Interaction of calcium with the human divalent metal-ion transporter-1

Abstract

Iron deficiency is the most prevalent micronutrient deficiency worldwide. Whereas dietary calcium is known to reduce the bioavailability of iron, the molecular basis of this interaction is not understood. We tested the hypothesis that divalent metal-ion transporter-1 (DMT1—the principal or only mechanism by which nonheme iron is taken up at the intestinal brush border—is shared also by calcium. We expressed human DMT1 in RNA-injected Xenopus oocytes and examined its activity using radiotracer assays and the voltage clamp. DMT1 did not mediate ⁴⁵Ca²⁺ uptake. Instead, we found that Ca²⁺ blocked the Fe²⁺-evoked currents and inhibited ⁵⁵Fe²⁺ uptake in a noncompetitive manner (K_i ≈ 20 mM). The mechanism of inhibition was independent of voltage and did not involve intracellular Ca²⁺ signaling. The alkaline earth metal ions Ba²⁺, Sr²⁺ and Mg²⁺ also inhibited DMT1-mediated iron-transport activity. We conclude that Ca²⁺ is a low-affinity noncompetitive inhibitor—but not a transported substrate—of DMT1, explaining in part the effect of high dietary calcium on iron bioavailability.

Introduction

Despite recent progress in our understanding of iron absorption, iron deficiency remains the most prevalent micronutrient deficiency worldwide. Whereas dietary calcium is known to reduce the bioavailability of iron [1,2,7,11-13], the molecular basis of this interaction is not understood. Divalent metal-ion transporter-1 (DMT1) (Ref. [5,10]) is the principal or only mechanism by which nonheme iron is taken up at the intestinal brush border [9], reviewed [15]. This widely-expressed transporter is also critical for the recovery of iron from recycling endosomes during transferrin receptor-mediated cellular iron uptake in erythroid precursor cells [4,5,9].

Since DMT1 is a proton-coupled ferrous-ion (Fe²⁺) transporter that is shared by certain other divalent transition metal ions [6,10,15-17,23], we considered whether calcium could also interact with DMT1. The literature contains at least two reports that suggest an association between Ca²⁺ transport and DMT1. For example, a DMT homolog from the scallop is reported to operate as a calcium transporter [24] and a glycine-to-arginine substitution (at residue 185 of mouse DMT1)—which disrupts DMT1-mediated iron transport and results in microcytic anemia in the mk mouse [5,22]—is thought to expose a calcium-entry pathway [25]. Therefore, we tested in vitro the hypothesis that Ca²⁺ is a transported substrate of human DMT1. We also assessed Ca²⁺ inhibition of DMT1-mediated Fe²⁺ and H⁺ transport activities, and explored whether the other group 2 alkaline earth metal ions (Mg²⁺, Sr²⁺, Ba²⁺) mimic the effects of Ca²⁺.

Materials and Methods

Expression of human DMT1 in Xenopus oocytes

We performed laparotomy and ovariectomy on adult female Xenopus laevis frogs (Nasco, Fort Atkinson, WI) under 2-aminoethylbenzoate methanesulfonate anesthesia (0.1% in 1:1 water/ice, by immersion) following a protocol approved by the University of Cincinnati Institutional Animal Care and Use Committee. Ovarian tissue was isolated and treated with collagenase A (Roche Diagnostics Corp., Indianapolis, IN), and oocytes were isolated and stored at 17 °C in modified Barths’ medium as described [14]. We expressed in Xenopus oocytes the 1A/IRE(+) isoform of DMT1, the product of the human SLC11A2 gene. We linearized the pOX(+) plasmid containing 1A/IRE(+) DMT1 cDNA under the SP6 promoter (as described [16]) using SnaBI (New England Biolabs Inc, Ipswich, MA) and synthesized RNA in vitro using the mMESSAGE mMACHINE / SP6 RNA polymerase kit (Applied Biosystems / Ambion, Austin, TX) according to the manufacturers’ protocols. Defolliculate stage V-VI oocytes were injected with approximately 50 ng of DMT1 RNA and incubated 6 to 7 days before being used in functional assays.

Reagents and media

Reagents were obtained from Sigma-Aldrich Corp. (St. Louis, MO) or Research Products International Corp. (Prospect, IL) unless otherwise indicated. Oocytes were superfused or incubated at room temperature (22-24 °C) in transport media containing 100 mM NaCl, 1 mM KCl, 1 mM l-ascorbic acid, buffered using 0-5 mM 2-(N-morpholino)ethanesulfonic acid (MES) and 0-5 mM N′,N′-diethylpiperazine (DEPP) (GFS Chemicals, Columbus, OH) to obtain pH 5.5-7.5 as indicated. Where indicated, we added up to 50 mM Ca²⁺, or up to 20 mM Mg²⁺, Sr²⁺ or Ba²⁺ by equimolar replacement of Na⁺.

Radiotracer assays

We obtained radiochemicals from Perkin-Elmer Life Science Products (Boston, MA). We used ⁵⁵Fe at final specific activity 0.31-1.99 GBq.mg⁻¹ (added as FeCl₃) and ⁴⁵Ca at 75-263 MBq.mg⁻¹ (added as CaCl₂). Radiotracer metal-ion uptake was measured over 10 min, within the linear portion of the timecourse of ⁵⁵Fe²⁺ uptake [16]. We terminated radiotracer uptake by rapidly washing the oocytes three times in ice-cold pH 5.5 transport medium containing 1 mM l-ascorbic acid. Oocytes were then solubilized using 5% (w/v) sodium dodecylsulfate and radiotracer content assayed by liquid-scintillation counting using Scintisafe-30% liquid-scintillation cocktail (Fisher Scientific, Pittsburgh, PA). Uptake data for 0.38-38 μM ⁵⁵Fe²⁺ in the absence of Ca²⁺ or in the presence of 20 mM Ca²⁺ were fit by a modified Hill function (equation 1) for which V_max is the maximal velocity of ⁵⁵Fe²⁺ uptake (V), S is the concentration of substrate S (Fe²⁺), K_0.5^S is the concentration of S at which V is half-maximal, and n_H is the Hill coefficient.

$$V=\frac{V_{\max }\cdot S^{n_{\mathrm{H}}}}{(K_{0.5}^{\mathrm{S}})+S^{n_{\mathrm{H}}}}$$ (Eq. 1)

Saturation kinetics parameters derived using equation 1 were used to estimate K_i^Ca, the Ca²⁺ concentration resulting in half-maximal inhibition of ⁵⁵Fe²⁺ transport, according to equation 2 (see Ref. [19]) for which V_max and V_maxi are respectively the maximal velocities in the absence and presence of inhibitor, [I] is the inhibitor concentration (in this case, 20 mM Ca²⁺).

$$K_{\mathrm{i}}^{\mathrm{Ca}}=\frac{[\mathrm{I}]}{(V_{\max }/V_{{\max }_{\mathrm{i}}})-1}$$ (Eq. 2)

Voltage-clamp experiments

We used the two-microelectrode voltage clamp (Dagan CA-1B amplifier) to measure currents in control oocytes and oocytes expressing DMT1. Microelectrodes (resistance 3-6 MΩ) were filled with 3 M KCl. Voltage-clamp experiments comprised two protocols: (i) Continuous current recordings were made at a membrane potential (V_m) of −70 mV, digitized at 10 Hz and low-pass filtered at 1 Hz. (ii) Oocytes were clamped at V_m = −50 mV, and step-changes in V_m were applied from +50 mV to −150 mV (in 20-mV increments) each for a duration of 200 ms, before and after the addition of metal-ion substrate. Current was digitized at 5 kHz and low-pass filtered at 500 Hz. Steady-state data were obtained by averaging the points over the final 16.7 ms at each V_m step. Data from protocol (i) exploring the inhibition of the H⁺ leak current by Ca²⁺ (see Fig. 3B) were fit by equation 3 for which I is the Ca²⁺-induced shift in current, I_max the derived current maximum, S the concentration of Ca²⁺, and K_i^Ca the Ca²⁺ concentration resulting in half-maximal inhibition of the H⁺ leak current.

Fig. 3.

Calcium inhibition of the DMT1-mediated H⁺ leak and Fe²⁺-evoked currents. (A) Typical continuous current recordings at membrane potential (V_m) of 70mV in a control oocyte (i) and in an oocyte expressing DMT1 (ii). Oocytes were superfused with Ca²⁺-free medium at pH 7.5 for the periods shown by the empty bars, or with Ca²⁺-free medium at pH 5.5 (hatched bars). We presented 20 mM Ca²⁺ as indicated by the gray bars and 20 μM or 2 μM Fe²⁺ (black bars). (B) Ca²⁺-induced current (I^Ca) in the absence of Fe²⁺ metal ion as a function of Ca²⁺ concentration. Data were fit by equation 3 to describe the Ca²⁺ inhibition of the DMT1-mediated H⁺ leak current: I_max = 16.4 ± 1.0 nA, K_i^Ca = 1.1 ± 0.3 mM (adjusted r² = 0.88, P < 0.001). (C) Current-voltage relationships of the currents induced by 2 μM Fe²⁺ alone (Fe, filled symbols) or 2 μM Fe²⁺ plus 20 mM Ca²⁺ (Fe + Ca, empty symbols) (n = 4, paired). (D) Effect of intracellular Ca²⁺ chelation on the Ca²⁺ inhibition of the DMT1-mediated Fe²⁺-evoked current. We voltage-clamped at −70 mV four oocytes expressing DMT1, and measured currents evoked by 2 μM Fe²⁺ alone or 2 μM Fe²⁺ plus 20 mM Ca²⁺ before (−2 min) and after (10, 20, 30 min) microinjection with EGTA. Each symbol shape represents an individual oocyte. The mean Fe²⁺-evoked current at −2 min was −88 (SD 64) nA and did not change with time (one-way repeated measures ANOVA, P = 0.83). We obtained percent inhibition by Ca²⁺ and found no effects over time (one-way repeated measures ANOVA, P = 0.18).

$$I=\frac{I_{\max }\cdot S}{(K_{\mathrm{i}}^{\mathrm{Ca}}+S)}$$ (Eq. 3)

Intracellular calcium chelation

To chelate intracellular calcium in oocytes expressing DMT1 (see Fig. 3D), we microinjected oocytes with 23 nl of a 50-mM solution of ethylene glycol-bis(2-aminoethylether)-N,N,N′,N′-tetraacetic acid (EGTA), adjusted to pH 7.5 using KOH, to obtain a nominal final intracellular EGTA concentration of 1 mM.

Statistical and regression analyses

We performed statistical analyses using SigmaPlot version 11 (Systat Software) with critical significance level α = 0.05. We have presented our data as mean and standard deviation (SD) for n independent observations and used parametric tests in their analysis since our data either (1) passed the Shapiro-Wilk test for normality, or (2) n ≥ 30 in each group, so normality was not tested. Data were fit by equations 1 and 3 using the least-squares method of nonlinear regression analysis, the results of which are expressed as the estimates of fit parameters ± standard error (SE); adjusted r² is the adjusted regression coefficient and P describes the significance of the fit. CI is the 95% confidence interval. Between-group comparisons were made using one-way or two-way analysis of variance (ANOVA) followed by pairwise multiple comparisons by the Holm-Šidák test. Data in Fig. 3D were analyzed using repeated-measures one-way ANOVA.

Results

DMT1 does not mediate calcium transport

We expressed human DMT1 in Xenopus oocytes and assayed metal-ion transport activity. Under conditions that stimulated over 60-fold the uptake of ⁵⁵Fe²⁺ compared with control (i.e. 2 μM metal ion at pH 5.5), the expression of DMT1 did not increase ⁴⁵Ca²⁺ transport compared with control (Fig. 1A). We considered that DMT1 could perhaps exhibit much lower affinity for Ca²⁺ than for Fe²⁺, so we tested whether significant ⁴⁵Ca²⁺ transport could be observed at a higher concentration. Expression of DMT1 resulted in a minute increase in the uptake of ⁴⁵Ca²⁺ at 100 μM (Fig. 1B); the DMT1-induced ⁴⁵Ca²⁺ uptake was only 1.2% ± 0.3% (SE) the DMT1-induced ⁵⁵Fe²⁺ uptake observed at 1/50th the concentration. Moreover, we found this increase to be nonspecific (possibly due to increased ‘leakiness’ of the oocytes overexpressing an exogenous protein) since the DMT1-induced ⁴⁵Ca²⁺ uptake was not inhibited by a saturating concentration (100 μM) of Fe²⁺. We therefore conclude that human DMT1 is not capable of transporting calcium.

Fig. 1.

DMT1 does not mediate ⁴⁵Ca²⁺ transport in Xenopus oocytes. (A) Radiotracer metal-ion (*Me²⁺) uptake (pH 5.5, 1 mM Mg²⁺) in control oocytes and oocytes expressing human DMT1 (n = 9-13). We compared uptake of 2 μM ⁵⁵Fe²⁺ with that of 2 μM ⁴⁵Ca²⁺. Two-way ANOVA revealed an interaction (P < 0.001); ^a within ⁴⁵Ca²⁺ uptake, DMT1 did not differ from control (unadjusted P = 0.933). (B) Uptake of 100 μM ⁴⁵Ca²⁺ (pH 5.5) in the absence (None) or presence (+Fe²⁺) of 100 μM Fe²⁺ in control oocytes and oocytes expressing human DMT1 (n = 22-32). (Note the y-axis scale is one tenth that in A.) Two-way ANOVA revealed a main effect of DMT1 (P < 0.001) but not of [Fe²⁺] (P = 0.99), and no interaction (P = 0.094).

Calcium is a noncompetitive inhibitor of DMT1-mediated iron transport

We further explored the interaction of calcium with DMT1 by testing the ability of calcium to inhibit DMT1-mediated transport activities. The addition of 20 mM Ca²⁺ inhibited by 52% ± 5% (SE) the uptake of 2 μM ⁵⁵Fe²⁺ in oocytes expressing DMT1 (Fig. 2A). We examined the saturation kinetics of ⁵⁵Fe²⁺ uptake in the absence and presence of calcium and found that 20 mM Ca²⁺ decreased the V_max for DMT1-specific ⁵⁵Fe²⁺ uptake by 48% ± 6% (SE, P < 0.001) without effect on K_0.5 for Fe²⁺ (not significant, P = 0.12), indicating that calcium is a noncompetitive inhibitor of DMT1-mediated ⁵⁵Fe²⁺ transport (Fig. 2B). In contrast, a competitive inhibitor is expected to increase the apparent K_0.5 without effect on V_max [19]. Using equation 2, we estimated K_i^Ca to be 21.7 (CI 17.5, 25.9) mM in this preparation. In a second preparation (not shown), we obtained K_i^Ca = 17.8 (CI 9.8, 25.8) mM. Therefore Ca²⁺ is a noncompetitive inhibitor of DMT1-mediated Fe²⁺ transport with K_i^Ca ≈ 20 mM.

Fig. 2.

Ca²⁺ is a noncompetitive inhibitor of DMT1-mediated Fe²⁺ transport. (A) Uptake of 2 μM ⁵⁵Fe²⁺ (pH 5.5, 1 mM Mg²⁺) in the absence or presence of 20 mM Ca²⁺ in control oocytes and oocytes expressing DMT1 (n = 10-12). Two-way ANOVA revealed an interaction (P < 0.001). (B) Saturation kinetics of ⁵⁵Fe²⁺ uptake (pH 5.5, 1 mM Mg²⁺) in the absence (circles) or presence (triangles) of 20 mM Ca²⁺ (n = 10-14 per group at each concentration). We obtained the specific DMT1-mediated ⁵⁵Fe²⁺ uptake by subtracting from the ⁵⁵Fe²⁺ uptake in oocytes expressing DMT1 (filled symbols) the endogenous ⁵⁵Fe²⁺ uptake derived from linear fits (dashed lines) of the ⁵⁵Fe²⁺ uptake in control oocytes (empty symbols). Specific DMT1-mediated ⁵⁵Fe²⁺ uptake data were fit (solid lines) by equation 1, yielding for [Ca²⁺] = 0 the parameters V_max = 5.0 ± 0.3 pmol.min⁻¹, n_H = 1.2 ± 0.1, and K_0.5 = 8.4 ± 1.0 μM (adjusted r² = 0.96, P <0.001) and for [Ca²⁺] = 20 mM the parameters V_max = 2.6 ± 0.1 pmol.min⁻¹, n_H = 1.2 ± 0.1, and K_0.5 = 10.2 ± 1.2 μM (adjusted r² = 0.97, P <0.001).

Calcium inhibition of DMT1-mediated H⁺ leak and Fe²⁺-evoked currents

We examined the effects of calcium on the currents associated with expression of human DMT1 in Xenopus oocytes (Fig. 3). In control oocytes voltage clamped at membrane potential (V_m) of −70 mV, we observed a small inward current upon switching the superfusion medium from pH 7.5 to 5.5; however, neither Ca²⁺ nor Fe²⁺ had any effect on current at pH 5.5 in control oocytes (Fig. 3A). Switching from pH 7.5 to pH 5.5 induced a more pronounced inward current in oocytes expressing DMT1; this current—in the absence of metal-ion—is attributed to a H⁺ leak pathway associated with DMT1 as previously described [10,16,17]. The DMT1-associated H⁺ leak current was inhibited in the presence of 20 mM Ca²⁺. The magnitude of the inhibition was dependent upon the Ca²⁺ concentration (Fig. 3B); using equation 3, we obtained K_i^Ca = 1.1 ± 0.3 mM. That the K_i^Ca for inhibition of the H⁺ leak was lower than the K_i^Ca for inhibition of ⁵⁵Fe²⁺ uptake most likely reflects a higher affinity for Ca²⁺ at V_m = −70 mV compared with that at depolarized potentials expected in nonclamped oocytes in the presence of ⁵⁵Fe²⁺.

We found that 20 μM Fe²⁺ evoked large, reversible inward currents in oocytes expressing DMT1 (Fig. 3A), as previously observed [10,16,17]. The currents evoked by 2 μM Fe²⁺ were strongly inhibited by the addition of 20 mM Ca²⁺. Ca²⁺ inhibited the Fe²⁺-evoked currents at any given V_m between −150 mV and +50 mV (Fig. 3C), strengthening the conclusion that Ca²⁺ specifically interacts with DMT1. To test whether the effect of Ca²⁺ on the Fe²⁺-evoked currents was mediated via intracellular signaling, we measured in four oocytes expressing DMT1 the Fe²⁺-evoked current before and after microinjection with EGTA to chelate the intracellular Ca²⁺ (Fig. 3D). The inhibitory effect of Ca²⁺ persisted at least up to 30 min following EGTA injection, indicating that the Ca²⁺-inhibitory effect is not exerted via intracellular signaling.

Inhibition of DMT1-mediated iron transport by other alkaline earth metals

We explored whether other alkaline earth metal ions could also inhibit DMT1-mediated transport activity. Again, 20 mM Ca²⁺ inhibited the uptake of 2 μM ⁵⁵Fe²⁺ by 53% ± 5% (SE) and Sr²⁺ inhibited ⁵⁵Fe²⁺ transport to a similar extent (58% ± 5%, SE) (Fig. 4). Ba²⁺ resulted in stronger inhibition (73% ± 5%, SE) and Mg²⁺ more weakly inhibited ⁵⁵Fe²⁺ uptake (28% ± 6%, SE).

Fig. 4.

Inhibition of DMT1-mediated Fe²⁺ transport by the alkaline-earth metals. We measured uptake of 2 μM ⁵⁵Fe²⁺ at pH 5.5 in standard transport medium (None) or in the presence of 20 mM Ca²⁺, Ba²⁺, Mg²⁺, or Sr²⁺ in control oocytes and oocytes expressing DMT1 (n = 12-16). Two-way ANOVA revealed an interaction (P < 0.001); ^a within DMT1, all treatments with alkaline-earth metals differed from no treatment (P < 0.001).

Discussion

We have expressed human DMT1 in RNA-injected Xenopus oocytes and found that DMT1 did not mediate ⁴⁵Ca²⁺ uptake, in contrast to the scallop DMT homolog which transports calcium [24]. We found, however, that Ca²⁺ (i) blocked the DMT1-mediated H⁺ leak currents, with K_i^Ca ≈ 1 mM, (ii) blocked the Fe²⁺-evoked currents, and (iii) inhibited ⁵⁵Fe²⁺ uptake in a noncompetitive manner, with K_i^Ca ≈ 20 mM. The other group 2 alkaline earth metals—Ba²⁺, Mg²⁺ and Sr²⁺—also inhibited iron transport. Since DMT1 interacts with Ca²⁺ at relatively low affinity, it is unlikely that Ca²⁺ would be a useful experimental tool with which to block DMT1 transport in vitro. Ba²⁺ afforded stronger inhibition at 20 mM; however, further investigation is needed to determine whether Ba²⁺ could serve as a sufficiently-specific blocker of DMT1 in vitro.

DMT1 is expressed in erythroid precursor cells where it transports iron from recycling endosomes to the cytoplasm following transferrin receptor-mediated cellular iron uptake. Intracellular Ca²⁺ is thought to regulate iron uptake in K562 cells, an erythroid-cell model, by accelerating transferrin/transferrin-receptor cycling [3]. Noting this observation, we explored whether intracellular Ca²⁺ may also regulate the activity of DMT1 expressed in the oocyte. We found, however, that chelation of intracellular Ca²⁺ (by EGTA injection) affected neither the magnitude of the Fe²⁺-evoked currents nor the magnitude of the Ca²⁺-inhibitory effect. We conclude that the Ca²⁺-inhibitory effect is not mediated via intracellular Ca²⁺ signaling and propose instead that Ca²⁺ acts at an extracellular site on DMT1.

The transferrin cycle also accounts for iron uptake in most other cell types. Since relatively large concentrations of Ca²⁺ are required to inhibit DMT1 transport activities (K_i^ca = 1-20 mM) in the oocyte it is unlikely that Ca²⁺ should interfere with DMT1-mediated transport of Fe²⁺ in most cell types. One notable exception is the enterocyte, in which DMT1 is expressed at the apical (luminal) membrane, since intestinal luminal Ca²⁺ concentrations may periodically reach the millimolar range as a result of dietary calcium. For example, the free Ca²⁺ concentration of milk (of various sources) is estimated at 3.2-4.2 mM by the use of an ion-sensitive electrode [21]; total soluble Ca is about 12.5 mM and the total calcium concentration (including that bound by casein) in milk is around 30 mM [8]. Whereas the luminal contents will be diluted as a result of intestinal secretions, the concentration of calcium from dietary sources such as milk may be sufficient to inhibit iron absorption significantly [7,13] and calcium supplementation has been shown to reduce iron absorption [12,13,18]. Our data indicate that an interaction of calcium with DMT1 could, at least in part, provide the molecular basis to the observation that dietary calcium inhibits iron absorption, and calcium inhibition of DMT1 may be taken into consideration when developing strategies for improved nutrition, milk formulation and iron fortification.