Copper Uptake In Mammalian Cells

A human protein potentially involved in high-affinity copper uptake, hCtr1, has been identified by complementation of a yeast Ctrl ctr3 double mutant (52). Expression of hCTRl in the uptake-deficient yeast mutant restores high-affinity iron uptake and permits growth on nonfermentable carbon sources. Furthermore, the human gene renders the yeast cells sensitive to copper overload, corrects the deficiency of Cu,Zn superoxide dismutase, and increases the cellular copper level as measured by atomic absorption spectrophotometry (52). Recently, human fibroblasts, transfected with hCTRl cDNA have been shown to have a dramatically increased capacity for 64Cu uptake, providing direct evidence that hCtr1 can function in copper uptake in human cells (see Fig. 1). So far, no human disorders of copper metabolism have been associated with the hCTRl gene, so it is presently unclear if other transporters might contribute to the uptake of copper ions in human cells.

A murine homolog of the hCTRl gene (mCTRl) has recently been cloned and shown to functionally complement yeast cells deficient in high-affinity copper uptake (54). The possible generation of a mCTRl knockout mouse might provide additional insights into the physiological role of Ctr1 in mammalian copper metabolism. The mouse protein is highly homologous to hCtr1 (Fig. 2), and both are encoded by four exons separated by introns that are located at identical positions (53,54). Interestingly, both proteins are also highly homologous to the predicted protein product of a rat cDNA


Fig. 2. Comparison of the protein sequences of hCtr1, mCtr1, and rCtr1, which consist of 190, 196, and 187 amino acid residues, respectively. Only derivations from the hCtr1 sequence are shown. Hyphens indicate the absence of an amino acid at the corresponding position. The predicted transmembrane segments in the hCtr1 sequence are shown in bold (52). Nucleotides flanking the three introns in the hCtr1 and mCtr1 coding sequences respectively are underlined (53,54). The Genbank accession no. of the hCTR1 and the rCTR1 sequences are U83460 and AF268030, respectively. The murine sequence was compiled from four EST sequences as described in ref. 53 (Genbank accession nos. AA124593, AA4867805, AI195572, AA250186).

Fig. 3. Comparison of the DNA direct repeats in the beginning of the hCTR1, mCTR1, and rCTR1 coding sequences. The initiation codon of the three genes is boxed. A 9-bp insertion in the first repeat of the human and rat sequence and a 3-bp insertion in the third repeat of the murine sequence have been "pulled out" to emphasize the similarity of the repeats. The methionines are shown in bold.

Fig. 3. Comparison of the DNA direct repeats in the beginning of the hCTR1, mCTR1, and rCTR1 coding sequences. The initiation codon of the three genes is boxed. A 9-bp insertion in the first repeat of the human and rat sequence and a 3-bp insertion in the third repeat of the murine sequence have been "pulled out" to emphasize the similarity of the repeats. The methionines are shown in bold.

sequence, rCTR1 (Fig. 2), but there is presently no experimental evidence relating to the function of this protein. Conflicting predictions have been presented for the initiation codon of the murine CTR1 gene (53,54). In Figs. 2 and 3, we have assumed that the initiation codon is the first ATG codon in the mCTR1 reading frame of the mouse cDNA.

The human Ctr1 protein is only 190 amino acids long and is predicted to contain three transmembrane segments like the yeast transporters. Interestingly, the primary structures of hCtr1 and a high-affinity copper transporter, Ctr4, from S. pombe suggest that these proteins have evolved by fusion of domains corresponding to the N-terminal domain of Ctr1 and the C-terminal domain of Ctr3 (48). It is remarkable that the hCtr1 protein can substitute for the yeast Ctr1 protein, even though the N-terminal domain of the human protein is considerably shorter and contains only two putative copper-binding motifs, MXXM, whereas the yeast protein has 11 copies of this motif.

Interestingly, the beginning of the mouse and the rat CTR1 coding sequences contains four imperfect repeats of a 15-bp-long consensus sequence ATG (A/C)AC CA(T/C) ATG G(G/A)(G/C) that may encode an MXXM motif, whereas only two of these repeats are present at the beginning of the human CTR1 sequence (Fig. 3). Thus, it appears that the number of MXXM motifs even in these very closely related proteins may vary because of the expansion or deletion of such direct repeats in the DNA sequence. In this connection, the considerable sequence deviation between the N-terminal domains of the hCtr1, mCtr1, and rCtr1 proteins should be contrasted with the completely conserved amino acid sequences of their C-terminal domains (Fig. 2).

By DNA sequence analysis, Zhou and Gitschier (52) has identified another human gene, hCTR2, which is similar to hCTR1. The hCtr2 protein consists of 143 amino acid residues and is predicted to have the same membrane topology as hCtr1, but it lacks the counterpart of the putative copper-binding N-terminal domain. So far, there is no evidence of an involvement of hCtr2 in copper uptake, as expression of the hCTR2 gene failed to complement the uptake deficiency of yeast mutant strains (52). In agreement with this, copper uptake by human cells was not increased upon transfection with an hCTR2 expressing plasmid (Fig. 1). The hCtr2 protein resembles the S. cerevisiae Ctr2 protein, as they both lack the counterpart the N-terminal domain of the Ctr1 proteins, the domain that has been implicated in the binding of copper by these proteins. Consequently, it has been speculated that the physiological function of hCtr2 and Ctr2 may be in the transport of substrates other than copper ions (49,53).


In S. cerevisiae, high-affinity copper uptake mediated by Ctrl and Ctr3 appears to be specific for transport of Cu(I) formed by prior reduction by the cell-surface Cu/Fe reductases Frel/Fre2. The Frel and Fre2 proteins were originally identified as plasma membrane ferric reductases required for iron uptake (55,56). Subsequent studies showed that the ability of Ctrl and Ctr3 to function in high-affinity Cu uptake is dependent on the Frel reductase, as Cu uptake mediated by Ctrl and Ctr3 was reduced to 90% and 70%, respectively, in the absence of Frel (45). Comparison of copper uptake by mutant strains lacking Frel or Fre2 or both activities showed that Fre2 provides a minor but significant contribution to transport-related copper reduction under conditions of low Cu(II) concentration (57). However, even the fre1 fre2 double mutant was not as deficient with respect to copper uptake as mutants lacking the Ctrl and Ctr3 proteins, perhaps owing to residual Cu reductase activity that might be ascribed to the FRE3-FRE7 gene products. Furthermore, it has been shown that extracellular reductans such as ascorbate may suppress the dependence of Cu uptake on the plasma membrane reductase activities, presumably by facilitating nonenzymatic reduction of Cu(II) (57,58).

In mammals, there is no consensus concerning the oxidation state of copper during uptake or the possible involvement of plasma membrane reductase activities. Several reports have suggested that copper is preferentially taken up in the form of Cu(II). Thus, the Cu(I)-specific chelator bathocuprione sulfonate did not inhibit the uptake of copper in the form of CuHis2 into trophoblast cells isolated from human term placenta (59). Ascorbate, which reduces Cu(II) to Cu(I), actually impaired apical uptake of copper from CuCl2 into human differentiated intestinal Caco-2 cells (60). In contrast, Percival and Harris (61) reported that ascorbate enhances copper transport from ceruloplasmin into human erythroleukemic K-562 cells, indicating that copper is taken up as Cu(I).


It has not been generally settled whether copper uptake is an energy-requiring process. In S. cerevisiae, it was found that Cu accumulation was saturable and subject to inhibition by the respiratory inhibitor sodium azide and the uncoupler dinitrophenol, particularly when cells were grown on nonfermentable carbon sources (25). These results suggested that ATP is required to sustain net uptake of copper. Conflicting results have been reported concerning the energy requirement of copper uptake into mammalian cells. McArdle et al. (62) found no evidence for an ATP requirement in mouse hepatocytes, as inhibitors that substantially reduced intracellular ATP levels failed to reduce copper uptake. Studies of cultured trophoblast cells from human placenta also indicated that ATP was not required for copper uptake (59). A recent study has suggested that apical uptake of Cu in human differentiated intestinal Caco-2 cells may not even be carrier mediated, as uptake appeared to be nonsaturable at copper concentrations up to 80 |M (60). However, Arrendo et al. (63) reported that copper uptake into Caco-2 cells was saturated at external copper concentrations of 4-6 |M.

Given the similarity and functional equivalence of the human and yeast Ctr proteins, Caco-2 cells would be expected to have similar substrate specificity and uptake mechanism as yeast cells. The different results concerning the requirements for energy and substrate reduction in yeast and mammalian cells might, in part, derive from the failure to distinguish between passive adsorption of Cu on the cell surface and true intracellular uptake of Cu. It appears that the great majority of copper associated with cells of S. cerevisiae belongs to a pool of rapidly exchangeable copper adsorbed on the cell surface and that this exchange was not energy dependent (25,64,65). Most of the cell-associated copper in mouse hepatocytes (66) and in Escherichia coli cells (67) also appears to be associated with such a rapidly exchangeable pool, which may severely complicate studies of the copper-uptake process.


In yeast cells, copper uptake appears to be limited by the expression of the CTR1 and CTR3 genes (45,46) and this limitation is part of an intricate regulatory loop that controls the intracellular copper concentration. Thus, copper transcriptionally downregulates the CTR1 and CTR3 genes as well as the FRE1 gene, because their expression depends on a copper-sensitive transcriptional activator Macl (68,69). Furthermore, the presence of a high copper concentration also inhibit copper uptake by inducing rapid degradation of both Macl and Ctrl (70,71), whereas copper does not affect the stability of Ctr3 (47). Internalization of Ctrl protein can be seen when cells are exposed to copper; however, the degradation of Ctrl seems to be independent of endocytosis, as it also takes place in strains defective in endocytosis and vacuolar degradation. Together, these mechanisms assure a tight control of the intracellular copper concentration in yeast cells. It is presently unknown how copper uptake in mammalian cells is regulated. The finding that 64Cu accumulation by human fibroblasts was greatly increased by transfection with the hCTR1 gene suggests that Cu uptake in human cells, in analogy with yeast, is normally limited by the expression of the hCTR1 gene (Fig. 3). However, expression of the mammalian CTR1 genes is apparently not regulated in response to the cellular copper availability (54). In this connection, it is noteworthy that efflux-deficient fibroblasts from Menkes disease patients hyperaccumulate copper in contrast to yeast cells deficient with respect to the homologous Ccc2 protein (72). This difference might suggest that the control of copper uptake in human cells is less stringent than in yeast cells.

In yeast cells, copper is indirectly required for ferrous iron uptake, because iron uptake depends on a multicopper ferroxidase, Fet3, located in the plasma membrane. This link between copper and iron metabolism is reflected in the finding that FET3 and several of the genes involved in transport of copper to Fet3 are regulated by iron. Thus, the FRE1 and FRE2 genes encoding Fe (III)/Cu(II) membrane reductases as well as the CCC2 and ATX1 genes, which encode the copper pump, Ccc2 and its chaperone, respectively, are all regulated by the iron-responsive transcription factor Aftl (28,73,74).

The human homolog of Fet3, ceruloplasmin, is also involved in iron transport, but the major function of ceruloplasmin seems to be mediation of iron export via the protein transferrin (75,76). However, recent results suggest that ceruloplasmin also stimulates cellular iron uptake (77). In analogy with its yeast homolog the human ceruloplasmin gene is activated by iron deficiency (77,78), but conflicting results have been reported concerning the regulation of this gene by copper. McArdle et al. (17) found no effect of copper on ceruloplasmin expression in mouse hepatocytes, whereas Daffada and Young (79) observed an increase of ceruloplasmin mRNA in HepG2 cells after administration of copper.

0 0

Post a comment