Undetectable Intracellular Free Copper: The Requirement of a Copper Chaperone for Superoxide Dismutase

See allHide authors and affiliations

Science  30 Apr 1999:
Vol. 284, Issue 5415, pp. 805-808
DOI: 10.1126/science.284.5415.805


The copper chaperone for the superoxide dismutase (CCS) gene is necessary for expression of an active, copper-bound form of superoxide dismutase (SOD1) in vivo in spite of the high affinity of SOD1 for copper (dissociation constant = 6 fM) and the high intracellular concentrations of both SOD1 (10 μM in yeast) and copper (70 μM in yeast). In vitro studies demonstrated that purified Cu(I)-yCCS protein is sufficient for direct copper activation of apo-ySOD1 but is necessary only when the concentration of free copper ions ([Cu]free) is strictly limited. Moreover, the physiological requirement for yCCS in vivo was readily bypassed by elevated copper concentrations and abrogation of intracellular copper-scavenging systems such as the metallothioneins. This metallochaperone protein activates the target enzyme through direct insertion of the copper cofactor and apparently functions to protect the metal ion from binding to intracellular copper scavengers. These results indicate that intracellular [Cu]free is limited to less than one free copper ion per cell and suggest that a pool of free copper ions is not used in physiological activation of metalloenzymes.

Metalloproteins such as the antioxidant enzyme Cu,Zn-superoxide dismutase (SOD1) (1,2) are generally assumed to acquire the essential cofactors by passive diffusion, although recent studies indicate that more complex mechanisms are involved. A network of copper-trafficking genes including ATX1 and LYS7 of Saccharomyces cerevisiae and their mammalian homologs are required for copper incorporation of proteins in vivo (3–7). The LYS7 gene, which encodes the yeast copper chaperone for superoxide dismutase (yCCS) (4), is essential for expression of the functional, copper-bound form of yeast SOD1 (ySOD1) in vivo. Possible roles for yCCS in activation of ySOD1 include insertion of copper, stabilization of protein-folding intermediates, enhancement of protein stability, or modulation of intracellular copper availability in concert with other cellular factors. We found that yCCS directly inserts copper into the target enzyme and that this metallochaperone is necessary only if the ambient free copper concentration is exceedingly low. These results lead to the conclusion that the intracellular milieu has an extraordinary overcapacity for chelation of copper. In contrast to the emerging picture of energy-dependent iron uptake (8), we anticipate that this chelation capacity leads to a substantial driving force for copper passage into the cell.

The requirement for yCCS in vivo was first examined by measuring the intracellular concentrations of copper, yCCS protein, and the ySOD1 target enzyme (9–12). In agreement with previous studies (4), we found that the absence of yCCS did not decrease the total number of ySOD1 molecules yet substantially reduced the ratio of the active, copper-bound ySOD1 to the inactive apo form of the enzyme (Table 1). In addition, the total number of copper atoms in these strains was similar regardless of the presence of yCCS or ySOD1 (Table 1), demonstrating that despite the loss of either abundant copper-binding protein, the number of copper ions per cell remains constant. Thus, there is no apparent role for yCCS in modulation of copper uptake or target enzyme stability.

Table 1

Amounts of copper and SOD1 per cell and resistance to copper toxicity in various yeast strains (9). Amounts of copper are reported in atoms per cell as determined by elemental analysis of whole cells (10). Total SOD1 and yCCS were determined by protein immunoblotting and are given as the number of molecules of monomers per cell (11). Active SOD1 is given as an upper limit of active SOD1 monomers per cell as determined by gel analysis of SOD1 activity (12). Mean inhibitory con- centration (MIC) is the concentration of copper sulfate required to inhibit yeast growth in a synthetic dextrose minimal medium (27) by 50%. All measurements are given as a full range of values obtained over two to four independent samples except in the case of the copper determinations, where the uncertainty is the SD of two independent experiments of four measurements each. ND, not determined; null indicates total absence of protein due to gene deletion. WT, wild type.

View this table:

To determine if this metallochaperone is sufficient for direct activation of the metal-depleted ySOD1 (apo-ySOD1), we expressed yCCS in Escherichia coli, purified it to homogeneity (Fig. 1A) (13), and incubated the copper-loaded form with purified samples of denatured apo-ySOD1 (14, 15). Yeast SOD1 enzyme was acid denatured and lyophilized after metal removal and was then treated with Cu(I)-yCCS or CU(I)-GSH copper donors (Fig. 1, B and C). The glutathione (GSH) complex of Cu(I), a compound previously invoked as a possible physiological Cu donor (16), was used as a control. After incubation with copper donors, the resulting ySOD1 activity was assessed by the standard kinetic assay with cytochrome c (17). Apo-ySOD1 was activated upon treatment with either Cu(I)-yCCS or Cu(I)-GSH. Copper in either the cuprous [Cu(CH3CN)4PF6] or cupric (CuSO4) oxidation state also activated apo-ySOD1 under similar conditions. The yCCS metallochaperone is thus unnecessary for in vitro ySOD1 activation if a sufficient pool of copper is accessible to the apo-enzyme.

Figure 1

In vitro analysis of yCCS metallochaperone function in activation of Cu,ZnSOD (ySOD1). (A) SDS-PAGE of the overexpression and purification of yCCS from E. coli. Lane 1, molecular size standards; lane 2, lysate from noninduced cells; lane 3, lysate from cells with yCCS protein induction; and lane 4, purified yCCS. (B) Cytochrome c–xanthine oxidase assay for SOD activity after the indicated Cu(I) sources were incubated with apo-ySOD1 in the presence or absence of the Cu(I) chelate, BCS. Activity is reported as units per milligram of apo-ySOD1 protein in each sample (1 U = 50% inhibition of the rate of cytochromec reduction) with residual activity from an apo-ySOD1 control subtracted. (C) Native PAGE of the in vitro activation mixtures assayed in (B) stained with NBT, riboflavin, andN,N,N′,N′-tetramethylethylenediamine for SOD activity. Amounts of ySOD1 protein applied to each lane are 40 and 20 ng of holo-ySOD1 in lanes 1 and 7, respectively, and 240 ng of apo-ySOD1 in each of lanes 2 to 6.

In contrast, the in vivo measurements indicated that ySOD1 is inactive in the absence of yCCS, even though total cellular copper remains at normal concentrations (Table 1). The apparent disparity between in vitro and in vivo metal incorporation of ySOD1 is readily reconciled if the “free” copper concentration ([Cu]free) in the cytoplasm is strictly limited. In this model, yCCS functions both to sequester copper away from metal-scavenging agents and to direct the cofactor into the enzyme. To test this, we decreased [Cu]free in the assay solutions by the addition of bathocuproine sulfonate (BCS), a chelating agent that reacts rapidly with and has high affinity for aqueous Cu(I) ions [dissociation constant (K d) ∼ 10−20 M] (18). With [Cu]free minimized (∼10−17 M) by the presence of 0.2 mM BCS, Cu(I)-yCCS still activated the apo-enzyme, whereas activation by Cu(I)-GSH was nearly abolished (Fig. 1B). Similarly, activation by CuSO4donor was restricted with the addition of an appropriate Cu(II) chelate, such as EDTA (K d = 1.6 × 10−19 M). Thus, in the presence of copper-scavenging reagents, Cu(I)-CCS can activate ySOD1, but Cu(I)-GSH and CuSO4 cannot.

Samples of these same reconstitution mixtures were also assayed on native polyacrylamide gels stained for SOD activity with nitroblue tetrazolium (NBT) (Fig. 1C). The resulting SOD activity bands confirmed the previous conclusion and established that the bulk superoxide disproportionation activity detected in the kinetic assays arose from ySOD1 enzyme rather than other adventitious reactions. Control samples of Cu-yCCS or Cu-GSH that lacked apo-yCCS also displayed no SOD activity bands comparable to holo-ySOD1.

The in vitro results indicate that yCCS alone is sufficient for activation of the target enzyme. Furthermore, the metal insertion event apparently proceeds with direct transfer of the copper ion from yCCS to the ySOD1 target enzyme. The inability of Cu(I)-GSH to activate ySOD1 in the presence of BCS (Fig. 1B) precludes an indirect transfer mechanism in which copper is first released from yCCS into solution and subsequently binds to apo-SOD. Copper ions that may dissociate from yCCS or GSH are rapidly sequestered by the BCS competitor (present in 20-fold excess over the copper donor molecule), resulting in a complex that is incapable of donating copper to ySOD1 apo-enzyme.

The biochemical activation data also suggest that the strict genetic requirement for yCCS will only be manifested when the cytoplasmic concentration of free copper is negligible. To test this proposition in vivo, we increased cytoplasmic concentrations of available copper by two means: exposure of cells to elevated copper concentrations in the growth medium ([Cu]medium) or elimination of cytoplasmic copper-sequestering proteins.

Elevation of [Cu]medium from submicromolar concentrations to 5 mM failed to activate ySOD1 in cells lacking the yCCS metallochaperone. As [Cu]medium was further increased to toxic amounts, however, the total number of copper atoms per cell increased by 100-fold and ySOD1 activity became apparent in native gels assays of cytoplasmic extracts (Fig. 2A). Thus, a key limitation to ySOD1 activity is the low availability of copper within the cell. Given that cells lacking the metallochaperone have similar amounts of copper and ySOD1 (Table 1) to wild-type cells, it is clear that the copper must be bound tightly at other sites that are inaccessible to apo-ySOD1. Candidates for such copper-sequestration sites include the yeast metallothionein isoforms encoded byCUP1 and CRS5 (19, 20).

Figure 2

In vivo analysis of copper effects on SOD1 activation in the absence of yCCS. The indicated yeast strains were grown to an early logarithmic stage in an enriched YPD medium (27) supplemented with various concentrations of copper. Lanes 1 and 2, no copper added; lanes 3 to 7, [Cu]medium = 0.2, 0.5, 1.0, 2.0, and 5.0 mM CuSO4, respectively; and lanes 8 and 9, 8.0 and 15 mM CuSO4, respectively, added to MT+ strain cultures (these concentrations prohibit growth of MTyeast). Crude protein extract (20 μg) (26) was analyzed by the gel assay for SOD1 activity (28). (A) Analysis of MT+ strains expressing CUP1 andCRS5. (B) Analysis of strains deleted for both MT genes. As the [Cu]medium is elevated, the total intracellular copper concentrations are observed to increase in parallel. For these strains, the initial number of copper atoms per cell is 1 × 104 (lane 2), but at the [Cu]medium where SOD1 activity is observed, the number of copper atoms per cell is 1 × 106 (MT+strain, lane 8) and 1 × 105 (MT strain, lane 5). The strains used here [MT+ CCS+, 19.3C; MT+ CCS, PS117; MTCCS+, PS115; and MT CCS, PS116 (9)] are different from those in Table 1 and show typical variations in the extent of copper accumulation.

The role of yeast metallothioneins in restricting intracellular [Cu]free was evaluated in yeast strains with deletedCUP1 and CRS5 genes (designated MT) that also lacked yCCS (9). In the absence of yCCS, ySOD1 activity in these strains was dependent on the dose of copper provided in the medium (Fig. 2B). ySOD1 activity became apparent at [Cu]medium = 0.5 mM, at which point cellular copper accumulation increased by 10-fold. This represents an order of magnitude less than the total cellular copper or [Cu]medium required to activate a comparable amount of ySOD1 in the yeast strain expressing MTs (Fig. 2A). These results demonstrate that yeast metallothioneins limit copper availability to ySOD1. Because additional copper must still be added to the medium to observe SOD activity, MTs must provide only part of the copper-chelation capacity in the cell. Furthermore, the fact that ySOD1 can be activated by copper in vivo in the absence of yCCS (Fig. 2) argues against an essential role for yCCS in activation of apo-ySOD1 through protein-folding pathways. We conclude that the yCCS molecule specifically delivers ∼15% of the total cellular copper pool to saturate the ySOD1 pool (Table 1) in spite of an abundance of intracellular scavengers that maintain [Cu]free at very low values.

The data in Table 1 are a reliable gauge for the number of relevant molecules per cell and provide a starting point for evaluating copper availability in the cell. Using these data and a thermodynamic model (21), we estimate the [Cu]free in an unstressed cell to be less than in 10−18 M. For comparison, one free copper atom in the same volume (10−14liters) corresponds to a [Cu]free of 10−10M. It is unclear whether the intracellular copper-ySOD1 system is at equilibrium; however, the rate of copper binding to ySOD1 in vitro is quite rapid (22). Using a kinetic model (22), we arrived at the same conclusion: There is less than one free copper atom per cell under normal growth conditions. Given the high intracellular capacity for copper binding, we anticipate that kinetic aspects of specific metallochaperone interactions will control the extent of copper transfer to target versus nontarget molecules.

The results indicate that intracellular copper availability is extraordinarily restricted. The extensive chelation capacity allows too few free copper ions to become available to saturate the 60,000 molecules of SOD1 within the typical life-span of a yeast cell. This provides a context for understanding the driving force for metal uptake in general (8) but more specifically underscores the physiological requirement for metallochaperones: Copper-dependent enzymes require accessory factors to compete with chelators and processes that sequester essentially all intracellular free copper. In this way, the metallochaperones protect copper ions. It has been suggested that copper chaperones protect cells against copper toxicity (3, 5). We do observe some protective effects of yCCS with respect to elevated [Cu]medium (MIC values, Table 1), although this protection likely results from insertion of metal into ySOD1 (itself a cellular buffer for copper) rather than from sequestration by yCCS (23). The central function of this metallochaperone is to directly insert the cofactor into the target enzyme, thus converting the latter from an inactive to an active state. In this sense, the function of the yCCS protein is analogous to that of the molecular chaperones, although it is unclear at this time whether yCCS also increases the efficiency of ySOD1 folding. These issues may be important in understanding the pathological features of the large number of known mutations in human SOD1 that lead to amyotrophic lateral sclerosis (24,25).

Note added in proof: Since submission of this paper, Lyonset al. (31) have proposed that in the absence of yCCS, intracellular SOD1 exists with one zinc ion bound per SOD dimer. This is compatible with our results, which do not address whether zinc is bound by SOD1 before or after copper insertion by yCCS.

  • * To whom correspondence should be addressed. E-mail: vculotta{at}; t-ohalloran{at}


View Abstract

Navigate This Article