|Home | About | Journals | Submit | Contact Us | Français|
Bacteria possess cytosolic proteins (Csp3s) capable of binding large quantities of copper and preventing toxicity. Crystal structures of a Csp3 plus increasing amounts of CuI provide atomic‐level information about how a storage protein loads with metal ions. Many more sites are occupied than CuI equiv added, with binding by twelve central sites dominating. These can form [Cu4(S‐Cys)4] intermediates leading to [Cu4(S‐Cys)5]−, [Cu4(S‐Cys)6]2−, and [Cu4(S‐Cys)5(O‐Asn)]− clusters. Construction of the five CuI sites at the opening of the bundle lags behind the main core, and the two least accessible sites at the opposite end of the bundle are occupied last. Facile CuI cluster formation, reminiscent of that for inorganic complexes with organothiolate ligands, is largely avoided in biology but is used by proteins that store copper in the cytosol of prokaryotes and eukaryotes, where this reactivity is also key to toxicity.
Important metabolic enzymes in eukaryotes and prokaryotes require copper for their active sites.1 To prevent toxicity, eukaryotes store excess cytosolic copper using metallothioneins (MTs).2, 3, 4, 5 The ability of bacteria to maintain appreciable amounts of intracellular copper has only recently been discovered.6, 7 This is achieved by a family of copper storage proteins (the Csps), which are tetramers of four‐helix bundles possessing Cys‐lined cavities binding up to approximately 20 CuI ions per monomer. The twin‐arginine translocated Csps (MtCsp1 and MtCsp2) from the methanotroph Methylosinus trichosporium OB3b are involved in copper storage for the main methane‐oxidizing enzyme.6 This organism also possesses a Csp3 (MtCsp3), homologues of which are much more widespread and allow bacteria to accumulate cytosolic copper.7 Csp3s may safely store CuI for currently unknown cytosolic copper enzymes or for export by the copper‐transporting ATPase CopA. MtCsp1 has 13 Cys residues and binds a similar number of CuI ions,6 whereas MtCsp3 has 18 Cys residues and accommodates 19 CuI ions.7 Both bind CuI via sites with highly novel coordination chemistry. Filling the complete core with metal ions, as in the Csps, has not been observed previously for either naturally occurring or engineered four‐helix bundles.8, 9, 10, 11, 12 Thiolate‐coordinated CuI clusters are remarkably rare in biological systems, but the crystallization of MtCsp3 in the presence of increasing amounts of CuI provides unprecedented insight into how such species form in a protein and are used to drive copper storage.
The crystal structure of MtCsp3 plus ca. 2 molar equiv of CuI (see Supporting Information6, 7, 13, 14, 15) has four partially occupied sites (Figure 1 and Table S1 in the Supporting Information). These are Cu11 with an occupancy of 0.25, and Cu12, Cu13, and Cu14, all with occupancies of 0.35, which form a symmetrical tetranuclear cluster (Figure 1b,c). Cu12 and Cu14 are bound by the Cys residues of CXXXC motifs, that is, from the same α‐helix, whilst Cu11 and Cu13 are ligated by Cys residues on different helices (Cu13 is also weakly coordinated by Asn58). The thiolates of Cys97, Cys101, Cys114, and Cys118 bridge between two CuI ions (μ2‐S‐Cys). Additional stabilization is provided by CuI–CuI interactions (ca. 2.6–2.7 Å) between neighboring metal ions (Figure 1b). Upon adding ca. 9 equiv, CuI ions are found at 18 locations in MtCsp3 with a total occupancy of 8.2 (Figure 2 and Table S2 in the Supporting Information). Cu3 to Cu14 have the highest occupancies, with three tetranuclear CuI clusters present; Cu3–Cu6 and Cu7–Cu10, as well as Cu11–Cu14 (Figure 2b–d). The addition of ca. 17 equiv of CuI to MtCsp3 results in binding at 22 sites with a total occupancy of 13.8 (Figure 3 and Table S3 in the Supporting Information). Occupancies increase for all sites in the Cu3–Cu6, Cu7–Cu10, and Cu11–Cu14 clusters, which still constitute the majority of the CuI core (Figure 3b–d). Alternative forms (Cu5b, Cu7b, Cu9b, and Cu11b) are found at four of the six inter‐helical sites making up Cu3 to Cu14 and the short CuI to CuI distances (ca. 1.4 to 1.6 Å) between these and the nearest high occupancy site indicates that both cannot be present within the same molecule (the sum of van der Waals radii is 2.8 Å). These alternate sites contribute to [Cu4(μ2‐S‐Cys)4] clusters (Figures 2b–d and and3b–d)3b–d) similar to that observed for Cu11‐14 in the 2 equiv structure (Figure 1b). However, the major species are [Cu4(S‐Cys)5]− (Cu3–Cu6), [Cu4(S‐Cys)6]2− (Cu7–Cu10), and [Cu4(S‐Cys)5(O‐Asn)]− (Cu11–Cu14), all with three μ2‐S(Cys) and either two (Cu3–Cu6 and Cu11–Cu14) or three (Cu7–Cu10) Cys ligands that bind a single CuI ion at that particular cluster. The major clusters have more ligands and are less symmetrical, highlighted by increased variation in Cu to Cu distances (compare Figure 1b and and2b–d).2b–d). The position of Cu11 in 2 equiv MtCsp3 corresponds to Cu11b in the 8 and 14 equiv structures, which is therefore bound before Cu11a. Cu5b, Cu7b, Cu9b, and Cu11b are all minor species and [Cu4(μ2‐S‐Cys)4] intermediates probably form prior to the final clusters.
Cu15 to Cu19 (Figure 4 and Table S3 in the Supporting Information) towards the mouth of the bundle (Figure 1a) are all partially occupied in 14 equiv MtCsp3 (Cu15, Cu16, and Cu18 are in the 8 equiv structure). Cu18 exists in two equally occupied (0.40) two‐coordinate sites separated by 2.1 Å (not present in the same monomer). Both are ligated by His110, with either Cys111 (Cu18a) or Cys38 (Cu18b) as the second ligand. Cu18a corresponds to the site in the fully CuI‐loaded (19 equiv) structure (Figure 4c and Table S4 in the Supporting Information), whilst Cu18b is occupied in the 8 equiv structure (Figure 4a), and CuI ions bind at Cu18b before Cu18a. Cu19 is relocated by more than 1 Å in the 14 equiv compared to 19 equiv structure (Figure 4b,c), being more distant from His104 (the imidazole of His104 rotates by ca. 180° in the 19 equiv structure allowing its Nδ1 atom to coordinate Cu19). Cu17 is 1.8 Å from Cu19 in the 14 equiv structure and both cannot be occupied in the same molecule. Cu15 is two‐coordinate in the 8 equiv and 14 equiv structures, whilst in 19 equiv MtCsp3 it is the only site ligated by three Cys ligands,7 primarily due to altered conformations of Cys101 and Cys111. Five Cys and two His residues rearrange to accommodate the Cu15‐Cu19 cluster at the mouth of the four‐helix bundle as MtCsp3 fills with CuI.
More sites are occupied than the number of CuI equiv added in all of the MtCsp3 structures. Furthermore, no site is fully occupied until the 14 equiv structure (Cu12), and there is a general tendency that CuI favors binding at CXXXC motifs. CuI ions populate numerous sites as the bundle fills, with a clear preference for those towards the center, and particularly in the Cu3–Cu6, Cu7–Cu10, and Cu11–Cu14 tetranuclear clusters (Figure S1 in the Supporting Information). This behavior is consistent with in vitro CuI‐binding properties,7 and a more ordered uptake mechanism for MtCsp3 compared to MtCsp1 (for further discussion see the Supporting Information). The total occupancies of the Cu3–Cu6, Cu7–Cu10, and Cu11–Cu14 clusters range from 2.4 to 2.7 in the 8 equiv structure and increase to 3.5 to 3.7 in 14 equiv MtCsp3 (Figures 2, ,3,3, and Figure S2 and Tables S2, S3 in the Supporting Information). There is no preference for CuI binding at a particular tetranuclear cluster apart from in the structure plus ca. 2 equiv, in which only Cu11–Cu14 is occupied. This must be due to the fact that as CuI ions diffuse past Cu15 to Cu19 at the mouth of the bundle, Cu11–Cu14 are the first sites they encounter at which they can form a tetranuclear cluster. The three tetranuclear clusters probably have similar CuI‐binding affinities, and occupancy of the Cu3–Cu14 core appears thermodynamically favored over other sites.
Copper clusters in proteins are extremely rare, with the CuZ site of nitrous oxide reductase providing the only example of a tetranuclear site in an enzyme.16 However, this site lacks any Cys ligands, and the CuA center, also found in nitrous oxide reductase as well as cytochrome oxidases, is the highest nuclearity copper site involving Cys ligands in an enzyme (Cu2(S‐Cys)2(N‐His)2).16, 17 The scarcity of CuI clusters bound by organothiolates in biological systems, such as those that drive MtCsp3 CuI core formation, is surprising given their rich coordination chemistry18, 19 and the thiophilic nature of CuI. Proteins involved in copper homeostasis, particularly in Saccharomyces cerevisiae, have been found to bind tetranuclear CuI clusters in vitro using Cys residues, but no crystal structures are available.20, 21, 22, 23, 24 Extended X‐ray absorption fine structure data for these are similar to that of [Cu4(SPh)6]2−.20, 21, 23, 25 In this complex the CuI ions are three‐coordinate and all thiolates μ2‐S.18 The crystal structure of a side‐to‐side dimer of a cyanobacterial Atx113 binds a symmetrical [Cu4(μ2‐S‐Cys)4Cl2]2− cluster comparable to the [Cu4(μ2‐S‐Cys)4] intermediates we observe for MtCsp3, but a functional role for this Atx1 dimer remains to be established. Thiolate‐coordinated CuI‐cluster formation is physiologically important in the MTs. These, almost exclusively eukaryotic, Cys‐rich unstructured apo‐polypeptides store cytosolic CuI by folding around clusters.2, 3, 4, 5 The crystal structure of a truncated S. cerevisiae MT (Cup1) has eight CuI ions bound by ten Cys residues, with the majority of sites coordinatively saturated (three‐coordinate).4 These are present as two CuI 4 flattened tetrahedra, each with structures similar to [Cu4(SPh)6]2−.18 The presence of linked tetranuclear clusters in the MT structure is comparable to what we have observed for the Cu3 to Cu14 core of MtCsp3, but most sites in the Csps are two‐coordinate. This is partly due to a low Cys:CuI ratio in the Csps (always about 1) and their four‐helix bundle fold that prevents significant re‐positioning of the Cys residues. Maintaining two‐coordinate sites in Csps may be important either to facilitate CuI release or for the safe use of Cys residues for binding CuI clusters.
The avoidance of copper clusters in enzymes is probably due to the enhanced risk of toxicity. This could be exacerbated by the presence of Cys ligands, due to the potential for CuII‐catalyzed disulfide bond formation (a dedicated protein is required to keep the Cys residues reduced prior to copper insertion at the CuA site26). Biologically important Cys‐bound CuI clusters are found in cytosolic copper storage proteins (Csp3s and MTs) in which disulfides do not occur. In both cases, the reducing nature of the cytosol helps maintain thiols (exported Csps fold in the cytosol6). The rigidity of the four‐helix bundle fold also prevents this reactivity by geometrically constraining Cys residues in the Csps (the Cys residues of apo‐Csps do not readily form disulfide bonds in air6, 7), as well as providing additional protection of bound CuI ions. This is not the case for the unstructured and flexible MTs.
In conclusion, we provide a visual description of CuI loading in a storage protein, highlighting factors important for the formation of the CuI core, as well as the fluxionality and flexibility of CuI binding. We identify [Cu4(μ2‐S‐Cys)4] as an intermediate for the three clusters that are made up of the first sites (Cu3 to Cu14) to be occupied in MtCsp3. The ability to store and protect CuI ions in the cytosol is driven by the formation of thiolate‐coordinated tetranuclear clusters in both prokaryotes and eukaryotes, but is achieved with dramatically different protein structures. Paradoxically, the same facile cluster chemistry of CuI is also responsible for toxicity by the displacement of iron at Cys‐bound 4Fe–4S clusters.27, 28, 29, 30 MTs have been suggested to contribute to a “chelation” rather than “compartmentalization” mechanism to combat copper toxicity in eukaryotes.27 Csp3‐possessing bacteria have a cytosolic detoxification system for copper,7 which also relies on chelation. Thus copper handling by such organisms is more complex than originally thought and requires further investigation.
The authors declare no conflict of interest.
As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.
We thank staff of the DLS synchrotron radiation source for help with diffraction data collection. This work was supported by Biotechnology and Biological Sciences Research Council (grant BB/K008439/1 to C.D.) and by Newcastle University who part‐funded a PhD studentship (S.P.).
A. Baslé, S. Platsaki, C. Dennison, Angew. Chem. Int. Ed. 2017, 56, 8697.