Introduction

Copper is imported into prokaryotic cells by CPx-type ATPases. CPx-type ATPases have the transmembrane characteristics typical of P-type ATPases involved in the translocation of many ions. A conserved Cys-Pro-X (X = C or H) sequence within the transmembrane channel and a variable number of distinct amino-terminal domains define the CPx classification (1). The cytoplasmic subdomains of the CPx-ATPases have a MxCxxC or M/HxxMDHS/GxM metal-binding site (x = any amino acid). Intracellular copper is utilized in the activation of enzymes, such as cytochrome-c oxidase, superoxide dismutase, and lysyl oxidase. Copper also has the potential to cause cellular damage because of its redox properties. To overcome this dichotomy, the cell regulates copper levels and prevents toxicity with overlapping mechanisms: sequestration, export, and inhibition of entry.

Protection by sequestration can have numerous modes of action. Metals are secluded by binding to high-affinity peptides, such as metallothionein (2), or can be stored in vesicles with import facilitated by ATPase pumps [e.g., the Menkes and Wilson proteins (3)]. Sequestration mechanisms result in an intracellular environment rich in chelating agents that limit the concentration of potentially toxic metals. In fact, studies on baker's yeast Saccharomyces cerevisiae have shown that cells contain very low levels of available free copper (4). In spite of the chelating environment, essential copper reaches its target by utilizing a mechanism that both sequesters and directs the copper to copper-dependent proteins. This mechanism involves a class of proteins termed the copper chaperones.

Copper chaperones have been identified for the delivery of copper to copper-regulated DNA-binding proteins from bacteria, cytochrome-c oxidase from yeast, superoxide dismutase from yeast and humans, and Cu(I)-ATPases in yeast and humans (5-7). The utilization of homologous metallochaperones across such a diverse range of organisms suggests that a common method of copper trafficking has evolved. The copper chaperone, CopZ, has a central role in copper routing in the gram-positive bacterium Enterococcus hirae (8). The E. hirae copper homeostasis gene products are translated from the cop operon, which encodes four genes copZ-copY-copA-copB. The proteins expressed from the operon are the copper chaperone, a copper-regulated repressor CopY, and two ATPase pumps, CopA and CopB (9). The current understanding of the cop system is that cellular copper levels are controlled by the import and export pumps (CopA and CopB) whose expression is regulated by CopY (10).

Fig. 1. The solution structure of CopZ traces the same global PaPPaP fold found in the copper chaperone Atxl and the regulatory domain of phenylalanine hydroxylase. The two a-helices lay diagonally across four P-strands. CopZ and Atxl share the same metal-binding loop (MxCxxC) between the first P-strand and the a-helix. (PDB accession nos. 1CPZ, 1CC8, 1PHZ).

Fig. 1. The solution structure of CopZ traces the same global PaPPaP fold found in the copper chaperone Atxl and the regulatory domain of phenylalanine hydroxylase. The two a-helices lay diagonally across four P-strands. CopZ and Atxl share the same metal-binding loop (MxCxxC) between the first P-strand and the a-helix. (PDB accession nos. 1CPZ, 1CC8, 1PHZ).

Copper binds to CopZ in the cytoplasm and is subsequently transferred to CopY. The binding of copper to CopY leads to release of the repressor allowing expression of the operon to proceed.

2. COPZ: A COPPER CHAPERONE

CopZ has sequence homology to proteins involved in the homeostasis of numerous metals. These include the copper chaperone, Atxl (S. cerevisiae), the mercury chaperone, MerP (Shigella flexneri), a domain in the copper chaperone for superoxide dismutase, CCS (S. cerevisiae and Homo sapiens) and as the N-terminal domains of the CPx-ATPases: Menkes disease (MNK), Wilson's disease (WND), and CopA (H. sapiens and E. hirae). The solution structure of CopZ reveals a conserved PaPPaP global fold and an exposed flexible loop that contains the MxCxxC metal-binding motif (Fig. 1) (11). The two a-helices lie diagonally across four P-strands forming an "open-faced P-sand-wich" (12). This simple global fold is utilized for many different functions. The PaPPaP fold is found as a domain in phenylalanine hydroxylase (13) (Fig. 1), procarboxypeptidase (14), and ferre-doxin (15). Solution structures of MerP (16) and MNKr4 (17) and the crystal structures of Atx1 (18) and CCS (19) have the same PaPPaP fold and reinforce the importance of the fold in the metallochaperone family. The ubiquitous nature of this global fold contrasts to the specificity of the copper chaperones for their target.

Spectral analysis of CopZ has revealed that the protein binds copper and silver (20,21). Copper-binding stoichiometry of apoCopZ can be determined by monitoring the amplitude of the metal-ligand charge-transfer band (MLCT). The untraviolet (UV)-visible absorption spectra of Cu(I)CopZ displays the characteristic MLCT of a Cu(I)-sulfur transition near 250 nm (Fig. 2). The MLCT reaches a plateau at 1 mol equivalent of Cu(I). The cysteine residues of the metal-binding site are exposed on a loop between the first P-strand and the first a-helix with little protection from the surrounding environment (11). CopZ displays no copper-thiolate luminescence characteristic of a Cu(I)S protein in which the Cu(I) is shielded from the solvent environment (Fig. 2). Protected sites are found in some copper-binding proteins such as metallothionein and the yeast transcription factor Ace1 (22).

X-ray absorption studies of Cu(I)-CopZ have confirmed the coordination of Cu(I) through two sulfur ligands, with an average bond length of 2.244 A (21). Extended X-ray absorption fine-structure (EXAFS) studies of metals coordinated to other PaPPaP folded proteins have all resulted in large Debye-Waller factors indicative of static disorder (23-26). All of these studies have derived bond lengths suggesting a coordination environment with a high proportion of Cu(I) bound to two cysteinyl sulfurs. Digonal coordination of cuprous copper in proteins was, until the discovery of

Fig. 2. Reconstitution of Cu(I) into the copper chaperone CopZ. The increase at 250 nm is the result of the formation of a Cu(I)-S metal-ligand charge-transfer band. Inset: Absence of Cu(I)-S luminescence at 600 nm in Cu(I)-CopZ (excited at 295 nm).

250 300 350

Wavelength (nm)

Fig. 2. Reconstitution of Cu(I) into the copper chaperone CopZ. The increase at 250 nm is the result of the formation of a Cu(I)-S metal-ligand charge-transfer band. Inset: Absence of Cu(I)-S luminescence at 600 nm in Cu(I)-CopZ (excited at 295 nm).

these proteins, thought to be rare although whole-cell EXAFS performed on S. cerevisiae did predict a high proportion of digonal copper (27). By providing only two adjacent sulfurs as ligands, the copper chaperone has exerted pressure on copper to remain digonal. Because the d10 electronic configuration of Cu(I) enforces no stereochemical demands, the coordination is largely determined by the electrostatic and molecular mechanical factors (28). A combination of direct ligands and the electrostatic configuration around the metal-binding loop provides a site suited to two-coordinate Cu(I). This site also allows for the ligand-exchange reactions thought to be involved in the metal transfer from the metallochaperones to the target. Ligand-exchange mechanisms have been proposed for the transfer of Cu(I) from a two-coordinate site to a two-coordinate site (Atx1-Ccc2) (23), from a two-coordinate to a three-coordinate site (CopZ-CopY) (21), and for the transfer of mercury (MerP-MerT) (29).

3. COPY: A COPPER-REGULATED REPRESSOR

The expression of the cop operon is dependent on the intracellular concentration of copper. The expression of the operon is biphasic because the operon encodes both the import and export pumps. Transcription is represssed under normal copper conditions by a DNA-binding protein, CopY. CopY is a homodimer repressor that binds across the transcription start site to a 27-base-pair inverted repeat. DNA footprinting and site-directed mutagenesis of the regulatory region has revealed an ACA triplet at -61 and -30 (from translation start site) to be critical to DNA binding (10). The secondary structure predicted for CopY suggests a multidomain protein. The first domain is predicted to have a helical bundle structure made up of three repeated a-helices making up a DNA-binding domain. A second domain is proposed as a metal-binding domain; this region lacks regular secondary structure and encompasses a CxCx4CxC motif (Fig. 3). The CxCx4-5CxC motif is repeated in the metallothionein family and in a number of transcription factors involved in the expression of metalloregulated proteins. These transcription factors include Ace1 (S. cerevisiae) (22), AMT (Candida glabrata) (30), Macl (S. cerevisiae) (31), and Grisea (Podospora anserina) (32). Interestingly, the Cys motif is found in a human homolog of unknown function (accession no. A161517) (Fig. 4).

CopY is one of the target molecules for Cu(I)-CopZ. DNA-binding activity of CopY is dependent on metal occupancy of the CxCx4CxC motif (8,10). In the metallotransfer reaction, CopZ transfers univalent copper ions into the site occupied by the single divalent zinc, displacing the zinc. The size difference between CopZ/CopY coupled with atomic absorption spectroscopy enables metal transfer to be monitored (8). Additionally, the CxCx^5CxC homologs have been demonstrated to have lumi-

Fig. 3. CopY secondary-structure prediction and proposed domains. The cysteine residues of the metal-binding domain are highlighted.

AMT1 qekgitieedmlmsgnmdmClCvrgepCrCharrkrtqksnkk

ACE1 cmcasarrpavgskedetrCrCdegepCkChtkrkssrks

Macl evlthkgif IstqCsC. edesCpCvnclihrseeelns

MT2 P domain mdpnCsCatggsCtCtgsckckeckcns

Fig. 4. Sequence alignment of CxCx(4-5)CxC motifs from bacteria, yeast, and humans. CopY is a repressor protein from E. hirae. AMT1 is a transcription activator for metallothionien from C. glabrata. ACE1 is a transcription activator for metallothionein from S. cerevisiae. Macl is transcription factor for the CTR/FRE proteins from S. cerevisiae. Grisea is a Macl ortholog from Podospora ansernia. MT2 P-domain is the N-terminal domain of metallothionien from H. sapiens. AF161517 is a human protein of unknown function.

nescent Cu(I)-S cores (33). Copper coordinated to sulfur in a solvent-shielded environment is luminescent at room temperature (28). Cu(I)-S luminescence is observed as the CopZ/CopY metal transfer proceeds. This Cu(I)-S luminescence is the result of Cu(I) binding to CopY, as the Cu(I)-S site in CopZ is solvent accessible. This can be exploited to determine the copper-binding stoichiometry of CopY. The increase in Cu(I)-CopY luminescence plateaus at 2.0 Cu(I)/CopY (Fig. 5). AMT titrated with copper beyond the plateau results in a decreased specific activity of the transcription factor (30). The binding of these copper ions occurs in an all-or-nothing-type mechanism that has been identified previously in homologous CxCx4-5CxC proteins (33).

X-ray absorption studies of Cu2CopY confirmed the cuprous state and provided evidence for Cu(I)-S bonds of 2.26 A. This bond length is best fit with three sulfur bonds; therefore, Cu(I) coordination in CopY requires the sharing of two sulfur ligands. A secondary scatter at 2.69 A, usually attributed to a metal-metal interaction, is also evident from the EXAFS analysis (21). These X-ray data suggest a tight Cu(I)2S4 core. The distorted planar geometry of three-coordinate copper and the sharing of ligands mean that the four sulfur groups should be mostly planar. Native CopY is isolated with 1 mol equivalent of Zn(II) per CopY. This zinc is displaced upon copper binding. In contrast to the Cu(I)-S planar geometry the zinc in Zn(II)-CopY should be coordinated in a tetrahedral geometry, with three sulfurs in a plane and one sulfur axial. The structural changes in the metal-binding domain when Cu(I) binds is dramatic enough to effect the DNA-binding activity of the protein. The addition of copper and reorientation of the metal-binding sulfurs affects DNA binding but does not

Fig. 5. Copper titration of CopY with Cu(I) as Cu(I)-CopZ. (A) The dotted line represents Zn(II)-CopY; the solid line represents 2Cu(I)-CopY. (B) The stepwise increase in relative luminescence at 600 nm for Cu(I)-CopY. The plateau represents the formation of a stable Cu(I)2 S4 core in CopY.

Fig. 5. Copper titration of CopY with Cu(I) as Cu(I)-CopZ. (A) The dotted line represents Zn(II)-CopY; the solid line represents 2Cu(I)-CopY. (B) The stepwise increase in relative luminescence at 600 nm for Cu(I)-CopY. The plateau represents the formation of a stable Cu(I)2 S4 core in CopY.

affect the dimer interaction of CopY. Gel filtration of the CopY species in the transfer reactions indicates no observable difference in size.

4. THE CHAPERONE-TARGET INTERACTION

Metallochaperones facilitate the transport of metal around the cell to its site of utilization. Therefore, metallochaperones may have at least dual specificity: one for the site of metal uptake and another for the target protein. The S. cerevisiae chaperones Atxl, Cox17, and CCS identified for CCC2, CCO, and SOD1 may deliver copper to a wide variety of copper-dependent enzymes, including the transcription factors Acel and Macl. Current evidence suggests that the copper chaperones have limited targets and that the specificity for the target is high (19). The yeast chaperone-target interactions, identified to date, appear to involve similarly structured domains and/or metal binding sites in the copper-transfer reaction. The Atx1-CCC2 interaction proceeds between two open-faced P-sand-wiches with MxCxxC binding sites (same domains/same sites) (34). The CCS-SOD1 interaction requires the formation of a heterodimer between CCS-SOD1 and results in the movement from a Cys-Cu-bridged complex to the SOD1 Cu site (same domains/different sites) (35,36). The only confirmed target for E. hirae's CopZ is CopY. The other potential targets for CopZ include the import and export ATPase pumps. The import pump, CopA, is similar to the Menkes, Wilson, and CCC2 ATPases in that the N-terminus contains a domain with a MxCxxC metal-binding site (37). CopA has only one PaPPaP domain compared to six in the Menkes/Wilson proteins and two in CCC2. The interaction between CopA and CopZ represents a potential same domain/same site type interaction, similar to the Atx1-CCC2 interaction. CopB, the copper export ATPase, has a cytoplasmic domain rich in histidine residues (38). If CopZ is the only copper chaperone in E. hirae, then it may have three different targets. A copper homeostasis model for E. hirae is summarized in Fig. 6, with the assumption that the CopZ is the only copper chaperone.

5. THE MECHANICS OF THE CHAPERONE-TARGET INTERACTION

CopZ transfers copper to CopY, leading to a structural change in CopY and a concomitant decrease in DNA-binding activity. A structural homolog of CopZ, MNKr2, the second Cu(I)-binding subdomain at the amino terminus of the Menkes protein, is unable to transfer copper to CopY (8). CopZ does not transfer or exchange metal to either MNKr2 or the complete amino terminal of the

Fig. 6. Schematic representation of copper routing in E. hirae. The import and export ATPase pumps CopA and CopB control the intracellular concentration of copper by pumping metal across the membrane into and out of the cytoplasm. Cytoplasmic copper is transferred, by CopZ, to the repressor, CopY, which induces expression of the operon.

Fig. 6. Schematic representation of copper routing in E. hirae. The import and export ATPase pumps CopA and CopB control the intracellular concentration of copper by pumping metal across the membrane into and out of the cytoplasm. Cytoplasmic copper is transferred, by CopZ, to the repressor, CopY, which induces expression of the operon.

Menkes protein (unpublished data). The subdomains at the amino terminus of the Menkes ATPase have been shown to have a role in the copper regulated translocation of the Menkes protein from the trans-Golgi network to the plasma membrane (39). Individually, the structure and metal-binding properties of the subdomains are identical to the reported properties of the metallochaperones (40). Because MNKr2 is predicted to have the PaPPaP global fold and the MxCxxC motif seen in CopZ, the ability to recognize a target is not dependent on these general elements. The electrostatic surface of CopZ and its homologs have charged residues grouped into charged faces (11,18). The organization of these faces appears to be individual for the different chaperones. Electrostatic interaction has been widely implicated in the interaction of the chaperones with targets. A charge relationship between CopZ and CopY was originally proposed when the cop operon was first described (9). The structure of Atx1 and modeling of its target, CCC2, has revealed oppositely charged faces that are proposed to interact (18). Site-directed mutagenesis of Atx1 has revealed the importance of particular lysine residues forming patches at the start of the first helix and the end of the second helix. Mutagenesis of these two patches resulted in a dysfunctional Atx1. Mutation of a single lysine residue located close to the MxCxxC motif indicated that this residue is necessary for both delivery of copper to CCC2 and Atx1's antioxidant role (41). In the Menkes modules, MNKr4 and MNKr2, a phenylalanine occupies this position. A large number of Nuclear Overhauser Effect (NOE) contacts from this Phe to the methionine of the MxCxxC motif suggest that this Phe's role is to assist in the stabilization of the metal-binding loop (17). Indeed, this residue may orientate the loop for correct copper binding.

The in vitro assay for copper transfer between CopZ and CopY linked to mutations of MNKr2 provides us with a powerful system to exploit gain-of-function mutations (42). The Atx1 lysine residues, critical for chaperone-target docking in Atx1-CCC2, are not found at the same position in CopZ (Fig. 7). MNKr2 shares similar lysine positions to Atx1 but lacks the lysine arrangement of CopZ (Fig. 7). This may explain the inability of MNKr2 to deliver Cu(I) to CopY. Whereas Atx1 has lysine residues on the helical regions, CopZ has lysine patches on the P-sheet and these patches were

Fig. 7. The lysine arrangement in CopZ, MNKr2, and Atxl. The residues critical to the chaperone function in Atxl (highlighted) are not present in CopZ. The residues highlighted in CopZ at positions 31, 32, 37, and 38 have been engineered into MNKr2. The structures of CopZ and Atxl show that the lysine residues are on opposite sides of the molecules.

Fig. 7. The lysine arrangement in CopZ, MNKr2, and Atxl. The residues critical to the chaperone function in Atxl (highlighted) are not present in CopZ. The residues highlighted in CopZ at positions 31, 32, 37, and 38 have been engineered into MNKr2. The structures of CopZ and Atxl show that the lysine residues are on opposite sides of the molecules.

3 hJ

0 0

Post a comment