Purified human Cp (Cp; 30 ng) was used as standard; the intact 132-kD protein is indicated by an arrow.

Figure 3

Regulation of Cp mRNA steady-state levels by iron chelators. (A) Confluent HepG2 cells (5 × 106 cells) were treated for 4, 8, or 16 hours with 1 mM BP or DF or with medium (Med.) alone. Total RNA was harvested with RNAzol, and equal amounts of RNA (20 μg) were fractionated by formaldehyde agarose gel electrophoresis (1%) and then blotted on to nylon membranes. The blot was probed first with a 646–base pair fragment (Bst X1 and Bam H1) of a Cp cDNA labeled by random priming (top). Subsequently, the blot was stripped and reprobed with a random prime-labeled human GAPDH cDNA probe as a control (bottom). (B) Confluent HepG2 cells were pretreated for 1 hour with CHX (10 μg/ml), then with 1 mM BP or DF for an additional 14 hours. Total RNA was collected and subjected to RNA blot analysis; the blot was probed with a human Cp cDNA (top), and equal RNA loading was shown by the 28S ribosomal RNA bands (bottom).

Figure 4

Transcriptional regulation of Cp in iron-deficient HepG2 cells. Confluent HepG2 cultures were treated for 10 hours with medium alone (A), with 1 mM BP (B), or with 1 mM DF (C). Actinomycin D (5 μg/ml) was added, and the cells were harvested after 4, 7, and 10 hours. Total cellular RNA was isolated, and 20 μg of RNA were fractionated on an agarose gel, transferred to nylon membrane, and hybridized with human Cp cDNA (top) and with a human GAPDH cDNA probe (bottom). (D) Cp mRNA was normalized to GAPDH after quantitation by densitometry. Control cells (•) and cells treated with BP (○) or with DF (▵) are shown. (E) Confluent HepG2 cells were treated with 1 mM BP or DF for 8 hours, and Cp transcription was measured by nuclear run-on assay (33).

The finding that Cp stimulates cellular iron influx was surprising in the context of recent studies of aceruloplasminemia patients. One possible explanation is that defective Cp-mediated iron uptake by the liver results in plasma iron accumulation [possibly in low-molecular-weight iron complexes, as in hereditary (28) and secondary (29) hemochromatosis] to a level that drives nonspecific uptake by most tissues. The stimulation of iron uptake by Cp in hepatic cells is consistent with the uptake system described inS. cerevisiae and S. pombe. Those results are consistent with a model in which iron depletion of normal hepatocytes causes increased transcription of Cp that interacts with an iron transporter (also up-regulated by iron depletion), resulting in increased iron influx. The iron depletion requirement for Cp-stimulated Fe uptake, and the inhibition by cycloheximide, are consistent with a requirement for an inducible iron transporter. Taken together, the experiments in yeast and hepatic cells demonstrate a remarkable evolutionary conservation of the mechanisms that underlie the pathway controlling eukaryotic iron metabolism by copper. However, there are noteworthy differences: in yeast, the Cp homolog is a membrane protein that is co-transported to the cell surface and is in continuous contact with the iron transporter, whereas in mammalian tissues Cp is a secreted protein, and any interaction with a transporter is likely to be transient. This difference may be related to paracrine requirements for Cp in multicellular organisms.

  • * To whom correspondence should be addressed. E-mail: foxp{at}


View Abstract

Stay Connected to Science

Navigate This Article