Phorbol esters stimulate non-transferrin iron uptake by K562 cells.

Characterization of non-transferrin (non-Tf) iron transport by K562 cells has revealed unique properties relative to iron uptake mechanisms present in other cell types (Inman, R. S., and Wessling-Resnick, M.(1993) J. Biol. Chem. 268, 8521-8528). Since treatment of K562 cells with phorbol esters promotes stable megakaryocytic differentiation, we examined the uptake of non-Tf iron in response to protein kinase C activation. Treatment of K562 cells with phorbol esters increased the cellular uptake of Fe 4-6-fold compared with untreated cells. The phorbol ester-induced stimulation of Fe uptake was time- and dose-dependent, with significantly enhanced transport observed only after prolonged administration of 50 nM phorbol 12,13-dibutyrate (>8 h). These effects can be attributed to an increased Vmax of transport (14.0 ± 5 versus 0.6 ± 0.2 pmol/min/106 cells) as well as a 6-fold increase in the apparent K (1.2 ± 0.4 versus 0.2 ± 0.06 μM). It is thought that the reduction of Fe to Fe is required as a first step in the uptake mechanism, and the associated ferrireductase activity of K562 cells is also enhanced with phorbol ester treatment by 5-10-fold (337 ± 53 versus 43 ± 3 pmol/min/106 cells). Bryostatin-1, a protein kinase C activator that fails to induce differentiation of K562 cells, did not promote this effect, indicating that the enhanced transport activity is dependent on the differentiation response. The idea that synthesis of a new class of transporters is responsible for this effect is supported by the observation that actinomycin D blocks up-regulation of non-Tf iron transport. The increased transport and ferrireductase activity induced upon differentiation also correlate with the appearance of saturable iron-binding sites on the surface of K562 cells with K 0.4 μM. These results indicate that non-Tf iron transport activity and the expression of cell-surface iron-binding proteins can be controlled by environmental factors that promote megakaryocytic differentiation of K562 cells.

ent iron transport systems for many different cell types suggest that there may be two classes: "high" K m (5-30 M) transport has been described in studies of fibroblasts (1,2), HeLa cells (1,2), Chinese hamster ovary cells (1), hepatocytes (3), HepG2 cells (4), and L1210 cells (5), while "low" K m (0.3-0.5 M) iron transport has been observed for reticulocytes (6,7) and K562 cells (8). Our knowledge about the factors that regulate the appearance of these alternative pathways for iron acquisition is limited, but it has been reported that non-Tf iron uptake by fibroblasts can be up-regulated upon exposure to heavy metals (2), suggesting that environmental factors can influence iron assimilation through Tf-independent mechanisms. In contrast, modulation of Tf receptor activity is well documented and, in particular, has been found to be associated with cancer cell differentiation in Friend, M1, HL-60, and K562 cell leukemias (9 -15). Among other chemical inducing agents, administration of tumor-promoting phorbol esters is known to promote a rapid down-regulation of surface Tf receptors, with an associated decline in receptor synthesis upon prolonged exposure (Ͼ12 h) (16,17). For K562 cells, such alterations are accompanied by morphological and functional changes due to megakaryocytic differentiation, including the expression of glycoprotein IIIa (a platelet-specific antigen), the loss of glycophorin A (an erythrocyte-specific lineage antigen), increased platelet peroxidase activity, and secretion of granulocyte-macrophage colony-stimulating factor and interleukin-6 (18). In this report, we examine whether or not phorbol esters can also influence the properties of the non-Tf iron transport system of K562 erythroleukemia cells under these conditions. Our results indicate that while Tf-mediated delivery is down-regulated when K562 cells are induced to differentiate, non-Tf iron delivery is up-regulated. Thus, these two different iron uptake mechanisms may be coordinately regulated.

MATERIALS AND METHODS
Iron Transport Measurements-Iron was complexed to nitrilotriacetic acid (NTA) with a solution of 5.73 mM 55 FeCl 3 prepared in the presence of 4-fold molar excess of NTA in 20 mM HEPES, 20 mM Tris, pH 6.0, 100 mM NaCl and brought to neutral pH by titration with 10 N NaOH. K562 cells (grown in ␣-minimal essential medium with 7% fetal bovine serum) were collected by centrifugation at 500 ϫ g for 5 min, washed three times with PBS, and resuspended in uptake buffer (25 mM HEPES, pH 7.4, 150 mM NaCl, 1 mg/ml dextrose). Transport assays were performed in triplicate with 1-10 ϫ 10 6 cells/ml. Iron uptake was initiated by adding 55 FeNTA (75 nM to 1 M) to K562 cells in uptake buffer with a total volume of 300 l. After a 5-min incubation at 4 or 37°C, 250-l aliquots were removed, added to 750 l of ice-cold quench buffer (1 mM FeNTA, 25 mM HEPES, pH 7.4, 150 mM NaCl), and left on ice for 20 min. 950 l of the quenched assay mixture was then filtered onto nitrocellulose discs and washed twice with 3 ml of 150 mM NaCl. Cell-associated radioactivity was determined by dissolving the filters in 10 ml of scintillation fluid in borosilicate glass vials and liquid scintillation counting. The cell-associated radioactivity measured at 4°C was subtracted from that measured at 37°C to obtain the specific transport of 55 Fe. Results of experiments are reported Ϯ S.E., and presented data are representative of results obtained on at least three separate occasions.
* This work was supported in part by Grant DK95737 from the National Institutes of Health and Grant CB15 from the American Cancer Society. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
Ferrireductase Assay-K562 cell-mediated reduction of ferricyanide to ferrocyanide was measured using the method of Avron and Shavit (19) to detect the reduced product. PBS-washed K562 cells were resuspended in Hanks' buffer (2-4 ϫ 10 6 cells/ml). The ferrireductase assay was carried out upon the addition of potassium ferricyanide (50 M final concentration, 1-ml total volume) with incubation for 30 min at 37 or 4°C. The reaction was stopped by chilling the cells on ice and immediate centrifugation. A 700-l aliquot of supernatant was transferred into a spectrophotometer cuvette to which the following were added: 100 l of 3 M sodium acetate, pH 6.4, 100 l of 0.2 M citric acid, 50 l of 3.34 mg/ml bathophenanthroline sulfonate, and 50 l of 3.3 mM FeCl 3 . The samples were mixed well, and the absorbance at 535 nm was measured after color development (10 -60 min). The difference in measurements at 37 and 4°C was taken as the specific activity of the cell-associated ferrireductase.
Iron Binding Studies-K562 cells were washed three times with PBS and resuspended in 25 mM HEPES, pH 7.4, 150 mM NaCl, 1 mg/ml bovine serum albumin. Binding assays (250 l) were prepared with 0.5-1 ϫ 10 6 cells and 0.1-5 M 55 FeNTA in the presence or absence of unlabeled ligand (5 mM). The reaction mixtures were incubated for 1 h at 4°C. The cells were then pelleted by centrifugation, washed once with PBS, and resuspended in 25 mM HEPES, pH 7.4, 150 mM NaCl; and cell-associated radioactivity was measured. The amount of nonspecific binding was determined in the presence of unlabeled FeNTA and subtracted from each point. The K d for binding and the number of binding sites were determined by Scatchard analysis (20).

RESULTS
To assess the effects of phorbol esters on non-Tf iron uptake, K562 cells were treated with the concentrations of PDBu indicated in Fig. 1 for 16 h. The ability of the cells to transport non-Tf iron was then measured as described under "Materials and Methods." As shown in Fig. 1, the rate of 55 Fe uptake was significantly increased upon exposure to 25 nM PDBu, with the maximum increase in non-Tf transport activity induced by 50 nM PDBu. Treatment with PDBu promoted a maximal stimulation of transport 3-5-fold over that observed for control (untreated) K562 cells in a dose-dependent manner.
Effects of protein kinase C activation by phorbol esters can be manifested within several minutes after administration; however, treatment of K562 cells with phorbol esters is known to induce cellular differentiation, which requires several hours to days before observable changes occur (18). To establish the earliest time point for induction of the increase in non-Tf iron uptake, K562 cells were treated with 50 nM PDBu for up to 16 h and then assayed for 55 Fe transport activity. The results presented in Fig. 2 demonstrate that increased transport rates are not observed until after exposure to PDBu for Ͼ8 h. The increase in 55 Fe uptake could result from changes in the V max of uptake, the apparent K m , or both of these kinetic parameters. We therefore determined these values for treated and control K562 cells as described previously (8). Doublereciprocal plots of initial transport rate as a function of [FeNTA] were employed to establish the V max and apparent K m . As shown by data presented in Fig. 3, both kinetic parameters were altered upon overnight treatment with PDBu. Table  I presents a summary of the results from these experiments. While control cells display a V max comparable to previously reported activity (8), PDBu-treated cells have a V max that is ϳ20-fold greater ( Table I). The K m of transport was also increased 4 -6-fold compared with control K562 cells.
In contrast to the observed changes induced by phorbol esters, treatment of K562 cells with bryostatin-1 fails to induce any significant alteration in the properties of non-Tf iron uptake (Table I). Like phorbol esters, bryostatin-1 is an activator of protein kinase C; however, the latter agent does not induce megakaryocytic differentiation of K562 cells (21). Thus, our observations suggest that up-regulation of non-Tf iron transport activity is specifically associated with phorbol ester-mediated differentiation of K562 cells and the concomitant expression of phenotypic markers of megakaryocytic lineage. Since actinomycin D, an inhibitor of mRNA synthesis, blocks phorbol ester-stimulated transport activity (Fig. 4), we conclude that transcriptional activation associated with megakaryocytic differentiation must be involved in the observed increase in the V max and K m of non-Tf iron transport. The simplest interpretation of these results is that synthesis of a new class of transporters is induced under these conditions, although it is not possible to rule out other post-translational effects that may regulate the activity of the existing iron uptake system.
It is interesting to note that the higher K m of iron transport by phorbol ester-treated K562 cells correlates well with apparent K m values reported for several other cell types. This may suggest that upon differentiation, K562 cells express an uptake mechanism akin to that observed for hepatocytes, fibroblasts, and Chinese hamster ovary and HeLa cells (1)(2)(3). In the latter studies, supraphysiological concentrations of calcium have been shown to promote significant increases in non-Tf iron uptake. Moreover, depletion of intracellular Ca 2ϩ also correlates with decreased transport activity. To establish whether the changes in transport induced by PDBu may reflect an altered response of non-Tf iron uptake to intra-or extracellular Ca 2ϩ levels, treated and control K562 cells were washed with 1 mM EGTA to deplete intracellular Ca 2ϩ or exposed to extracellular Ca 2ϩ (1 mM). As shown by the results presented in Table  II, neither treatment had a significant effect on the transport activity measured for PDBu-treated or control cells. These observations are consistent with previous data from our laboratory indicating that the non-Tf iron uptake system in K562 cells is calcium-independent (8); most important, the results demonstrate that stimulation of transport activity by PDBu does not reflect an altered Ca 2ϩ response of the non-Tf iron uptake mechanism. The specificity of the non-Tf iron transporter in K562 cells with respect to other transition metals is also distinct from other uptake systems that have been characterized. Whereas the transporter in K562 cells may be specific only for iron and cadmium (8), the non-Tf iron transporter in HeLa cells may be able to transport cadmium, manganese, copper, and zinc as well (1,2). Since the evidence presented above indicates fundamental changes in transport characteristics, we further examined the effects of 200 M Cd 2ϩ , Co 2ϩ , Cu 2ϩ , Ni 2ϩ , and Mn 2ϩ in the uptake assay. As shown by the results presented in Table II, only Cd 2ϩ and Cu 2ϩ significantly inhibited non-Tf iron uptake by control K562 cells, consistent with previous observations (8). None of the other cations were potent inhibitors of transport, unlike observations made for non-Tf iron uptake by HeLa cells, Chinese hamster ovary cells, and fibroblasts (1). The stimulated transport activity of PDButreated K562 cells was also resistant to inhibitory effects of these metals; moreover, the efficacy of Cd 2ϩ and Cu 2ϩ to inhibit uptake was reduced (Table II). This result may reflect the increase in apparent K m of transport upon phorbol ester treatment. Thus, the specificity of non-Tf iron transport by K562 cells does not appear to be altered upon differentiation.
A cell-surface ferrireductase activity has been associated with non-Tf iron uptake by K562 cells (8). Reduction of Fe 3ϩ to Fe 2ϩ is thought to be the first step in the uptake mechanism, and inhibition of the ferrireductase activity corresponds to a block in non-Tf iron uptake (22). Thus, the observed increase in non-Tf iron uptake could be associated with a parallel stimulation of the ferrireductase activity. To test this hypothesis, cell-surface ferrireductase activity was measured in PDButreated and control K562 cells, with the results shown in Fig. 5.

FIG. 4. Actinomycin D blocks phorbol ester-induced non-Tf iron transport by K562 cells.
K562 cells were incubated with or without 50 nM PDBu and 1 g/ml actinomycin D for 16 h. At the end of the incubation period, the cells were collected by centrifugation, and non-Tf iron uptake was measured exactly as described for Fig. 1. PMA, phorbol 12-myristate 13-acetate; ***, significantly different from control (p Ͻ 0.001).

TABLE I Kinetic parameters of non-Tf iron transport by K562 cells
The apparent K m and V max of non-Tf iron uptake by K562 cells were determined as previously described (8 was stimulated 5-10-fold compared with control K562 cells, with measured activities of 337 Ϯ 53 and 43 Ϯ 3 pmol/min/10 6 cells, respectively (n ϭ 4). The enhanced ferrireductase activity correlates well with the observed increase in non-Tf iron uptake after overnight PDBu treatment and is compatible with the idea that this function is closely associated with the uptake system. Since reduction of Fe 3ϩ to Fe 2ϩ is presumably a prerequisite for non-Tf iron transport (22,23), it is possible that the observed stimulation of uptake is a functional consequence of enhanced ferrireductase activity, although direct evidence that the latter step is rate-limiting for transport is lacking.
To identify proteins potentially involved in the phorbol esterinduced transport activity, iron binding studies were performed. As indicated by the results of Fig. 6, K562 cells do not display a significant number of 55 Fe-binding sites under basal conditions. However, upon megakaryocytic differentiation, saturable binding of 55 Fe to the surface of phorbol ester-treated K562 cells can be measured. Scatchard analysis of binding data reveals a K d for 55 Fe binding of 0.37 Ϯ 0.06 nM, with a single class of ϳ5.4 ϫ 10 7 binding sites/cell. Although the functional relationship with the observed stimulation of non-Tf iron up-take remains unknown, it is possible that the iron-binding proteins are components of a new class of transporters expressed by K562 cells upon megakaryocytic differentiation. DISCUSSION Tf-mediated iron delivery is known to be modulated by ironinduced changes in the half-life of receptor mRNA, as well as by regulation of the rate of transcription (24). In contrast, non-Tf iron uptake does not appear to be influenced by intracellular iron levels (8), indicating fundamental differences in the regulation of these two pathways for iron assimilation. Regulation of Tf receptor activity has also been correlated with leukemia cell differentiation and the cessation of cellular proliferation (9 -15). When human erythroleukemia K562 cells are exposed to phorbol esters, a rapid down-regulation of surface Tf receptors to intracellular compartments is observed (14,15), and prolonged exposure will diminish receptor synthesis (17). In contrast, our results show that the non-Tf iron transport system is up-regulated by phorbol esters. Combined, these observations suggest that the cellular expression of these two different iron uptake mechanisms may be coordinately regulated. It is interesting to note that despite the down-regulation of the K562 cell Tf receptors under these conditions, the Tf-mediated iron delivery is only slightly suppressed (14,25), supporting the idea that the coordinate regulation between Tf-dependent and -independent pathways does not reflect changes in intracellular iron content.
Phorbol esters promote other changes in the phenotype of K562 cells as they cease to divide and begin to differentiate, expressing megakaryocytic markers (18). The fact that actinomycin D blocks the PDBu-induced stimulation of Tf-independent transport indicates that mRNA and most likely protein synthesis are both required for this effect. Thus, altered transport activity appears to be the consequence of changes in cellular function due to the differentiation program. In fact, bryostatin-1, which activates protein kinase C but does not induce K562 cell differentiation (21), fails to stimulate non-Tf iron transport activity. Two potential explanations can account for stimulation of non-Tf iron transport. 1) Expression of regula- FIG. 5. Ferricyanide reductase activity is stimulated in PDButreated K562 cells. K562 cells were treated with 50 nM PDBu overnight (16 h) and then assayed for cell-surface ferrireductase activity as described under "Materials and Methods." Briefly, the capacity of cells to reduce membrane-impermeant ferricyanide was measured by the production of ferrocyanide upon incubation at 37°C. Background activity was measured at 4°C in parallel reactions, and this value was subtracted to yield specific cell-associated ferricyanide reductase activity. Data are the average activity Ϯ S.E. from four independent experiments expressed as pmol/min/10 6 cells.  tory factors may modify the activity of existing plasma membrane iron transporters, and/or 2) synthesis of a new class of non-Tf transporters is induced to increase iron uptake. We favor the latter hypothesis due to the appearance of cell-surface iron-binding sites upon phorbol ester treatment of K562 cells; this activity may reflect properties of the new class of transport molecules. Conrad et al. (26) have proposed that ␤ 3 -integrin is involved in non-Tf iron transport by K562 cells, although stringent functional criteria to support a role for this extracellular matrix adhesion factor in iron uptake are lacking. Since, upon megakaryocytic differentiation of K562 cells, expression of glycoprotein IIIa or ␤ 3 -integrin is up-regulated (18), it is possible that the binding activity we observe is due to association of 55 Fe with ␤ 3 -integrin. The latter does indeed bind cations; in fact, bound cation is displaced upon the binding of ␤ 3 -integrin by its ligand, typically fibrinogen or another adhesion molecule that contains the RGD amino acid motif (27). We have examined the ability of RGD-containing peptides to block both non-Tf iron transport and cell-surface iron binding in basal and stimulated K562 cells. Since the RGD peptides do not compete for uptake or binding, we conclude that neither of these activities involves the ligand-displaceable cation-binding site of ␤ 3integrin. 2 Clearly, further investigation is necessary to define the function of the cell-surface iron-binding proteins that are induced by phorbol esters. As indicated above, we speculate that these binding sites are functionally related to the enhanced non-Tf iron system observed in K562 cells upon megakaryocytic differentiation. Future efforts will not only help to identify the molecules involved in iron assimilation, but may also explain why the Tf-independent transport system is up-regulated during megakaryocytopoiesis.