The Mechanism of Copper Uptake Mediated by Human CTR1

Cellular copper uptake is a prerequisite for the biosynthesis of many copper-dependent enzymes; disruption of copper uptake results in embryonic lethality. In humans, copper is transported into cells by hCTR1, a membrane protein, composed of 190 amino acids with only three trans-membrane segments. To provide insight into the mechanism of this unusual transporter, we characterized the functional properties of various hCTR1 mutants stably expressed in Sf9 cells. Most single amino acid substitutions involving charged and potential copper-coordinating residues have some influence on the Vmax and Km values for copper uptake but do not greatly alter hCTR1-mediated copper transport. However, there were two notable exceptions. Replacement of Tyr156 with Ala greatly reduced the maximal transport rate without effect on the Km value for copper. Also, replacement of His139 in the second trans-membrane segment with Arg caused a dramatic increase in the rate of copper uptake and a large increase in the Km value for copper. This effect was not seen with an Ala replacement, pointing to the role of a positive charge in modulating copper exit from the pathway. Truncated mutants demonstrated that the deletion of a large portion of the N-terminal domain only slightly decreased the apparent Km value for copper and decreased the rate of transport. Similar effects were observed with the removal of the last 11 C-terminal residues. The results suggested that the N and C termini, although nonessential for transport, may have an important role in facilitating the delivery of copper to and retrieving copper from, respectively, the translocation pathway. A model of how hCTR1 mediates copper entry into cells was proposed that included a rate-limiting site in the pore close to the intracellular exit.

Copper is an essential element required for many important cellular reactions that rely on the redox properties of this metal (1)(2)(3). Like many of the essential trace metals, inappropriately high concentrations of copper are toxic, both at the cellular and whole organism level. Thus intracellular copper is tightly regulated so that sufficient amounts are absorbed and excess copper is efficiently excreted (2, 4 -10). Copper homeostasis at the cellular level involves the entry and exit of copper as well as the processes whereby copper is delivered to specific subcellular locations. The regulatory mechanisms that govern the cellular localization of proteins involved in copper transport are currently under investigation, and some progress has been made in understanding these path-ways (11)(12)(13)(14). However, studies focused on the molecular mechanism of copper transporters, mediating both entry and exit of copper, are still at an early stage.
Copper entry into cells may be mediated by either of two transporters, divalent metal transporter 1 (DMT1) 2 and copper transporter 1 (CTR1). DMT1, a relatively nonspecific divalent metal ion transporter, is believed to play a role in copper(II) uptake in the small intestine (15), whereas CTR1, a more specific transporter for copper (and perhaps silver), is thought to mediate Cu(I) uptake in many tissues. The physiological importance of hCTR1 is demonstrated by its essential role in embryonic development (16,17). Recent studies have also provided the first insights into structure-function relationships in hCTR1 (18,19).
hCTR1 is a glycosylated trans-membrane protein, which consists of 190 amino acids and has three putative trans-membrane segments. The N terminus is extracellular and has a single N-linked glycosylation site at Asn 15 , and the C terminus is localized within the cell (20). Recent studies demonstrated that mutations of the methionines in the N-terminal domain (Met 40 -Met 45 ) and in the second trans-membrane segment (Met 150 and Met 154 ) decrease the ability of hCTR1 to transport copper into cells (19). Complementation assays applied to a yeast homologue of hCTR1 (yCtr3) also demonstrated the structural importance of residues in the third trans-membrane segment. The glycine residues Gly 167 and Gly 171 in this segment have been suggested to contribute to helix packing interactions and oligomerization of hCTR1 to form trimers (18). Additionally, Aller et al. (18) used tryptophan scanning of the third trans-membrane segment of yCtr3 to provide evidence that other residues (Ile 196 ,Ile 197 ,Ser 198 ,Cys 199 , and Arg 207 ) may have a role in protein folding and structural integrity. Because of the sequence homology between yCtr3 and hCTR1, these residues may also be relevant to hCTR1.
Although these initial studies provided first glimpses into the molecular organization of CTR1, the mechanism of copper translocation by this transporter remains uncharacterized. Are there specific copperbinding sites that are alternately exposed to the outside and the inside of the cell via protein conformational changes? Are there amino acid residues that control entry of copper into a translocation pathway? Does copper migrate through CTR1 via a chain of "essential" residues or is there a "channel-like" pathway through which copper migrates without forming strong contacts with the pore-lining amino acid residues? Does the exit of copper from the transporter represent a regulated event? What is the rate-limiting step of transport? The answers to these questions are currently unknown. To begin addressing these important mechanistic issues, it is essential to go beyond complementation studies and provide direct and quantitative analyses of hCTR1 transport function. Consequently, we have utilized stable overexpression of hCTR1 in insect cells and carried out measurements of radioisotopic copper uptake over a range of copper concentrations for various hCTR1 mutants. The roles of the intramembrane charged residues, residues potentially involved in copper coordination, and both the N and the C terminus were characterized. We combine our results with those of previous studies to propose a model of how hCTR1 mediates copper entry into cells.

EXPERIMENTAL PROCEDURES
Cell Maintenance-Sf9 cells were cultured in Ex-Cell 420 media (JRH Biosciences, Lenexa, KS). Cells were maintained attached in T-75 flasks and were split when 90% confluent. Mutant hCTR1 constructs were maintained in Ex-Cell 420 media that was supplemented with 0.015 mg/ml blasticidin S (Invitrogen) to ensure retention of the expression vector.
Generation of Mutant hCTR1 Constructs-hCTR1 constructs with various mutations and deletions were generated by PCR using the fulllength hCTR1 cDNA as a template (20). The outside primers (5Ј forward primer, 5Ј-GAATTCATGGATCATTCCC-3Ј, and 3Ј reverse primer, 5Ј-CCGCGGAACAACTTCCCACTGC-3Ј, which contain EcoRI and SacII restriction sites at the 5Ј and 3Ј ends, respectively) and internal primers specific for the mutation generated (TABLE ONE) were utilized in the Pfu polymerase (Stratagene, La Jolla, CA.)-mediated reaction for full-length constructs. Primers designed to generate truncation constructs were used with either the 5Ј forward or 3Ј reverse primers, respectively. The PCR products were ligated into pIB/H5 TOPO vector using the TOPO-cloning technology (Invitrogen). The two C-terminal truncations were PCR-amplified using a construct containing Asn 15 mutated to glutamine (20). This enabled us to monitor expression of the constructs that lacked the epitope for the C-terminally directed antibody. For these constructs we had to utilize an antibody directed against the N terminus that recognizes only the unglycosylated form of hCTR1. Plasma membrane protein expression of N15Q is the same as the plasma membrane expression of Wt hCTR1 in this system. The presence of desired mutations was confirmed by sequencing the entire cDNA.
Analysis of Protein Expression-For protein expression, each pIB plasmid encoding the mutant hCTR1 was transfected into Sf9 cells using Cellfectin reagent following the manufacturer's protocol. Cells were selected using 80 g/ml blasticidin-HCl for 2 weeks. Each stable cell line was then examined for hCTR1 protein expression by Western blot analysis using antibody directed against the C terminus of hCTR1 or, in the case of the two C-terminal deletions, an antibody raised against the N-terminal peptide (SHHHPTTSASHSHG). For these experiments, cells were collected and homogenized using a Dounce homogenizer. The "total" membrane preparations were fractionated using a sucrose gradient, as described previously (20). To visualize hCTR1, 20 g of plasma membrane protein was resolved using a 12% SDS-PAGE Laemmli gel. The proteins were electroblotted to a nitrocel- lulose membrane and blocked using phosphate-buffered saline, pH 7.4, containing 5% milk. After blocking, the membranes were incubated with anti-C-terminal antibody (1:25,000, phosphate-buffered saline, pH 7.4, 0.1% Tween 20, 0.5% milk), and detection was carried out using a secondary goat anti-rabbit antibody (1:10,000, phosphate-buffered saline, pH 7.4, 0.1% Tween 20, 0.5% milk). There was some variation in the plasma membrane abundance among the various mutant proteins. The plasma membrane expression levels in the stably expressing Sf9 cell lines were normalized with respect to the expression of wild type hCTR1. The normalization allowed a direct comparison of the V max values among the various mutants. The quantitation was accomplished by densitometric determination of nonsaturated Western blots of the plasma membrane fractions using the Alpha Innotech FluorChem 5500 imaging system (San Leandro, CA). The V max values for transport were then corrected by utilizing the normalized value for protein expression for the mutants compared with wild type hCTR1 expression. A notable feature of the proteins expressed in stable cell lines is the relative lower level of higher order multimers that are stable in SDS gels compared with virally infected cells (20). 64 Cu Uptake Experiments-Experiments to evaluate the ability of mutant hCTR1 proteins to transport copper were carried out as described previously (20) with slight modifications. Briefly, copper uptake was performed in 12-well plates seeded with 1 ϫ 10 6 cells per well and allowed to incubate at 27°C with cell media. After the cells had attached, they were washed with uptake buffer (150 mM NaCl, 2.5 mM MgCl 2 , 25 mM Hepes, pH 7.4) and incubated at 27°C for 30 min. To initiate transport, cells were incubated with 64 Cu (MIR Radiological Sciences) in uptake buffer for either 5 or 45 min at six different copper concentrations (0.5, 1.0, 2.0, 5.0, 10.0, and 20.0 M CuCl 2 ). To halt transport at the end of the uptake period, cells were washed three times with ice-cold incubation buffer supplemented with 10 mM EDTA. Cells were then lysed using 0.1 N NaOH, and aliquots were counted using a ␥-counter (Packard Cobra II) and also used to determine protein concentrations. Radioactive copper uptake was calculated as reported previously (20). Briefly, the amount of radioactivity remaining in the cells, following the washes, is converted to copper influx by multiplying by the specific activity of the 64 Cu at that copper concentration. The copper uptake is normalized by the protein concentrations for each well. This normalization procedure takes into account fluctuations in cell number at the start of transport measurements and any loss of cells during washes. In order to readily compare the effects of modifications in hCTR1, the rate of uptake of copper into Sf9 cells that did not contain transfected hCTR1 was subtracted from the total copper uptake into the transfected cells.

Stable Expression of hCTR1 in Sf9 Cells Improves Quantitative Measurements of Copper Uptake-
The trans-membrane portion of CTR1 contains a number of amino acid residues that may critically contribute to the translocation of Cu(I) across the membrane (Fig. 1). These include residues that could be involved in copper coordination, such as Cys, His, and Met, as well as negatively charged and polar residues (Asp, Glu, and Tyr). In addition, both the C terminus and especially the N terminus contain copper-coordinating residues with only a poorly defined role in copper uptake. To investigate the functional role of these various amino acids, we generated the appropriate mutations and truncations and stably expressed these proteins as well as wild type hCTR1 in Sf9 cells. Stable expression was expected to facilitate comparative analysis of mutants, because variations in virus titer, unavoidable in the transient transfection experiments, can be eliminated. Because func-tional expression of hCTR1 in stably transfected cells has not been described previously, we first characterized the properties of wild type hCTR1 in these cells. Fig. 2 (inset) illustrates the distribution of wild type hCTR1 in membrane fractions of Sf9 cells. The stably expressed hCTR1 is present predominantly in the plasma membrane fraction consistent with its function and localization reported previously in mammalian cells (14). Two protein bands are observed following immunostaining with anti-hCTR1 antibody, a 28-kDa band corresponding to glycosylated hCTR1 monomer, and a 24-kDa band that is unglycosylated hCTR1. In some membrane preparations a C-terminal degradation product of 17 kDa is also seen. No protein bands are seen in any of the fractions isolated from Sf9 cells lacking hCTR1 expression plasmids.
The stably expressed hCTR1 is functional, as evident from the kinetics of copper uptake into cells shown in Fig. 2. Specifically, expression of hCTR1 protein is accompanied by a large (about 5-fold) increase in copper uptake (see also TABLE TWO). In this figure we also show the hCTR1-dependent uptake that results from the subtraction of the copper uptake into untransfected cells from the uptake into cells expressing hCTR1. All subsequent data presented on the kinetics of copper uptake into Sf9 cells have been corrected by this subtraction procedure.
Expression and Function of N-terminally Truncated hCTR1-It was hypothesized that Met-and His-rich sequences in the N-terminal domain of hCTR1 could be important for copper binding and concentrating metal prior to entry into the trans-membrane pathway. To examine the role of multiple Met residues and His residues in guiding copper toward the permeation pathway, we generated three N-terminal deletions of hCTR1, and we investigated the effect of these deletions on the V max of transport and the K m value for copper for each construct. The truncated constructs include the following: (i) hCTR1 with the deletion of the first 34 amino acid residues (MG34), which removes the first Met repeat and His-rich segment (Fig. 1A); (ii) the construct in which the second methionine repeat is also deleted (MN53); and (iii) hCTR1 that lacks the entire N terminus (Met 69 ) (Fig. 1A). Each of the three N-terminal truncations is expressed well (Fig. 3A, inset) and was localized to the plasma membrane, indicating that the N-terminal region is not essential for protein stability or plasma membrane delivery, at least in Sf9 cells.
The truncated hCTR1 molecules showed very different functional properties. Copper uptake by MG34 was slightly elevated compared with the wild type (Wt) protein ( Fig. 3A and TABLE THREE). The V max value of this mutant is close to the Wt rate, and the K m value shows only a 2-fold increase. The two other truncation mutants, MN53 and Met 69 , were significantly impaired in their ability to facilitate copper uptake. In the case of MN53, about 60 -70% reduction in copper transport was observed; nevertheless, transport activity can be measured, in contrast to the results of yeast complementation studies of a similar mutant (19) (see "Discussion"). The MN53 mutant shows a slightly decreased K m for copper (TABLE THREE). The phenotype of the Met 69 mutant is even more pronounced. Copper uptake by cells expressing Met 69 is only about 10% of cells expressing Wt (TABLE THREE) and is only about 20% greater than Sf9 cells that do not express hCTR1. Thus, cells expressing hCTR1 lacking the entire N terminus have essentially lost the ability to transport copper.
Expression and Function of C-terminally Truncated hCTR1-We also investigated the potential functional importance of the C-terminal intracellular tail of hCTR1. We have shown previously, in baculovirusinfected Sf9 cells, that the C terminus can be epitope-tagged without significant disruption of hCTR1 function (20). However, the C terminus was shown to be involved in protein-protein interactions with the cop-Copper Transport Mediated by hCTR1 NOVEMBER 4, 2005 • VOLUME 280 • NUMBER 44 per chaperone, Atx1, in vitro (21,22), suggesting that it may play a role in copper uptake. To test this hypothesis, two C-terminal truncations were generated, CTK178 and CTD184 (Fig. 1A). CTK178 is a construct in which hCTR1 is terminated at Lys 178 , and CTD184 is a construct in which hCTR1 is terminated at Asp 184 . The CTD184 construct lacks Cys 189 , a putative copper-coordinating residue, and CTK178 has most of the C terminus deleted up to a Lys 177 -Lys 178 sequence, which may be important for establishing membrane topology. Both C-terminally truncated constructs are expressed at comparable levels, and both are delivered to the plasma membrane (Fig. 3B, inset).
The functional characterization of these mutants showed that CTK178 and CTD184 were able to mediate copper uptake, although at reduced levels compared with Wt ( Fig. 3B and TABLE THREE). The transport capacity of the expressed CTD184 is moderately affected (its transport activity is at about 70% of the Wt hCTR1 levels and the K m value is very similar to the Wt). However, removal of six additional amino acids significantly reduces the capacity for copper uptake (40% of Wt hCTR1). Most interestingly, the effect of C-terminal deletion on the apparent K m values is opposite the effect of N-terminal truncations, i.e. the K m value for copper is increased (TABLE THREE), although the significance of only a 2-3-fold change in the K m values for copper is difficult to interpret. Taken together, these results suggest that the C terminus does influence, to some extent, the efficiency of copper transport mediated by this protein, perhaps at the copper release step (see "Discussion").
Mutants of Met Residues in the Trans-membrane Domains-The trans-membrane portion of hCTR1 contains four Met residues. The first pair, Met 69 and Met 81 , is located in the first trans-membrane seg-FIGURE 1. Amino acid residues of hCTR1. A, diagram of hCTR1 showing relative placement of putative trans-membrane domains and hCTR1 residues mutated in this study. Methionine residues that are modified in this study are colored red. Charged residues are colored green. Tyrosine residues are colored yellow. Cysteine residues are colored blue. Truncation sites of mutants are distinguished by arrows. B, protein alignments of putative trans-membrane domains from selected mammalian Ctr1 proteins and yeast Ctr1/3. Residues of identity are displayed in red. Green residues are conserved residues. Blue residues are semi-conserved. ment (TMS1); the second pair, Met 150 and Met 154 , belongs to the second trans-membrane segment, TMS2 (Fig. 1). Met 150 and Met 154 are conserved in every high affinity copper transport protein of the Ctr family (Fig. 1B). To investigate the functional role of these residues, we expressed a double mutant, M140I//M154I, and we characterized its ability to transport copper. Protein levels of the double mutant at the plasma membrane were similar to the levels of Wt hCTR1 (data not shown); however, the rate of transport was reduced to about 30% of the rate of Wt protein (TABLE FOUR). Most interestingly, this decrease in transport occurs with no change in the K m for copper.
Two other methionine mutants, M69I and M81I, are capable of transporting copper: the former at 150%, when corrected for differences in protein expression at slightly higher rates than hCTR1, and the latter at about 50% of the rate of wild type transporter. Neither of these mutations resulted in a more than 2-fold change in the K m value for copper uptake (see TABLE FOUR). Thus, it seems unlikely that a specific and essential role can be assigned to either of these Met residues.
The Synergistic Effect of Cys Substitutions-The above results suggested that neither pair of the intramembrane Met residues was essential for copper transport, although clearly the second pair of methionines (Met 150 and Met 154 ) was more important for hCTR1 function. Sulfur-containing amino acids are frequently found at copper coordina-tion sites in proteins, and we turned our attention to the two Cys residues Cys 161 and Cys 189 . By using baculovirus-mediated infection of Sf9 cells, we showed earlier (20) that when each of these residues was mutated individually, the proteins were capable of transporting copper with a similar apparent affinity as Wt hCTR1 and with comparable V max values. By utilizing stable cell lines in the present work, we have confirmed that observation ( Fig. 4 and TABLE FOUR). We then generated the double C169S/C181S mutant and examined its properties. Unexpectedly, and in marked contrast to single amino acid substitutions, mutations of both cysteines markedly decreased the ability of this construct to transport copper across the membrane. The V max value of the double mutant was only 27% of the Wt with little change in the K m value for copper (TABLE FOUR). Clearly, simultaneous replacement of the

Kinetic constants for copper uptake by Wt-hCTR1
Data from copper uptake experiments at various copper concentrations were analyzed using the Michaelis-Menten equation to determine the copper concentration at which uptake was half-maximal (K m ). K m values (M) were derived from multiple experiments (at least four experiments for each protein, each experiment consisting of triplicate determinations). The standard error for the K m values is shown. V max values (pmol of copper/mg of protein⅐min) are shown for each mutant, and standard errors are given.   NOVEMBER 4, 2005 • VOLUME 280 • NUMBER 44 two cysteines, one in TMS3 and one in the C-terminal tail, resulted in a significant impairment of copper transport. It seems likely that these mutations cause an uncompensated change in local structure with negative consequences to transport (see "Discussion"). The Effect of Tyr Substitutions-Two trans-membrane tyrosine residues (Tyr 83 and Tyr 147 in hCTR1) are highly conserved among the Ctr family members (Fig. 1B). A third tyrosine residue, Tyr 156 , is located in the putative extracellular loop between TMS2 and TMS3 and may potentially play a role in copper entry into the trans-membrane pathway. To investigate the role of these residues, Tyr 83 and Tyr 147 were mutated to either alanine or phenylalanine and stably expressed in Sf9 cells. Analysis of protein expression and copper uptake (Fig. 5) demonstrated that the K m values of hCTR1 for copper were unaffected in these mutations, although we did observe some effect on the rate of transport.

NT and CT truncations
Kinetic constants for copper uptake by hCTR1 N-terminal and C-terminal truncation constructs. Kinetic constants are presented as in Table II (for specific details  see Table II). Additionally, mutant K m and V max values are presented as ratios with respect to Wt hCTR1 kinetic constants (mutant/Wt). Also, plasma membrane mutant protein abundance for each mutant is shown as a ratio of mutant protein to Wt hCTR1 protein measured as described under "Experimental Procedures." This ratio is used to calculate the normalized V max value for each mutant by correcting the experimentally measured mutant V max values for differences in protein abundance. In each case, the mutant V max values were 40 -50% of Wt for the Tyr Ͼ Phe substitutions and 60 -70% of Wt for the Tyr to Ala mutants, suggesting that these Tyr residues are unlikely to be directly involved in copper transport. In contrast, the Y156A mutation had a dramatic effect on copper uptake (TABLE FIVE and Fig. 5). The V max value of this hCTR1 variant was only 20% of the V max of the Wt construct, the most significant decrease in the V max among all single and double mutants examined in this study, and the K m value for copper was only 2-fold greater than the K m value of Wt protein.

Construct
Mutations of Charged Residues in the Trans-membrane Domains Result in Gain of Transport Function-There are three charged residues in the trans-membrane segments of hCTR1, Glu 68 , Glu 84 , and His 139 . The occurrence of charged residues in the trans-membrane segments of cation transport proteins is often associated either with important structural or functional roles, because such residues are normally more stable in highly polar environments. We mutated the two glutamic acid residues to either a leucine, which has a nonpolar side chain, or to a glutamine, having an uncharged polar side chain. The mutant proteins were expressed equally well and were delivered to the plasma membrane (Fig. 6, inset). Unlike the mutants described above, these mutant proteins transported copper at higher rates than the Wt hCTR1. E68L and E68Q transported copper at about 1.5-fold higher rate than the Wt hCTR1. The K m value for the Glu 68 mutants is 2-3-fold higher than that of Wt hCTR1 (TABLE SIX). Similarly, the mutants E84L and E84Q also transported copper at higher rates than Wt hCTR1 (1.1-and 1.7-fold, respectively).
The most dramatic and unexpected effect was observed for the H139R mutant, which was designed to alter the size but preserve the charge at this position. In this case, the K m value of the mutant for copper was close to 0.15 mM, a ϳ16-fold decrease in apparent affinity compared with Wt hCTR1 (TABLE TWO). In addition, the transport rate was markedly increased with the V max value for the mutant about 7.5-fold higher than for the Wt transporter.
To examine the possible role of His 139 in copper transport in more detail, we generated a H139A mutant to replace His with a small and uncharged residue. The Ala mutant showed little effect on K m for copper (TABLE SIX), suggesting that the effect of Arg substitution on the K m could be because of the size of this residue. At the same time, the H139A had lower (only 40%) transport activity compared with Wt hCTR1 (TABLE SIX). Thus, the difference in the type of residue at position 139 (Arg compared with Ala) generated in total a 20-fold difference in the rate of transport, with the presence of the (likely) fixed positive charge being highly stimulatory (see also "Discussion").

DISCUSSION
Members of the CTR1 family have no apparent homology to known transporters, and their transport mechanism is currently unknown. In the present work we have carried out mutational and functional analysis of hCTR1 and have made some first steps toward understanding how hCTR1 may mediate copper entry into cells. We have characterized directly and quantitatively the transport activity of various hCTR1 mutants using radioisotopic copper uptake measurements, allowing analysis of hCTR1 function at a greatly improved level of sensitivity and resolution. This was possible because of the utilization of stably expressing insect cells, which provide a permanent and highly reproducible source of the transport protein in uncompromised and directly accessible cell membranes. The unexpected and important finding of this study is that the majority of single residue replacements produce subtle changes in the transport properties of hCTR1. Most of the potential copper-coordinating residues, when deleted by truncation or replaced by substitution, have only a minor effect on the K m value for transport. Therefore, any binding at these sites does not represent the rate-limiting transport step. Substitution of the intramembrane His 139 by Arg, however, causes a large change in both K m (16-fold) and V max (8-fold) values and implies that rate-limiting interactions for permeation are likely to take place at this location in the pore. hCTR1 has only three trans-membrane segments (although the involvement of other portions of the protein into P-loop-like structures cannot be excluded) and forms oligomers with the functional unit being most likely a trimer (18,23,24). Neither ATP hydrolysis nor ion gradients seem to be necessary for copper uptake, suggesting that hCTR1 is neither an ion pump nor a secondary transporter. Two fundamentally  In one mechanism, specific, essential sites accessible to the extracellular membrane surface would bind copper and, following a conformational change, would release copper at the intracellular surface. In a second mechanism, a permeation pathway or pore is formed by the protein, and copper, once having gained access to this pore, passes through the membrane. Such a permanent pore is not a feature of the first mechanism. Access to the pore can be regulated via a gate that "moves" in response to copper loading or in response to some other factor. It seems most likely that a permanent pore structure is formed by the oligomerization of hCTR1. It is possible that some of the changes we have made, for example deletions in the extracellular N-terminal domain (Met 69 ), that result in a dramatic loss of transport function may be the result of destabilization of the functionally essential oligomerization of hCTR1. The apparent absence of essential residues within the intramembrane portion of the protein and the marked stimulation in V max values caused by the His to Arg substitution at the inner surface suggest a model in which hCTR1 acts through the regulated access to a pore, with rate-limiting interactions occurring close to where copper ions exit from the pore.
Copper Entry into hCTR1-The presence of multiple Met and His residues in the extracellular N-terminal portion of the transporter (the total in a trimer is a remarkable 54!) makes this region an attractive candidate for performing the role of copper "collection" at the mouth of a pore. Previous studies using the yeast complementation assay demonstrated that the second methionine motif (Met 40 -Met 45 ) was important for copper uptake, whereas the first Met repeat was not (19). Our measurements are consistent with distinct roles of the N-terminal Met residues in the copper transport process.
The deletion of the first 34 residues, which removes all His and half of the Met residues, has no effect on the V max value of transport and shows only a 2-fold increase in the apparent K m value for copper. Clearly, this region was not essential for copper entry into the pathway. It was also not essential for the structural integrity of the transporter. However, in   the organism, the delivery of copper to hCTR1 presumably occurs via interactions between extracellular donor protein(s) (the identity of which remains to be resolved) and the extracellular domain of the transporter. Thus, the extracellular domain (with multiple His and Met residues) may play an essential physiological role in accepting copper ions from copper donors. This step is circumvented in the in vitro assay by direct addition of copper ions. Further deletion of the N-terminal segment (up to 53 amino acid residues) removes all Met and His residues from the extracellular N-terminal domain and has more dramatic consequences for hCTR1 function. The rate of transport decreases significantly (by 70%), with only about a 3-fold change in the K m value for copper. The marked decrease in the V max value could be due either to infrequent opening/limited access to the pore or to slower permeation through the pathway. As the permeation pathway is likely to be via a pore through the membrane, any effects of removal of extracellular Met residues are likely to be indirect. The most extensive truncation, Met 69 , removes the entire extracellular N-terminal domain and results in loss of copper transport.
A similar role in facilitating copper entry can be envisioned for Met 150 and Met 154 . These residues are located at the extracellular portion of TMS2 and are conserved among all Ctr1 molecules ( Fig. 1 and Fig. 7). It is possible that three pairs of Met 150 and Met 154 form the entrance into the copper-translocation pathway and facilitate copper entry either through their ability to weakly bind copper and/or by providing an appropriate steric fit for the pore (Fig. 8). It should be noted that the K m value of the double mutant remains the same as Wt hCTR1, so that it is unlikely that the binding of copper in this location represents a ratelimiting step in the permeation process. Although important, these residues are nevertheless not critical for copper transport. Uptake of copper, although markedly diminished, still occurs at 30% of Wt hCTR1 when both residues are replaced with isoleucines.
These latter data appear at odds with the previous reports (19), which showed that substitutions of Met 150 and Met 154 inactivated copper uptake. This apparent discrepancy is most likely due to differences in experimental conditions. Puig et al. (19) measured uptake at an early time point and with a single copper concentration (2 M). We have measured copper uptake for a period of 40 min at multiple copper concentrations. If a mutant CTR1 protein transports copper at a very slow rate, comparable with untransfected cells, then a single early measurement may not reveal a small but significant transport capacity.
Tyrosine Mutants in the Trans-membrane Domains-It is known that tyrosine residues are important structurally as hydrogen bond partners in proteins and can also participate in metal ion coordination (25,26). Most interestingly, the single-site mutation of Tyr 156 had a dramatic effect on the transport of copper with the Y156A mutant only transporting at a rate of about 20% of Wt hCTR1. This residue is probably located in the short loop between TMS2 and TMS3 and is very close to Met 154 at the extracellular boundary of the putative transport pore (see Fig. 8). Consequently, it seems likely that Tyr 156 plays a structural role in helping to place Met 154 appropriately.
Charged Residues in the Trans-membrane Segments-There are only three charged residues in the intramembrane regions of hCTR1. These residues are located in the first trans-membrane segment (Glu 68 and Glu 84 ) and the second trans-membrane segment (His 139 ). Most interestingly, mutations that remove the negative charges on the glutamates increase the rate of transport by hCTR1. Not only is the rate of transport increased, the apparent affinity of these mutations for copper decreases somewhat. It is tempting to speculate that side chain carboxyls may bind to or slow the rate of passage of copper ions through the transport pore as they pass through the membrane. However, the effects are not large, and a more extensive analysis of these sites would be needed to draw definitive conclusions. The same can be said of the M69I substitution in the first trans-membrane segment that results in a slightly increased value (150%) for the V max for copper uptake.
Copper Exit from the Pore-Very little is known about the immediate fate of copper as it enters the cytoplasm. It has been speculated that copper chaperones accept copper directly from Ctr1 transporters. Recent in vitro studies demonstrated that the C terminus of yeast yCtr1 can interact with Atx1 and that copper is transferred between these two proteins (21,22). Our results show that mutations of C-terminal Cys 189 or the CT184 deletion have only a slight effect on the rate of transport (10 and 30% decrease, respectively). It should be pointed out that such transfer to chaperones, if it occurs, is unlikely to be rate-limiting for copper uptake. This conclusion can be drawn from our observation that the H139R mutant has a much higher (8-fold) V max value for transport. It would be unlikely that a mutation within the membrane portion of the transporter could greatly stimulate uptake if the transport process itself was not rate-limiting.
The effects of replacements at position His 139 are the most dramatic and very interesting. When His 139 at the cytosolic end of TM2 is mutated to Arg there is a large increase in the V max value for transport (Ala replacement does not have a similar effect). Such gain-of-function mutations are unusual and therefore are potentially very informative. The specific effect of replacement with Arg could be explained by possible differences in protonation of His (partial) compared with Arg. It is likely that in the Wt hCTR1, all three His 139 residues (one from each monomer) may not be protonated simultaneously. Mutation to Arg places three positive charges in the pore; this may result in charge repulsions distorting the pore at a critical site, lessening interactions between the pore and the permeating copper ion and increasing the maximal transport rate. The details of the interactions and the involvement of His and Arg residues in the pathway are the subjects of ongoing studies. It is also worth noting that His 139 is not strictly conserved between yeast and mammalian Ctr1 proteins (mammalian Ctr1 proteins have a histidine at this position, and yeast Ctr proteins do not). However, there is an arginine residue four amino acids away (approximately one turn of an ␣-helix) from this histidine in both yCtr1 and yCtr3, which perhaps plays a similar role.
A Working Model for hCTR1-Bringing together the results from this study and others we can begin developing a model for hCTR1 functional architecture (Fig. 8). This model does not yet address the important issue of the interactions of the extramembrane segments of hCTR1 with proteins involved in the delivery to and removal from hCTR1 of copper ions under in vivo conditions. Further experiments are underway to start addressing these questions.
The permeation pathway is formed by the oligomeric association of three hCTR1 molecules. The pore is bounded at the outer membrane surface by a ring of Met residues formed from Met 150 and Met 154 in TMS2. Two glycine residues, Gly 167 and Gly 171 , on the opposed surface of TM3 have been shown previously to be important for function via their role in helix packing of the trans-membrane domains. The appropriate configuration of the Met residues can be influenced by the nearby Tyr 156 . Cys 161 is implicated as a potential structurally important residue given our present results with the double mutant C161S/C189S and the results from others (18).
At the inner membrane surface Glu 84 , in TMS1, and His 139 , in TMS2, are close to the exit site from the pore. It is at this latter location that interactions between copper ions and the protein form rate-limiting associations, as evidenced by the dramatic increase in transport rate and K m value for copper when His is replaced by Arg. We have observed previously (via proteolytic cleavage studies) that the binding of copper ions (presumably at the extracellular N-terminal domain) causes a conformational change in the intracellular loop between the first and second trans-membrane segments (20). It remains to be shown how such a conformational change contributes to facilitating copper uptake into cells via the proposed hCTR1 pore structure.