Histidine-stimulated divalent metal uptake in human and in the erythroleukaemic cell line HEL.92.1.7

Abstract

The uptake of ⁶⁵Zn by human erythrocytes was investigated in the presence of high (40 mm) and low (5 mm) concentrations of histidine and 0–500 μm cobalt, nickel, manganese and zinc. Varying concentrations of metal mono- and bis-histidine complexes will be formed and the inhibition of ⁶⁵Zn uptake could be correlated with the calculated complex concentrations to investigate competition between metals. For each metal, the calculated concentrations of bis-histidine complex giving 50% inhibition of ⁶⁵Zn uptake were similar at both 5 mm and 40 mm histidine. Manganese–bis-histidine appeared to have a much higher affinity for the binding site than the other metal–bis-histidine complexes, which had similar affinities to each other. Studies of the inhibition of histidine-stimulated ⁵⁴Mn uptake by the addition of manganese confirmed that manganese–bis-histidine does act as a substrate for the transporter in a similar fashion to the other metals studied. In addition, human erythroleukaemic cells (HEL cells) were used as a model for erythroid precursor cells. l-histidine, but not d-histidine, stimulated ⁶⁵Zn uptake in a saturable fashion. The other metals competed with zinc in a similar manner to that seen in erythrocytes, and the affinity for manganese–bis-histidine was much greater than for the bis-histidine complexes of the other three metals. Both the capacity for metal transport per cell, and the affinity of the transporter for the metal–bis-histidine complexes, were much greater in the HEL cells than in the erythrocyte. It is suggested that histidine-stimulated metal transport may play a role in the supply of metals to maturing erythroid cells.

The normal zinc concentration in human blood plasma is approximately 15 μm. Of this total, about one-third is incorporated within metalloproteins such as α₂-macroglobulin and is not exchangeable with other plasma components (Giroux, 1975). The remaining zinc is labile and forms possible substrates for cellular uptake mechanisms. Approximately 3% of the total plasma zinc is in the form of coordination complexes formed between zinc, histidine and cysteine (Prasad & Oberleas, 1970; Harris & Keen, 1989). These are in a dynamic equilibrium with free ionic zinc and zinc bound to serum albumin. We have previously investigated the possible role of this amino acid-bound fraction in cellular zinc uptake into human erythrocytes. We reported that l-histidine increased ⁶⁵Zn uptake into both rat and human erythrocytes in a dose-dependent fashion, and that the rate of uptake correlated with the calculated concentration of the zinc–bis-histidine complex but not that of the zinc–mono-histidine complex or of free ionic zinc. Stimulation was only seen with the l-enantiomer; d-histidine simply acted as a chelator and reduced uptake. In these experiments, bovine serum albumin (BSA) was present as a metal ion buffer to maintain a low free ionic metal activity even at lower histidine concentrations (Horn et al. 1995). We also reported that the uptake of ¹⁰⁹Cd was stimulated by histidine concentrations up to 40 mm in a similar fashion, and that the uptake again correlated with the calculated bis-histidine complex concentration. In addition, in the presence of excess l-histidine, unlabelled zinc and cadmium competitively inhibited the histidine-stimulated ⁶⁵Zn uptake and the relationship between metal–bis-histidine concentration and uptake could be fitted by a saturable one-site binding (Michaelis-Menten) model. The calculated cadmium– and zinc–bis-histidine concentrations giving 50% inhibition (apparent K_i) were similar. The stimulation by histidine of metal uptake was not inhibited by pretreatment of the erythrocytes with N-ethylmaleimide, suggesting that the recognition site did not require functional sulphydryl groups (Horn & Thomas, 1996).

This paper reports further experimental studies on the specificity of the uptake mechanism and investigations of the relationship between the concentration of various metal–histidine complexes and inhibition of ⁶⁵Zn uptake. In these experiments we used a lower histidine concentration (5 mm) so that both mono- and bis-histidine complexes were formed. The metal ion buffer N-[2-acetamido]-2-iminodiacetic acid (ADA) was used in preference to BSA to control the ionic metal concentrations because it represented a single defined ligand for which reliable stability constants were available.

The human erythrocyte is a non-nucleated end-stage cell and its membrane transport properties reflect the survival of carriers which were important in earlier stages of development (Christensen & Killberg, 1987). The importance of metal uptake in the mature erythrocyte, if any, is unclear. Uptake of metals such as zinc, cobalt and copper would have been essential at the earlier, nucleated stages when enzyme synthesis was occurring. We wished to extend the studies to a model cell which would more closely represent the erythroid progenitor cells. We therefore used the human erythroleukaemic cell line HEL.92.1.7. This cell line was originally isolated from a patient with Hodgkin's lymphoma who then developed erythroleukaemia; the cells have been stable in culture ever since. These cells are described as erythroid-like cells which are capable of exhibiting both spontaneous and induced globin production, and may be induced to differentiate towards either erythroid or megakaryocytic phenotypes (Martin & Papayannopoulou, 1982). They are therefore a suitable model system for the study of metal metabolism in the earlier stages of differentiation when the cell (erythrocytic or leucocyte) will have a requirement for trace metals such as zinc and manganese for carbonic anhydrase and superoxide dismutase synthesis. They also represent a potential source of RNA and DNA for molecular biological studies of putative metal–amino acid uptake systems.

Our previous studies (Horn et al. 1995; Horn & Thomas, 1996, 1997) have demonstrated the existence in erythrocytes of a divalent metal uptake system which is dependent on the concentration of the metal–bis-histidine complex, is saturable and stereospecific. Histidine-stimulated metal uptake has also been described in several nucleated cell types, for example zinc into hepatocytes (Taylor & Simons, 1994), renal proximal tubule cells (Gachot et al. 1991) and brain in vivo (Buxani-Rice et al. 1994), and copper into hepatocytes (Bingham & McArdle, 1994) and hypothalamic slices (Barnea & Katz, 1990). However, all of these systems appear to differ from that in the erythrocyte by their lack of extreme stereospecificity. We wished to investigate whether nucleated cells with erythroid characteristics showed the metal uptake properties characteristic of erythrocytes which are not found in non-erythroid cells.

Methods

The methods used were similar to those described in detail by Horn et al. (1995). Briefly, blood was drawn, with local ethical committee approval, from volunteers. Coagulation was prevented by the addition of 100 mm EDTA to a final concentration of 10 mm. The red cell pellet was washed three times in 150 mm NaCl and then resuspended in suspension buffer containing (mm): NaCl 150, sucrose 125, Hepes 10 (pH 7.4) at 10% haematocrit. Incubations were conducted at 37°C and started by adding an aliquot of red cell suspension to tubes containing an equal volume of suspension buffer containing isotopic metals (⁶⁵Zn, specific activity 1.55 mCi mg⁻¹; ⁵⁴Mn, specific activity 240 mCi mg⁻¹; both from New England Nuclear), l-histidine, N-[2-acetamido]-2-iminodiacetic acid (ADA), BSA and unlabelled metals as appropriate at twice their final concentration. Aliquots (0.5 ml) were taken (in triplicate) at 0 and 12 min and transferred to 1.9-ml microfuge tubes (Eppendorf) containing 0.6 ml of 10 mm EDTA (to remove surface bound metal) in saline layered on top of 0.4 ml of silicone oil (Dow Corning 550). The microfuge tubes were spun at 7500 g for 1 min and the red cells pelleted through the silicone oil thus minimizing the amount of supernatant contamination in the pellet. The supernatant and most of the oil were then aspirated and the bottoms of the tubes, containing the red cell pellet, were cut off and allowed to fall into counting tubes. The HEL.92.1.7 cells were obtained from the European Collection of Cell Cultures, CAMR, Porton Down, UK, and grown as a suspension culture in media containing RPMI 1640 (Sigma R0883) supplemented with (mm): glucose 14, Hepes 10, sodium bicarbonate 1, l-glutamine 2, sodium pyruvate 1; and 10% fetal calf serum. Cells were incubated at 37°C in 5% CO₂ in air and passaged every 2–3 days. For the uptake experiments HEL cells were suspended in suspension buffer containing 1% BSA or 100 μm ADA at 37°C. The experiments were started by adding a 250 μl aliquot of cell suspension to an equal volume of suspension buffer containing ⁶⁵Zn, ⁵⁴Mn, l-histidine, d-histidine and 0–500 μm nickel, cobalt, zinc or manganese (as chlorides) as appropriate, at twice their final concentration, in a 3 ml polystyrene tube (Thermo-Luckham LP3) at 37°C. Experiments were terminated by the addition of 0.5 ml of ice-cold saline containing 10 mm EDTA in 150 mm NaCl to the tubes. This was intended to prevent further metal uptake and remove surface bound metal from the cells. The tubes were vortexed and then the contents were decanted into microfuge tubes (Eppendorf) containing 0.35 ml of silicone oil (a mixture of 40% 556: 60% 550, Dow Corning). The HEL cells were then pelleted through the oil by centrifugation at 7500 g. This minimized the amount of trapped fluid contained within the pellet. The remaining saline and incubation medium formed a layer above the oil. This layer and most of the oil was aspirated and the bottoms of the tubes containing the pellets were cut off and allowed to fall into counting tubes. Radioactivity in the pellets was measured using a gamma counter (Wallac). To allow for trapped fluid, surface bound isotope and background radiation samples were taken at 0 min in all experiments and the counts measured in these samples were subtracted from those measured at subsequent times. Measurements were made in triplicate in each experiment. Results are expressed as means ± s.e.m. where n is the number of experiments (in the case of experiments with HEL cells) or number of different blood donors (in the case of erythrocyte experiments). Significance of differences was assessed using Student's unpaired t test.

Variation of metal–histidine complex concentrations

In a metal–ligand system, all coordination complexes are labile and are in equilibrium. The log stability constant is a measure of the tendency for a particular complex to form, and may be used to calculate the concentrations of the various metal–ligand complexes in a system. This means that the concentration of a single complex cannot be varied independently. For example, the concentration of the metal–bis-histidine complex can be raised by increasing the ratio of histidine concentration to metal concentration, but this will simultaneously reduce the concentration of the metal–mono-histidine complex and the free ionic metal activity. In some of the experiments presented here, high histidine:metal ratios have been used to favour the formation of metal–bis-histidine complexes, while in others lower histidine concentrations in the presence of metal-ion buffers such as ADA have favoured the formation of metal–mono-histidine complexes while maintaining a low free metal activity to reduce uptake by mechanisms such as the Band 3 anion exchanger.

All calculations of complex concentrations have been performed by Basic computer programs kindly provided by Dr T. J. B. Simons, King's College London (personal communication). The various stability constants used are shown in Table 1.

Table 1.

Stability constants used in the calculation of metal–histidine and metal–ADA complex concentrations at pH 7.4

Metal–ligand complex	K1	β2
Zinc–ADA	7.10	9.22
Zinc–histidine	6.70	11.80
Nickel–ADA	7.86	11.61
Nickel–histidine	8.90	15.90
Cobalt–ADA	6.72	9.34
Cobalt–histidine	6.80	13.90
Manganese–ADA	4.72	6.93
Manganese–histidine	3.40	5.80
Cadmium–ADA	7.08	10.68
Cadmium–histidine	5.40	9.70
Copper–ADA	9.70	12.19
Copper–histidine	10.20	18.30

Results

In the initial experiments, erythrocytes were incubated with 15 μm ⁶⁵Zn and 5 mm l-histidine in the presence of 100 μm ADA (to maintain a low free zinc concentration) and varying (0–500 μm) concentrations of cobalt, nickel, manganese and zinc chlorides. ⁶⁵Zn uptake was determined after 12 min of incubation. Under these conditions (i.e. at 5 mm histidine), appreciable concentrations of both the mono- and bis- forms of the metal–histidine complexes are formed and it is thus possible to compare the degree of inhibition of ⁶⁵Zn uptake with the corresponding calculated concentrations of the free metal and metal–histidine complexes. In the presence of 5 mm histidine the change in free histidine concentration resulting from the addition of metals at the concentrations used (up to 500 μm) is small and thus there is very little change in the concentration of ⁶⁵Zn–histidine complexes available for uptake. Inhibition by the metals is thus being exerted at the membrane binding site rather than in competition for the histidine.

Figure 1A shows the inhibition of ⁶⁵Zn uptake plotted against the calculated concentration of metal–bis-histidine complexes in this experiment. It can be seen that in all cases the addition of increasing concentrations of metal has produced an inhibition of ⁶⁵Zn uptake. The inhibition curves for Zn–, Ni– and Co–bis-histidine complexes are almost superimposable with very similar apparent K_i values (Table 2). The apparent K_i for the manganese–bis-histidine complex is significantly lower (for zinc P = 0.0004, for nickel P = 0.0155, for cobalt P = 0.001) suggesting that the binding site of the transporter has a much higher affinity for this form of complex. If similar graphs are plotted (Fig. 1B) using the corresponding calculated metal–mono-histidine complex concentrations then curves are produced but they do not superimpose and the K_i values for the various complexes are disparate (Table 2). Similarly there is no clear relationship between the degree of inhibition of ⁶⁵Zn uptake and the calculated free metal concentrations.

Figure 1. The inhibition of ⁶⁵Zn uptake into human erythrocytes after 12-min incubation in the presence of 15 μm ⁶⁵Zn, 5 mm l-histidine and 100 μm ADA and of added Zn (□), Ni (Δ), Co (▪) and Mn (•) (0–500 μm).

The abscissae show the calculated metal–bis-histidine concentrations (A) and the calculated metal–mono-histidine concentrations (B). Values are expressed as means ± s.e.m.;n = 4. The curves and corresponding K_i values were calculated using a one-site binding model.

Table 2.

The apparent metal complex K_i values (μm); that is, the calculated metal–mono-histidine and metal–bis-histidine concentrations corresponding to 50% inhibition of human erythrocyte ⁶⁵Zn uptake, in the presence of 5 mm l-histidine (with 100 μm ADA) or 40 mm l-histidine

	5 mm histidine and 100 μm ADA		40 mm histidine
Metal	Ki metal–mono-histidine	Ki metal–bis-histidine	Ki metal–bis-histidine
Zinc	162 ± 0.36 μm	31.7 ± 3.96 μm	32.9 ± 4.0 μm
Nickel	0.10 ± 0.03 μm	64.5 ± 18.7 μm	54.7 ± 6.6 μm
Cobalt	0.04 ± 0.01 μm	37.2 ± 5.77 μm	57.2 ± 12 μm
Manganese	2.86 ± 0.8 μm	1.78 ± 1.27 μm	4.84 ± 0.7 μm

Values are expressed as means ± s.e.m.; n = 4.

These results would suggest that with the exception of manganese, the transporter discriminates poorly between bis-histidine complexes containing different metals. The concentration of metal–mono-histidine formed at various metal concentrations is proportional to the concentration of metal–bis-histidine formed under the same conditions and thus it is possible to plot ‘apparent’ inhibition curves for the metal–mono-histidine complexes (Fig. 1B) even though they do not represent any underlying mechanism. In a second set of experiments, erythrocytes were incubated with ⁶⁵Zn, various metal concentrations in the presence of 40 mm histidine and in the absence of ADA. Under these conditions the high histidine concentration forces the production almost exclusively of metal–bis-histidine complexes. The metal–bis-histidine inhibition curves thus produced are shown in Fig. 2.

Figure 2. The inhibition of ⁶⁵Zn uptake into human erythrocytes after 12-min incubation in the presence of 15 μm ⁶⁵Zn, 40 mm l-histidine and added Zn (□), Ni (Δ), Co (▪) and Mn (•) (0–500 μm).

The abscissa shows the calculated metal–bis-histidine concentrations. Values are expressed as means ± s.e.m.; n = 5. The curves and corresponding K_i values were calculated using a one-site binding model.

The calculated K_i values for the metal–bis-histidine concentrations (Table 2) are very similar to those generated in the presence of 5 mm histidine and 100 μm ADA. Again the K_i value for Mn–bis-histidine is much lower.

The above experiments, where the relative concentrations of the various metal–histidine complexes are varied by changing the metal or histidine concentrations, suggest that the metal–bis-histidine complexes, with the exception of manganese, have similar affinities for the transporter site. In order to provide further evidence for the zinc–bis-histidine complex being the preferred substrate for the transporter, ⁶⁵Zn uptake was measured over a range of histidine concentrations (0–10 mm) in the presence of varying concentrations of the buffering ligands (ADA or BSA). This will result in different concentrations of the various forms of zinc–histidine complex being formed at each concentration of histidine. This particular experiment, in which absolute uptakes are plotted, was conducted using erythrocytes obtained from one individual because we have previously found that there is considerable variability between individuals in the extent to which histidine stimulates metal uptake. These differences between individuals are consistent over long periods of time and presumably reflect varying degrees of expression of the transporter.

Figure 3A shows the uptake of ⁶⁵Zn at various concentrations of ADA and BSA plotted against the concentration of added histidine. It can be seen that there is a considerable scatter in the points at the various histidine concentrations. In Fig. 3B the ⁶⁵Zn uptake is plotted against the calculated concentrations of the Zn–bis-histidine complex formed under the various conditions and the scatter is much less, again suggesting that the metal–bis-histidine is the form of complex recognized by the transporter.

Figure 3. Zinc uptake, in nmol (10¹³ cells)⁻¹, in human erythrocytes incubated for 12 min in the presence of 0–10 mm l-histidine and various concentrations of ADA (25 μm (□), 50 μm (Δ), 200 μm (▪) and 400 μm (•)) or 120 μm BSA (▴).

Zinc uptake is plotted against histidine concentration (A) and the calculated zinc–bis-histidine concentration (B).

In the experiments depicted in Figs 1 and 2 the K_i values for the Zn–, Ni– and Co–bis-histidine complexes are similar but that for the Mn–bis-histidine complex is much lower suggesting that the transporter has a high affinity for this particular complex. It is possible that the inhibitory effect of manganese is not a result of a Mn–bis-histidine complex acting at the binding site but that it is instead a direct or indirect effect of manganese ions on the transporter. However ⁵⁴Mn uptake is stimulated by histidine in a dose-dependent manner and inhibition studies of ⁵⁴Mn uptake conducted in the presence of 40 mm histidine using 0–500 μm unlabelled manganese (Fig. 4) show that this histidine-stimulated uptake is saturable. The calculated K_i for the inhibition by Mn–bis-histidine of histidine-stimulated ⁵⁴Mn uptake is 12.7 μm which is comparable with the K_i of 4.8 μm (Table 2) for the effect of Mn–bis-histidine on histidine-stimulated ⁶⁵Zn uptake.

Figure 4. The inhibition of ⁵⁴Mn uptake into human erythrocytes after 12-min incubation in the presence of ⁵⁴Mn, 40 mm histidine and added (0–500 μm) Mn.

The abscissa shows the calculated Mn–bis-histidine concentrations. Values are expressed as means ± s.e.m.; n = 4. The curve was calculated using a one-site binding model. The calculated K_i is 17.1 ± 4.0 μm.

It has previously been shown that l-histidine at concentrations greater than 2 mm, in the presence of 1% BSA, stimulates ⁶⁵Zn uptake measured over 12 min in human erythrocytes (Horn et al. 1995). However the uptake of ⁶⁵Zn over 12 min at lower, more physiological, histidine concentrations is not significantly greater than that seen in the absence of histidine.

In Fig. 5A the duration of the erythrocyte incubation has been extended to 30 min and it can be seen that there does appear to be a stimulation of ⁶⁵Zn uptake by l-histidine at these lower concentrations. However, the effect of l-histidine is not significantly different from that of d-histidine until a concentration of 1 mm (P = 0.003) by which time the Zn uptake is 5 μmoles (10¹³ cells)⁻¹. In contrast, d-histidine merely chelates ⁶⁵Zn but does not stimulate uptake. The fact that the d-histidine line is flat rather than declining suggests that the BSA is working effectively in its role of buffering the free ionic zinc concentration. When these experiments were repeated using HEL cells, a stereospecific histidine stimulation of zinc uptake was seen (Fig. 5B). It is clear that the absolute zinc uptake per HEL cell was much greater than uptake per erythrocyte. It is probable that these nucleated, protein-synthesizing, erythroid-precursor cells have many more copies of the transporter expressed on their plasma membranes than the ‘end-stage’ erythrocytes. It is also possible that the more recently synthesized transporters have a higher affinity for metal–histidine complexes, perhaps because a functional element or unit is lost as the transporter ages.

Figure 5. Zinc uptake, in nmol (10¹³ cells)⁻¹, into cells incubated for 30 min with 15 μm ⁶⁵Zn and l-histidine (▪) or d-histidine (Δ) in the presence of 1% BSA.

Values are expressed as means ± s.e.m.; A, shows uptake into human erythrocytes;n = 5; B, shows uptake into human erythroleukaemic cells (HEL.92.1.7);n = 4.

Uptake inhibition studies were conducted to investigate whether the specificity and metal complex affinities of the HEL transporter were similar to those seen in human erythrocytes (Fig. 6). In these studies 5 mm histidine was used and the free ⁶⁵Zn concentration was buffered using ADA.

Figure 6. Zinc uptake, in nmol (10¹³ cells)⁻¹, into human erythroleukaemic cells (HEL.92.1.7) incubated for 30 min in the presence of 15 μm ⁶⁵Zn, 5 mm l-histidine and 100 μm ADA together with added Zn (□), Ni (Δ), Co (▪) and Mn (•) (0–500 μm).

The abscissa shows the calculated metal–bis-histidine concentrations. Values are expressed as means ± s.e.m.;n = 4. The curves and corresponding K_i values were calculated using a one-site binding model.

In Fig. 6 it can be seen that Zn, Ni, Co and Mn all inhibit histidine-stimulated ⁶⁵Zn uptake. If the degree of inhibition is plotted against the calculated metal–bis-histidine concentration, the values for the apparent K_i are similar in the case of Zn, Ni and Co but the value for Mn is much lower. As in the case of the erythrocytes there was wide disparity between the apparent K_i values for the various metal–mono-histidine complex concentrations (Table 3) suggesting that the mono-form of the complex is not a substrate for uptake. The metal–bis-histidine K_i values obtained with the HEL cells (Table 3) may be compared with the values for erythrocytes under identical conditions (Table 2). Whilst the HEL cell K_i values for the metal–bis-complexes, with the exception of manganese, are similar to each other, the HEL cell K_i values are much lower than the corresponding erythrocyte values.

Table 3.

The apparent metal complex K_i values; that is, the calculated metal–mono-histidine and metal–bis-histidine concentrations corresponding to 50% inhibition of ⁶⁵Zn uptake, in the presence of 5 mm histidine and 100 μm ADA. in human erythroleukaemic cells, HEL.92.1.7

Metal	Ki HEL Zn(his)2	Ki HEL Zn(his)
Zinc	2.64 ± 0.76 μm	650 ± 230 nm
Nickel	0.66 ± 0.38 μm	0.42 ± 0.34 nm
Cobalt	1.94 ± 0.67 μm	3.0 ± 0.1 nm
Manganese	0.0004 ± 0.0003 μm	90 ± 50 nm

Values are expressed as means ± s.e.m.; n = 4.

As with the erythrocyte, the question arises as to whether the inhibitory effect of manganese on histidine-stimulated zinc uptake is a result of competition by a Mn–bis-histidine complex or a direct effect of manganese ions on the transporter. In attempting to establish whether histidine will promote ⁵⁴Mn uptake the problem arises that the affinity of Mn for histidine and ADA is low compared to other metals. At low histidine concentrations, even in the presence of 100 μm ADA, there is a substantial concentration of free ionic ⁵⁴Mn which is also readily taken up by the HEL cells. As the concentration of histidine is increased, more of the histidine complex forms but this is at the expense of the concentration of free ⁵⁴Mn and the overall effect is that uptake changes little. In order to separate ionic manganese uptake from histidine complex-dependent uptake, the uptake of ⁵⁴Mn into HEL cells after 12 min was determined during incubation with a high (40 mm) concentration of l-histidine or d-histidine (Fig. 7).

Figure 7. Uptake of manganese, in nmol (10¹³ cells)⁻¹, after 12 min in the human erythroleukaemic cell line HEL.92.1.7.

/Cells were incubated in the presence of 0 mm, 40 mmd-histidine or 40 mm l-histidine. Values are expressed as means ± s.e.m., n = 4.

With d-histidine the ⁵⁴Mn uptake is inhibited, presumably because the free ionic ⁵⁴Mn concentration is reduced while the ⁵⁴Mn–bis-d-histidine complex is not a substrate for the transporter. However ⁵⁴Mn uptake increases slightly with 40 mm l-histidine because the free ionic ⁵⁴Mn concentration is reduced to the same extent as with d-histidine but the ⁵⁴Mn–bis-l-histidine formed is also a substrate for the transporter. The difference between the effects of l-histidine and d-histidine is significant (P < 0.02) confirming the existence of a histidine-stimulated manganese uptake into HEL cells.

Discussion

The experiments reported here and our previous studies have demonstrated that histidine promotes the uptake of various divalent metals into erythrocytes in a concentration-dependent, saturable and stereospecific manner. In addition, we showed that certain metals appear to compete with zinc for the histidine-stimulated uptake mechanism and that the rate of uptake and the degree of competition showed a clearer relationship to the concentration of the metal–bis-histidine complexes rather than that of the mono-histidine complex.

In the present study, we have investigated the ability of a range of metals to inhibit the uptake of ⁶⁵Zn in the presence of 5 mm histidine and the metal-ion buffer ADA. Under these conditions significant amounts of both mono- and bis-histidine complexes are formed. For each metal, the calculated concentrations of bis-histidine complex giving 50% inhibition of ⁶⁵Zn uptake were similar at both 5 mm and 40 mm histidine (i.e. independent of the concentrations of free ionic metal and of the mono-histidine complex). This is in agreement with our previous finding that the rate of metal uptake is best correlated with the calculated bis-histidine concentration (Horn et al. 1995; Horn & Thomas, 1996, 1997). It appears that the apparent K_i values of zinc–, nickel– and cobalt–bis-histidine complexes for inhibition of uptake of ⁶⁵Zn are very similar, indicating that the uptake site does not differentiate well between these metals when they are present as bis-histidine complexes. Further evidence for the importance of the metal–bis-histidine complexes comes from experiments carried out in the presence of 40 mm histidine, where very little metal–mono-histidine complex is formed. The calculated concentrations of zinc–, nickel– and cobalt–bis-histidine giving 50% inhibition of ⁶⁵Zn uptake under these conditions are similar to those found in the low-histidine experiments (Table 2), indicating that the presence of mono-histidine complexes is not required for binding to the transporter. This conclusion is supported by the results of experiments in which the concentration of the zinc–bis-histidine complex is varied by varying the concentration of the buffering ligand (Fig. 3).

The calculated Mn–bis-histidine K_i values in both the 5 mm and 40 mm histidine experiments are approximately 10 times lower than those for zinc, nickel and cobalt. This suggests that the binding site has a higher affinity for Mn–histidine complexes than for histidine complexes with zinc, nickel and cobalt. However, it is possible that manganese is simply having a non-specific inhibitory effect on histidine-linked zinc uptake, perhaps by a channel-blocking effect. That this is not the case is supported by the finding that unlabelled manganese inhibits ⁵⁴Mn uptake in the presence of 40 mm l-histidine (Fig. 4), with a K_i of 12.7 μm, indicating competition for the binding site.

The conclusion that the metal–bis-histidine complex is a substrate for uptake in erythroid cells is interesting in light of the reports that histidine can promote uptake of divalent metals into a variety of non-erythroid cells. Taylor & Simons (1994) investigated zinc uptake into cultured rat hepatocytes, but did not find evidence for significant promotion of uptake by zinc–ligand complexes as opposed to ionic zinc. Gachot et al. (1991) studied zinc uptake into rabbit renal proximal tubule cells, and found that histidine and cysteine gave a sodium-dependent stimulation of uptake at ligand:metal molar ratios between 1:1 and 8:1. Above a ratio of 16:1, inhibition was seen suggesting competition at a binding site. Glover & Hogstrand (2002) found evidence for a stereospecific zinc–bis-histidine-mediated pathway for absorption of zinc in rainbow trout intestine. Schmitt et al. (1983) described saturable non-active uptake of copper into isolated rat hepatocytes from copper–histidine solutions having a 1:1 molar ratio. They reported competition by cadmium, zinc, cobalt and manganese and concluded that ionic copper was the transported species, but did not consider the significance of the wide differences in affinity for histidine of the different metals, which would have affected the relative concentrations of complexes formed. A subsequent study (Darwish et al. 1984) used a mixture of ⁶⁴Cu and ³H-labelled histidine with a sufficient excess of histidine to favour the formation of copper–bis-histidine. They concluded that the copper entered the hepatocytes but the histidine did not, and that histidine promoted copper uptake from plasma by mobilization as copper–bis-histidine from binding proteins such as albumin.

In a study of copper uptake into rat hypothalamic tissue slices, Hartter & Barnea (1988) reported the presence of two separate histidine-linked uptake sites with differing characteristics. For both sites, the presence of histidine gave a greater uptake than an equivalent concentration of copper alone, suggesting a genuine facilitatory effect. A subsequent study using the same preparation (Katz & Barnea, 1990) did not confirm the previous finding of two distinct uptake sites. They suggested that the key feature necessary for stimulation of metal uptake was coordination of copper with at least three nitrogen atoms, explaining why the copper–bis-histidine complex with coordination across the imidazole nitrogens was so effective. The effect of histidine was not stereospecific. Histidine, in excess of the theoretical concentration necessary for complex formation, inhibited copper uptake, suggesting competition at a histidine-binding site. Dual-label studies with histidine and copper complexes suggested that the uptake kinetics of histidine and copper were different, implying dissociation of the complex before internalization. In previous experiments we have shown that copper is able to inhibit histidine-stimulated zinc uptake into human erythrocytes in the presence of 40 mm histidine. The calculated K_i for copper–bis-histidine was 88.5 μm (Horn & Thomas, 1997). In all our experiments, metal uptake increased at histidine concentrations up to 40 mm, showing that free histidine does not inhibit the transporter.

McArdle et al. (1988) reported that histidine stimulated copper uptake in acutely dissociated mouse hepatocytes. The effect was maximal at a histidine concentration which would have given predominant formation of the copper–bis-histidine complex. The stimulation effects were not energy-dependent but were inhibited by N-ethylmaleimide, implying a role for sulphydryl groups. We have found that histidine-stimulated zinc uptake into human erythrocytes is not inhibited by N-ethylmaleimide treatment (Horn & Thomas, 1996). Bingham & McArdle (1994) described copper uptake from Cu–bis-histidine in rat liver plasma membrane vesicles and in cultured rat hepatocytes. The uptake was saturable and inhibited by zinc, albumin and N-ethylmaleimide. Hilton et al. (1995) investigated the role of the ceruloplasmin receptor in vesicles prepared from human placental tissue, and suggested that uptake of copper from the bis-histidine complex involved binding sites on this receptor. This receptor does not appear to correspond to any of the ‘ceruloplasmin receptors’ previously reported in erythrocytes, but the authors suggested that it was similar to the previously described receptor from liver endothelium.

The functional significance of divalent metal uptake in the mature erythrocyte is unclear. It is possible that erythrocyte metal uptake allows the cell to act as a reservoir of metal and play a role in inter-organ metal transport. However, the erythrocyte is an end-stage cell and it is possible that any uptake systems remaining in the cell membrane are simply vestiges of those transporters essential at earlier stages. With this in mind, we have investigated histidine-stimulated metal uptake in the nucleated human erythroleukaemic cell line HEL.92.1.7, which was considered to represent the earlier stages of blood cell formation where the erythrocyte or leucocyte precursor is carrying out protein synthesis. At this time the cell will have a requirement for zinc for the synthesis of metalloproteins such as carbonic anhydrase and superoxide dismutase, and zinc in the form of histidine complexes is a potential substrate for uptake. Direct evidence for production of manganese superoxide dismutase in HEL cells is lacking, but another human erythroleukaemic cell line (K562 cL6) has been shown to contain MnSOD (Jia et al. 1995), and it would be expected that normal earlier erythroid cells would synthesize mitochondrial MnSOD. As new membrane transport proteins cannot be synthesized and renewed in the mature erythrocyte, it would be predicted that the capacity for zinc and manganese uptake would decline as the cells age. The HEL cell (radius, 12.1 ± 1.8 μm; Schafer & Buettner, 1999) has approximately 3.4 times the surface area of an erythrocyte, and thus our results show that the maximum rate of the histidine-stimulated zinc uptake per unit of cell surface area in HEL cells is approximately 10 times greater than that in erythrocytes. In addition, the apparent affinities of the uptake site for zinc–, nickel– and cobalt–bis-histidine complexes in HEL cells are similar to each other and are 50- to 100-fold greater than those found in erythrocytes (Table 1). The affinities reported here for metal–bis-histidine uptake into HEL cells are similar to those reported elsewhere for histidine-stimulated copper uptake, for example Darwish et al. (1984) reported a K_m of 15 μm for uptake of copper–bis-histidine into hepatocytes. The fact that the affinity for metal–bis-histidine complexes in the erythrocyte precursor cell is much greater that in the mature erythrocyte suggests that these complexes might provide a functional substrate for metal uptake in erythrocyte precursor cells continuously exposed to plasma histidine concentrations. Thus the higher affinity and greater capacity of the transporter as expressed in the erythroid precursor cells is commensurate with a role in the supply of metals for metalloprotein synthesis.

In conclusion, we have shown that the mechanisms of histidine-stimulated metal uptake in erythrocytes and erythroleukaemic cells share many properties; they are saturable and extremely stereospecific, and show a preference for the bis-histidine complex rather than the mono-histidine complex (although binding of the mono-histidine complex cannot be excluded). The uptake mechanism does not discriminate between several biologically important divalent metals, but both cell types have a much higher affinity for manganese–histidine complexes than those of the other metals. However, the extreme stereospecificity found in both cell types appears to differentiate them from non-erythroid cells, suggesting that the uptake mechanism responsible is more highly expressed in cells of the leucocyte and erythroid lineages.