Copper Activation of Superoxide Dismutase 1 (SOD1) in Vivo

Insertion of copper into superoxide dismutase 1 (SOD1) in vivo requires the copper chaperone for SOD1 (CCS). CCS encompasses three protein domains: copper binding Domains I and III at the amino and carboxyl termini, and a central Domain II homologous to SOD1. Using a yeast interaction mating system, yeast CCS was seen to physically interact with SOD1, and this interaction required sequences at the predicted dimer interface of CCS Domain II. Interactions with SOD1 also required sequences of Domain III, but not Domain I. Mutations were introduced at the dimer interface of yeast SOD1, and the corresponding mutant failed to interact with CCS. When loaded with copper independent of CCS, this mutant SOD1 exhibited superoxide scavenging activity, but was normally inactive in vivo because CCS failed to recognize the enzyme. Activation of SOD1 by CCS was also examined using an in vivo assay for copper incorporation into SOD1. Yeast CCS was observed to insert copper into a pre-existing pool of apoSOD1 without the need for new SOD1 synthesis or for protein unfolding by the major SSA cytosolic heat shock proteins. Our data are consistent with a model in which prefolded dimers of apoSOD1 serve as substrate for the CCS copper chaperone.

Eukaryotic cells employ a family of copper-carrying proteins entitled "metallochaperones" that escort the metal ion to distinct locations within the cell (reviewed in Ref. 1). Delivery of copper to the secretory pathway involves the action of the cytosolic metallochaperone ATX1, also known as ATOX1 or HAH1 (2)(3)(4)(5). In a separate pathway, COX17 facilitates the delivery of copper to the mitochondria where it is ultimately incorporated into cytochrome oxidase (6 -9). In addition, copper delivery and incorporation into copper-containing superoxide dismutase 1 (SOD1) 1 is mediated by the cytosolic copper carrier, CCS (copper chaperone for SOD1) (10,11).
The SOD1 target of the CCS metallochaperone is a homodimeric copper-and zinc-requiring enzyme that acts to disproportionate deleterious superoxide anions to hydrogen peroxide and oxygen in a reaction catalyzed by the redox cycling of the bound copper (12). Each monomer of SOD1 contains a single copper atom buried well within the active site of the molecule. SOD1 binds copper with high affinity in vitro; however, the yeast enzyme is apo for copper in Saccharomyces cerevisiae cells lacking the copper chaperone (10,13). Additionally, mammalian SOD1 is largely inactive in a null mouse model for murine CCS (14). Based on studies in yeast, there appears to be an absence of free copper ions available to SOD1 due to a potent copper-chelating capacity of the cell (15,16). Hence, metallochaperones such as CCS ensure activation of metalloenzymes under conditions where the free metal is not readily available.
Three functionally distinct polypeptide domains are evident in CCS. The amino-terminal region (Domain I) harbors the same MXCXXC copper binding site typical of the ATX1 copper chaperone, and the two molecules adopt similar tertiary folds as revealed by x-ray crystallography (17,18). However, the ATX1-like region of CCS is not obligatory for activity and only seems necessary under conditions of extreme copper limitation (19). The central Domain II of CCS bears remarkable homology to SOD1, yet lacks the SOD1-copper site as demonstrated in three-dimensional structures resolved for human and yeast CCS (18,20,21). It has been postulated that Domain II plays a critical role in recognition of the SOD1 target (18 -20, 22). Finally, at the carboxyl terminus of CCS, a short Domain III peptide bears a CXC copper binding site that may interact with Domain I and is crucial for inserting copper into SOD1, in vivo (19,23).
A number of questions remain regarding the mechanism of CCS action. For example, how does this metallochaperone specifically recognize its SOD1 partner? Although Domain II of CCS is predicted to form a complex with SOD1, the role of the copper binding Domains I and III in this process have not been explored. A second question regards the nature of the SOD1 polypeptide that serves as a substrate for CCS. Can CCS insert the metal into pre-assembled SOD1 dimers, or is the metal inserted co-translationally with SOD1 polypeptide synthesis? To begin to address these issues, we have utilized a yeast model system to examine the physical interactions between CCS and SOD1 and to monitor the copper insertion process in vivo.
Plasmids-Plasmids for the yeast interaction mating analysis were obtained from Roger Brent. pSH18-34 contains the bacterial lacZ reporter gene driven by the lexA promoter. Plasmid pJG4-5 (prey) was used to create fusions of CCS or SOD1 to the B-42 transcriptional activation domain tagged with a hemagglutinin (HA) epitope, and subcloning with pEG202 (bait) resulted in similar fusions to the LexA DNA binding domain. In both cases, protein fusions were introduced at the amino terminus of CCS or SOD1. CCS and SOD1 coding sequences from human or yeast were amplified using upstream primers that introduced an EcoRI site adjacent to the start codon (gaattcATG) and downstream primers that flanked the corresponding stop codons. The polymerase chain reaction products were cloned into the TA-TOPO vector (Invitrogen) per the manufacturer's instructions, mobilized by EcoRI digestion (yCCS required a partial digestion due to an internal EcoRI site) and inserted at the EcoRI sites of pJG4-5 and pEG202. To construct the yCCS II, III-prey expression plasmid, an internal EcoRI site was adjusted to an in-frame position using the QuikChange site-directed mutagenesis kit (Stratagene) per the manufacturer's instructions. Domain I was then removed by EcoRI digestion followed by re-ligation. Plasmids for expression of yCCS I, II-prey and yCCS II-prey fusions were prepared through introduction of an opal stop codon at yCCS residue ϩ76. Site-directed mutagenesis was also used to derive plasmids for the expression of yCCS C229,231S-prey, yCCS K136E,G137E-prey, yCCS II, III C229,231S-prey and ySOD1 FG50,51EE-prey.
Biochemical Analysis-For interaction mating experiments, the prey plasmid was transformed into strain RFY206 (MATa) and the bait and pSH18-34 reporter plasmids were transformed into yeast strain EGY48 (MAT␣), also harboring a lexA-LEU2 integrant. Yeast were mated overnight on YPD plates and then replica-plated onto either synthetic dextrose-based medium supplemented with 5-bromo-4-chloro-3-indolyl ␤-D-galactopyranoside (X-gal; for colorimetric scoring of the lacZ reporter) or onto medium lacking leucine (for scoring expression of the LEU2 reporter by cell growth). When needed, 1 mM bathocuproinedisulfonic acid (BCS; Aldrich) was added to the growth medium.
Immunoblot analysis was conducted essentially as described (25). Expression of the interaction mating protein fusions was monitored using an anti-LexA antibody (1:20,000) (Invitrogen) and goat antirabbit horseradish peroxidase-linked secondary antibody (1:10,000; Bio-Rad) for bait molecules, and anti-HA (1:3000; Boehringer) and anti-mouse horseradish peroxidase-linked secondary antibody (1:10,000; Bio-Rad) for analysis of prey proteins. Expression of native yCCS and ySOD1 was detected using anti-CCS and anti-SOD1 IgG as described previously (25). SOD1 activity was visualized by an in situ gel assay through staining with nitro blue tetrazolium (NBT) (25). For immunoblot analysis of native gels, 10 g of protein were applied directly to a precast 14% polyacrylamide gel (Novex) without boiling. After electrophoresis, the gel was soaked in SDS gel running buffer for 1 h, followed by immunoblot detection of ySOD1 as described previously (25).
For kinetic studies of copper incorporation in SOD1 in vivo, cells were first grown to an approximate A 600 of 1.5 in YPD medium supplemented with 250 M BCS. 50 M CuSO 4 was then added, and 50-ml samples of cells were quickly harvested at various time points for preparation of cell lysates and analysis of SOD1 activity. Where indicated, 100 g/ml cycloheximide was added for 10 min prior to addition of copper to arrest protein synthesis. In the case of temperature-sensitive mutants, cultures were pregrown at a permissive temperature (25°C) then treated for 1 h at the non-permissive temperature (37°C) before addition of copper.

Interaction Mating Assay for Monitoring CCS-SOD1 Interactions in Vivo-To study the physical interactions between CCS
and SOD1 in vivo, we employed a yeast interaction mating system. Chimeric proteins were constructed in which the open reading frames of CCS or SOD1 (human or yeast) were fused to either the DNA binding domain of LexA ("Bait") or to the potent B42-HA transcriptional activation domain ("Prey"). Haploid strains expressing the corresponding bait or prey molecules were mated, and the resultant diploids harboring a lexA-LEU2 reporter were tested for bait-prey interactions by a leucineindependent growth test. As seen in Fig. 1A, the interaction mating assay revealed in vivo interactions between yeast CCS and SOD1 from both yeast and humans. Consistent with results obtained by Gitlin and colleagues (22), human CCS positively interacts with human SOD1. We additionally obtained evidence for a hCCS⅐ySOD1 complex using this interaction mating system (Fig. 1A). The metallochaperones themselves appeared to oligomerize as well. Human CCS exhibited interaction with itself and with yeast CCS, however, we failed to obtain evidence for yCCS⅐yCCS dimers (Fig. 1A).
CCS Sequences That Facilitate Interactions with SOD1-In x-ray crystallographic analyses of human and yeast CCS, Domain II was seen to form dimers, with a subunit interface similar to that of SOD1 homodimers (18,20,21). It has been shown by Bertini and colleagues (32,33) that the dimeric interface of human SOD1 can be disrupted by substituting Phe 50 and Gly 51 with Glu. Because these sequences are somewhat conserved in Domain II of human (Phe 133 Gly 134 ) and yeast (Lys 136 Gly 137 ) CCS molecules ( Fig. 2A), the analogous substitutions were introduced in the copper chaperones to test the role of Domain II sequences in CCS activity. The resultant K136E,G137E mutant of yCCS was stably expressed in yeast cells, but was defective for charging SOD1 with copper as seen in a gel assay for SOD1 activity (Fig. 2B)  weak activity represented only Յ5% of that seen with wild type yCCS, as determined by serial cell dilutions (Fig. 2C). Consistent with this loss in activity, the K136E,G137E variant of yeast CCS failed to interact with SOD1, as seen in the interaction mating test (Fig. 2D). Therefore, sequences of the predicted dimer interface region of CCS Domain II appear important for formation of a docked complex with SOD1 and for transfer of copper.
We next tested whether Domain II of yeast CCS (amino acid sequences 79 -249) is by itself sufficient to direct interactions with SOD1. As seen in Fig. 2D, the yCCS Domain II prey fusion failed to interact with the ySOD1 bait, and the same results were obtained in the converse experiment where Domain II was introduced as bait (not shown). This lack of interaction does not reflect low protein levels, because the Domain II fusions were stably expressed (not shown). Furthermore, the failure to interact is not due to interference by the aminoterminal fusion to Domain II; other chimeras harboring precisely the same fusion to Domain II gave positive signals in the interaction mating test (see ahead, Fig. 3). Thus, the SOD1-like Domain II of CCS is necessary, but not sufficient, to direct physical interactions with SOD1.
A number of truncations and point mutations were introduced into full-length yCCS to define additional sections of the protein that facilitate interactions with SOD1. Interaction mating analysis of protein interaction was determined both by the leucine-independent growth test (lexA-LEU2 as reporter; Fig.  3A) and by ␤-galactosidase measurements (lexA-lacZ as reporter; Fig. 3B). As seen in Fig. 3, a fusion protein spanning yCCS Domains I and II failed to interact with SOD1, even though the protein was stably expressed (not shown). By comparison, a fusion spanning Domains II and III did physically interact with its target (Fig. 3A); however, the interaction appeared somewhat weaker than the association with fulllength yCCS, as monitored by the diminished ␤-galactosidase signal (Fig. 3B). Hence, both Domains II and III of CCS, but not Domain I, were necessary for interactions with SOD1. Domain III contains a copper binding CXC motif that is crucial for yCCS activity in vivo (19). Interestingly, this motif was not required for physical interactions between yCCS and ySOD1. Substituting cysteines 229 and 231 with serine in both the full-length and the fusion spanning Domains II and III did not alter interactions with SOD1 (Fig. 3). Domain III of CCS therefore appears to facilitate interactions with SOD1 by methods that do not involve the crucial copper binding site.
To test further the role of copper in modulating CCS/SOD1 interactions, interaction mating analyses were conducted under conditions where medium copper was limited by treatment with BCS, a specific Cu ϩ chelator. This treatment indeed restricts copper incorporation into SOD1 (see Fig. 6). As revealed in Fig. 4, lowering copper availability did not abrogate interactions between yCCS fusions and ySOD1 (Fig. 4, A and B) nor between human and yeast SOD1 molecules (Fig. 4C). If anything, copper depletion appeared to have increased the strength of contacts between fusions spanning yCCS Domains II and III and SOD1 (Fig. 4B).
SOD1 Sequences Needed for Activation by CCS-In one model that has been proposed for CCS, a monomer of SOD1 forms a heterodimer with CCS Domain II in a manner that resembles the homodimer complex of SOD1 (18 -20, 34). We therefore addressed whether the dimer interface region of

FIG. 2. Mutations within Domain II of yeast CCS.
A, an amino acid sequence alignment of CCS and SOD1 sequences that surround the dimer interface mutation of human SOD1 (FG50,51EE) known to create monomeric SOD1 (32). Arrows demarcate amino acid changes. B, the lys7⌬ strain lacking CCS (SY2950) was transformed with plasmids expressing either wild type yeast CCS (pHAL7-413) or the K136E,G137E variant (pPS035), or with empty vector (pRS413) (Vec). Cell lysates were prepared, and 10 g of crude cell protein was subjected to either denaturing gel electrophoresis and Western blot analysis using an anti-yCCS antibody (top) or to non-denaturing gel electrophoresis and NBT staining for SOD activity (bottom) (25). An arrow indicates the position of ySOD1 activity. C, the transformants as described in B were tested for complementation of the aerobic lysine auxotrophy of lys7⌬ strains. Both lys7⌬ and sod1⌬ strains cannot grow aerobically on medium lacking lysine due to oxidative damage to components of the lysine biosynthetic pathway (10,11,42). 4 ϫ 10 5 and 2 ϫ 10 4 cells were spotted onto minimal media lacking lysine and then grown either aerobically or anaerobically for 4 days. D, interaction mating analysis of ySOD1-bait interactions with prey fusions to wild type yCCS, the K136E,G137E mutant, or to yCCS sequences spanning only Domain II. Analysis of interactions was analyzed by the leucineindependent growth test as in Fig. 1A.

FIG. 3. A role for CCS Domain III in mediating interactions with SOD1.
Interactions between ySOD1 prey fusions and the indicated yCCS bait fusions were tested by interaction mating using the leucine-independent growth test (A) or by a ␤-galactosidase assay as described under "Experimental Procedures" (B). Control ϭ pRFHMI. SOD1 is needed for interactions with CCS in vivo. In the experiment of Fig. 5, we analyzed a FG50,51EE mutant of ySOD1, which is analogous to the monomeric mutant of human SOD1 ( Fig. 2A). The yeast SOD1 mutant is stably expressed in yeast cells, although migrates aberrantly on denaturing gels (Fig. 5A, top), as has been seen with a number of SOD1 mutants (35)(36)(37). This FG50,51EE mutant of ySOD1 normally exhibits no detectable superoxide scavenging activity when expressed in yeast (Fig. 5A, bottom) and is also incapable of complementing the aerobic growth defect of a yeast sod1⌬ mutant (Fig. 5B). As expected, this protein failed to show any physical interaction with itself or with wild type SOD1 (Fig.  5C). The FG50,51EE variant of ySOD1 also failed to interact with yCCS in the mating interaction assay (data not shown).
The complete absence of superoxide scavenging activity associated with FG50,51EE ySOD1 could result either from loss of recognition by yCCS or from conformational changes that cripple the SOD1 active site. Indeed, the corresponding FG50,51EE mutant of human SOD1 has been shown to possess only 10% of wild type activity (32). We therefore attempted to activate the FG50,51EE mutant of ySOD1 in vivo independent of the CCS metallochaperone. We have previously shown that SOD1 can be charged with copper independent of CCS by growing yeast cells in the presence of exceedingly high concentrations of copper (15). As seen in Fig. 5D, lane 3, the FG50,51EE SOD1 mutant exhibited activity when expressed in cells treated with elevated copper, although the activity was reduced in comparison to wild type SOD1 (lane 1). The position of mutant ySOD1 (presumably monomeric) on this non-denaturing gel was confirmed through use of an anti-ySOD1 antibody (Fig. 5D, lane 4). Copper activation of FG50,51EE ySOD1 in vivo was indeed independent of CCS, because activity did not change in a lys7⌬ strain lacking the copper chaperone (Fig. 5D,  lane 6). Therefore, yeast FG50,51EE SOD1 is normally devoid of activity in vivo, because CCS fails to recognize this enzyme.
Monitoring Copper Incorporation in Vivo-To examine further the requirements for CCS activation of SOD1, we developed an in vivo assay for monitoring the time course of copper incorporation into yeast SOD1. A pool of apoSOD1 can be generated in yeast cells by limiting copper uptake through the addition of the copper chelator BCS to the growth medium (Fig.  6A, lane 2). This pool of apoSOD1 was readily charged with copper following the addition of copper salts back to the media (Fig. 6A, lane 3). Reconstitution of SOD1 in this manner was quite rapid and began within 5 min of copper addition to the growth medium (Fig. 6, B and C; also see Fig. 7). Furthermore, this in vivo metallation of SOD1 was strictly metallochaperonedependent, because, in the lys7⌬ strain lacking CCS, there was no detectable activity even following 3 h of copper supplementation (Fig. 6A, lane 6).
The in vivo metallation assay for SOD1 was used to address the possible role of heat shock proteins (HSPs) in CCS activation of SOD1. Wild type SOD1 is a strong homodimer, even as an apo protein (21,38), and formation of a putative CCS-SOD1 heterodimer might be facilitated through protein unfolding by HSPs. In yeast, the SSA collection of genes encodes soluble HSP70 molecules (28). Specifically a yeast mutant lacking all four SSA genes (SSA1, SSA2, SSA3, and SSA4) is inviable at FIG. 4. Interaction mating analysis of CCS⅐SOD1 interactions under copper limiting conditions. The mating interaction test was performed as in Fig. 1A, except, where indicated, medium was supplemented with 1 mM of the copper-specific chelator, BCS. A, interactions were monitored between yCCS prey fusions and bait fusions to human or yeast SOD1 and CCS molecules. Control ϭ pRFHMI. B, interactions tested between ySOD1 prey fusions and bait fusions to wild type or mutant yCCS molecules. C, test for interactions between ySOD1 prey molecules and bait fusions to human or yeast SOD1, as indicated.

FIG. 5. A dimer interface mutant of yeast SOD1.
A, the sod1⌬ strain VC107 was transformed with either pRS426 (Vec), pPS008 expressing wild type yeast SOD1 (WT), or with pPS069 expressing the yeast SOD1 FG50,51EE mutant. Cell lysates were prepared and were subjected to either denaturing gel electrophoresis and Western blot analysis using an anti-ySOD1 antibody (top) or to non-denaturing gel electrophoresis and NBT staining for SOD activity (bottom) (25). B, transformants as in A were tested for complementation of the aerobic lysine auxotrophy of sod1⌬ mutants as described in Fig. 2C. C, a yeast SOD1 prey fusion was tested in the mating interaction assay with bait fusions to either wild type ySOD1 or the FG50,51EE mutant as in Fig.  1A. D, the sod1⌬ strain VC107 (lanes 1-4) or the sod1⌬ lys7⌬ strain VC279 (lanes 5-7) expressing either wild type yeast SOD1 (lane 1) or the FG50,51EE variant (lanes 2-7) were grown in enriched medium that was untreated (lanes 1, 2, and 5) or treated with 10 mM CuSO 4 for 4 h (lanes 3, 4, 6, and 7). Crude cell protein was subjected to nondenaturing gel electrophoresis, and the gels were either stained for SOD1 activity (lanes 1-3, 5, and 6) or soaked in SDS, followed by immunoblot analysis for ySOD1 expression (lanes 4 and 7). Arrows mark positions of wild type (WT) ySOD1 or the FG50,51EE variant. all temperatures, presumably due to the requirement of the encoded HSPs in the proper folding of cytosolic proteins (39). Additionally, S. cerevisiae expresses HSP104, which helps to reactivate recalcitrant proteins that have been denatured and or have aggregated (27,40). To determine if these various HSPs are critical for the unfolding or stabilization of SOD1 during copper insertion, the corresponding yeast mutants were tested for metallation of SOD1 in vivo. As seen in Fig. 6B, the mutant strains lacking HSP104 or a combination of SSA1, -2, and -4 exhibited rapid metallation of SOD1 upon addition of copper to the growth medium. We additionally tested the temperaturesensitive ssa1 ts ssa2⌬ ssa3⌬ ssa4⌬ strain. For this study, strains expressing apoSOD1 were first cultured at the permissive temperature (25°C), followed by a shift to the non-permis-sive temperature (37°C) for 1 h to inactivate SSA1; copper was then added to the growth medium. As seen in Fig. 6C, these cells lacking all four functional SSA HSPs exhibited normal metallation of SOD1 that was complete within 5 min. These studies demonstrate that yeast CCS is capable of inserting copper into SOD without the assistance of the major SSA cytosolic HSPs. However, these studies do not exclude the possibility that other unknown ancillary proteins are needed to assemble holoSOD1 in vivo.
Post-translational Insertion of Copper into Prefolded SOD1-Based on the strong association of SOD1 as a homodimer, it has been suggested that CCS inserts the metal co-translationally or immediately following new SOD1 synthesis, prior to the assembly of a fully folded SOD1 homodimer (21). To address this possibility, the in vivo metallation experiment was carried out in the presence of the protein synthesis inhibitor cycloheximide under conditions known to block synthesis of yeast polypeptides (41). Over the course of treatment with cycloheximide, the polypeptide levels of SOD1 and CCS did not vary largely (Fig.  7, bottom), because these proteins are quite stable. By analysis of SOD1 activity (Fig. 7, top), there was no obvious change in the kinetics of SOD1 metallation in the presence versus absence of cycloheximide. These data demonstrate that CCS can insert copper post-translationally into SOD1 in vivo and suggest that pre-existing apo dimers of SOD1 serve as the target for CCS. DISCUSSION The method by which CCS and SOD1 form a docked complex and then commence cofactor exchange is poorly understood. We have begun to address this issue in vivo employing assays for monitoring physical interactions between SOD1 and CCS and the time course of copper incorporation into SOD1.
Using the yeast interaction mating system, we obtained evidence for a docked complex between CCS and SOD1 in vivo and find that this physical interaction requires two distinct structural features of yCCS: sequences at the predicted dimer interface region of Domain II and specific sequences within CCS Domain III. CCS recognition of SOD1 also appears to require the dimeric interface region of SOD1, because a "monomeric" mutant of ySOD1 (FG50,51EE) failed to be activated by yCCS. The requirement for the dimer interface regions of both SOD1 and CCS are in agreement with structural predictions made by Rosenzweig on a heterodimeric complex of CCS and SOD1 (18,20). However, we cannot exclude the possibility that these dimer interface sequences are only employed in forming CCS⅐CCS and SOD1⅐SOD1 homodimers, which in turn assemble into higher order complexes, as proposed by Hart and colleagues (21). In any case, these studies on dimeric interface mutants have provided us with an intriguing variant of SOD1. The FG50,51EE mutant of yeast SOD1 was capable of scavenging superoxide but failed to be recognized by CCS, and as such, represents a useful tool for genetically dissecting the CCS-dependent and CCS-independent mechanisms of activating SOD1 in vivo.
In addition to Domain II, CCS interactions with SOD1 were facilitated by sequences of Domain III and surprisingly, these sequences did not include the crucial CXC copper site. The entire region spanning Domain III is highly conserved among all organisms yet studied (19), and any number of these residues may be involved in formation of docked complexes with SOD1. Because the copper site of SOD1 is well buried, Domain III might extend into the active site of SOD1 as a prequel to copper transfer, thereby stabilizing interactions with SOD1.
The interaction mating analyses also proved effective for monitoring in vivo oligomerization of the copper chaperones. Human CCS showed evidence of dimerization with itself and with yeast CCS; however, the yeast version of CCS showed no evidence of homodimer formation in this in vivo assay. This apparent lack of dimerization may reflect sequences of Domain II, because this region in yeast CCS exhibits far less homology to SOD1 than its human CCS counterpart (20,25). Although yeast CCS showed no evidence of dimerization in vivo, O'Halloran and colleagues (19) have demonstrated that the purified protein dimerizes in vitro but only in the copper-loaded form. This would suggest that yeast CCS largely exists in vivo as the apo form and only transiently binds copper when activating SOD1. Binding of copper to CCS is not a prerequisite for docking with SOD1. We show here that, under copper starvation conditions, when SOD1 cannot be charged with copper, yCCS and ySOD1 still form a complex as monitored in the interaction mating assay. Furthermore, Giltin and coworkers (22) have shown that human CCS still interacts with human SOD1 when copper is depleted. The formation of this apparent futile complex may facilitate the rapid activation of SOD1 once copper becomes available.
Our studies strongly indicate that, in vivo, CCS inserts copper into a pre-existing pool of apoSOD1 dimers. The apo form of SOD1 is known to form a strong homodimer in vitro (38), and our interaction mating studies under copper depletion conditions indicate that apoSOD1 also dimerizes in vivo. We show here that a cellular pool of apoSOD1 is readily charged with copper by CCS and that this reaction occurs within minutes in the absence of new protein synthesis. If CCS activates SOD1 through formation of a transient heterodimer complex, then some mechanism must exist to first disassemble the tight SOD1 homodimers. As shown herein, the major cytosolic SSA HSPs are not required for this process. It is therefore possible that CCS itself helps to disassemble SOD1 homodimers, somewhat analogous to the action of molecular chaperones. In an alternative model (21), the SOD1 homodimer may remain intact during copper insertion by CCS. A more complete understanding of the nature of the docked CCS⅐SOD1 complex (heterodimer versus heterotetramer) awaits detailed biophysical analysis of the protein composite.