ACS Publications. Most Trusted. Most Cited. Most Read
Bacterial Metallostasis: Metal Sensing, Metalloproteome Remodeling, and Metal Trafficking
More

OR SEARCH CITATIONS

My Activity
Publications

Figure 1Loading Img
Download Hi-Res ImageDownload to MS-PowerPointCite This:Chem. Rev. 2024, 124, 24, 13574-13659
  • Open Access
Review

Bacterial Metallostasis: Metal Sensing, Metalloproteome Remodeling, and Metal Trafficking
Click to copy article linkArticle link copied!

Open PDF

Chemical Reviews

Cite this: Chem. Rev. 2024, 124, 24, 13574–13659
Click to copy citationCitation copied!
https://doi.org/10.1021/acs.chemrev.4c00264
Published December 10, 2024

Copyright © 2024 The Authors. Published by American Chemical Society. This publication is licensed under

CC-BY-NC-ND 4.0 .

Abstract

Click to copy section linkSection link copied!
High Resolution Image
Download MS PowerPoint Slide

Transition metals function as structural and catalytic cofactors for a large diversity of proteins and enzymes that collectively comprise the metalloproteome. Metallostasis considers all cellular processes, notably metal sensing, metalloproteome remodeling, and trafficking (or allocation) of metals that collectively ensure the functional integrity and adaptability of the metalloproteome. Bacteria employ both protein and RNA-based mechanisms that sense intracellular transition metal bioavailability and orchestrate systems-level outputs that maintain metallostasis. In this review, we contextualize metallostasis by briefly discussing the metalloproteome and specialized roles that metals play in biology. We then offer a comprehensive perspective on the diversity of metalloregulatory proteins and metal-sensing riboswitches, defining general principles within each sensor superfamily that capture how specificity is encoded in the sequence, and how selectivity can be leveraged in downstream synthetic biology and biotechnology applications. This is followed by a discussion of recent work that highlights selected metalloregulatory outputs, including metalloproteome remodeling and metal allocation by metallochaperones to both client proteins and compartments. We close by briefly discussing places where more work is needed to fill in gaps in our understanding of metallostasis.

This publication is licensed under

CC-BY-NC-ND 4.0 .
  • cc licence
  • by licence
  • nc licence
  • nd licence
Copyright © 2024 The Authors. Published by American Chemical Society

Subjects

what are subjects

1. Introduction

Click to copy section linkSection link copied!

1.1. Metallostasis

All cells, whether free-living in a particular microenvironment, or part of community of cells in a tissue or a microbial biofilm, must acquire metal ions from the extracellular milieu, bring them across a lipid bilayer and put them to use. Metals generally function as metallocofactors in enzymes that are required to catalyze an enormous diversity of chemical transformations at rates and efficiencies compatible with cellular needs, while also responsive to prevailing metabolic conditions. These metalloproteins, ranging from proteins that use a metal as an enzymatic cofactor to structural metals that stabilize the protein fold, play many roles in biological systems. (1) Approximately one-third to 40% of a typical proteome is a metalloprotein, (2) and cells must ensure that the metalloproteome is metalated with integrity, defined by binding of the cognate or native metal under all growth conditions, yet is sufficiently plastic or adaptable, so as to effectively respond to metal limitation while avoiding metal toxicity. (3)
Metallostasis is the cellular homeostasis of metal ions and encompasses all processes that ultimately impact the status or integrity of the metalloproteome. Metallostasis integrates all mechanisms that ensure proper metalation, including those that function at the transcriptional level, mediated by protein- and RNA-based metallosensors, cotranslationally during protein synthesis, or post-translationally, where metal is inserted into a folded apoprotein (Figure 1). Metallostasis is analogous to, and a subdomain of, proteostasis or protein homeostasis, defined as the dynamic regulation of a balanced and functional proteome (Figure 1). While proteostasis is governed by a “network” that collectively controls the biogenesis, folding, trafficking, and degradation of proteins present within and outside the cell, metallostasis is likewise controlled by a network of metal sensing, transport, speciation and ultimately trafficking of these finite inorganic resources to a specific task (Figure 1). Metallostasis becomes particularly important under environmental conditions that induce metal limitation or starvation. In bacteria, the evolution of protein “paralogs” that perform identical functions in cells, but with distinct inorganic cofactor requirements or none at all, is a particularly dramatic and sometimes overlooked example of metabolic and evolutionary adaptation to bioavailable metals. (4) Examples of these proteins can be found particularly in central carbon metabolism that becomes essential in an infection under these conditions (5,6) and in DNA synthesis (7,8) and, more generally, nucleic acid metabolism. (4,9) In humans, nutritional immunity, where host mechanisms have evolved to limit bacterial pathogen access to nutrient metals, alongside host-mediated metal intoxication, are central aspects of the “tug-of-war” between host and pathogen in infectious disease. (10−12) In addition, dietary metal nutrient deficiency, metal toxicity or impaired metal transport, trafficking and allocation caused by a wide range of human genetic disorders figure prominently in human biology and pathology. (13−17)

Figure 1

Figure 1. Metallostasis (blue) is a subdomain of the proteostasis network (black) that is integrated into the synthesis-folding-degradation paradigm. Metallosensors function at the apex of the metallostasis network and sense cellular bioavailable metal. This results in the regulation of the genes that encode components of the network, and includes proteins involved in metal transport, storage and allocation, and as enzyme paralogs (not shown here). The adaptive response to metal overload or starvation collectively mediated by the metallostasis network enforces an appropriate free energy of metalation (ΔΔG) of the metalloproteome, thereby maintaining its functional integrity.

High Resolution Image
Download MS PowerPoint Slide

1.2. Scope of this Review

Metallostasis is dedicated to ensuring the structural and functional integrity of the metalloproteome (Figure 1). We focus on specific mechanisms discovered from work published within the last 10 or so years, nearly exclusively in bacteria, as representative of the diversity of the adaptive responses in these simple organisms, with reference to eukaryotic mechanisms when parallels may inform bacterial metallostasis systems. In each section, we provide historical context for a discussion of the more recent work that follows, referring the reader to this earlier work as needed. We begin by briefly discussing the metalloproteome and specialized roles that metals play in biology, with a focus on general principles and specific illustrations of what constitutes structural, regulatory and catalytic metal sites in proteins. Metallostasis is tightly orchestrated by upstream “sentinels”, metallosensors, specialized proteins or RNAs specific for a cognate metal that detects the bioavailability of that metal in the cell, and by doing so, control the expression of genes that encode other players in the metallostasis network (Figure 1). We discuss these proteins in detail, emphasizing general concepts that have emerged from comprehensive studies of these model “allosteric” switches (18) and more recent efforts to employ metallosensors in synthetic biology and biotechnology applications. (19) This is followed by a discussion of enzyme paralogs as a remarkable example of metalloproteome remodeling, albeit one generally limited to bacteria. (4) We then discuss the trafficking of metals to downstream metalloenzyme targets, contextualizing this discussion around the concept of a metallochaperone, focusing first on the biogenesis or maturation of metallocofactors in specific enzymes, followed by moving metals across a membrane. We close by briefly discussing places where more work is needed to fill in gaps in our understanding of metallostasis, and how this new knowledge might be leveraged to develop new therapeutic strategies to combat infectious disease.

2. The Metalloproteome

Click to copy section linkSection link copied!

2.1. Metals in Biology

2.1.1. Overview

The metalloproteome comprises all of those proteins that bind a specific or cognate metal ion, or in rare cases a number of closely related metal ions, to execute a well-defined biological function. These metal ions can function as structural, regulatory or catalytic cofactors when bound to a protein through coordinate covalent bonds. In the context of this Review, we explicitly consider the first-row, late 3d-block metal ions from manganese (Mn; d (5) in the 2+ oxidation state or MnII) to zinc (Zn; d10) and including all the elements between these two: iron (Fe; d6), cobalt (Co; d7), nickel (Ni; d8) and copper (Cu; d9) (Figure 2). We limit our discussion to these metals, since, with the exception of nickel, (20) each plays important roles in all kingdoms of life and cells are known to sense and actively regulate their intracellular availability through the mechanisms we discuss here. (21,22) The reader is referred to other recent articles, reviews or collections of articles that discuss biological metals and processes beyond these metals, notably nitrogenase and the maturation of the molybdenum cofactor (Moco) (23) and other molybdenum-containing enzymes (24) in bacteria, (25,26) plants and photosynthetic organisms, (27,28) and humans, (29) as well as tungsten and vanadium-cofactored enzymes. (20,30) In addition, we do not discuss recent exciting findings on the chemistry and biology of trivalent lanthanides (31,32) and associated emerging work on employing the basic science of rare earth element molecular recognition to develop robust separations technologies designed for field-deployment. (33) Finally, we do not consider the major metalloid selenium (Se), which functions in proteins in a twenty-first amino acid, selenocysteine, with the selenoproteome playing important roles in both bacteria and eukaryotes. (34,35)

Figure 2

Figure 2. Abbreviated version of the periodic table of the elements highlighting the biologically important metals. s-block mono- and divalent metals (dark green); essential d-block metals (light green); d-block metals that play roles in some forms of life (yellow); chromium (orange) and the metalloid selenium (light cyan) are also indicated. The early lanthanides (tan; La) are used by some archaea, (32) notably the methylotrophs, which incorporate Ln into methanol dehydrogenases. Other elements from both d- and p-blocks are considered xenobiotics and sensed by metallosensors (light pink) (Section 3). The biologically relevant metals discussed are from the late 3d-block, from Mn to Zn. (21)

High Resolution Image
Download MS PowerPoint Slide

2.1.2. Structural Metal Sites

A structural metal ion stabilizes the protein fold and might be considered the inorganic analog of a long-lived disulfide bond that is found in proteins that reside in, or transit the endoplasmic reticulum in eukaryotic cells, and fulfills its function in the more oxidizing environment of the extracellular space. Since the reducing potential of the cytoplasm of a bacterial cell and the cytosolic compartment of a plant cell or a eukaryotic cell often precludes the widespread formation of thermodynamically stable disulfide bonds (see below), a metal ion like ZnII, which is stable in the 2+ oxidation state and refractory to oxidation–reduction chemistry, can perform such a function. Structural ZnII coordination sites typically adopt tetrahedral coordination geometries as classically observed in zinc finger (ZF) proteins or domains (Figure 3A). ZF domains typically incorporate cysteine (Cys) thiolate coordination but can also harbor histidine (His) imidazole and mono- or bidentate carboxylate donors from aspartate (Asp) and glutamate (Glu) in these sites. These structural ZF sites are coordinatively saturated, i.e., they lack a solvent-accessible or “open” coordination site and are thought to be kinetically and/or thermodynamically stable, exhibiting very slow off rates. (36,37) A structural ZnII atom nucleates folding around four metal–liganding residues that are often closely spaced in the primary structure, which then induces the folding of a subdomain of a protein, organizing a surface that promotes protein–protein or protein-nucleic acid (DNA or RNA) interactions. In the case of a ZF, these are “mini-folds” that lack a conventional hydrophobic core, and thus do not stably fold in the absence of metal–ligand bonds. Indeed, ZF proteins adopt to a wide diversity of folds that differ in the type and arrangement of metal ligands in the primary structure, e.g., CCHH, CCCH and CCHC, respectively, suggesting strongly divergent evolutionary solutions to a common protein folding problem (Figure 3A-D).

Figure 3

Figure 3. Zinc finger protein gallery illustrating the structural diversity of coordinatively saturated, approximately tetrahedral ZnII complexes involved in protein–protein and protein-nucleic acid interactions. A) Cross-brace-like (ring-finger-like) double zinc-finger domain of murine METAP1 (7SEK); (56) B) Zinc-sensing site of the yeast Zap1p, illustrating a tandem array of canonical Cys2-His2 sites (1ZW8); (40) C) Tandem array of the two canonical CCHC retroviral-type zinc-fingers (1AAF); (57) D) SARS-CoV-2 nsp13 ring finger domain superficially related to the METAP1 domain, harboring a third CCHC ZnII binding domain that is known to coordinate a [4Fe-4S] cluster in solution (6ZSL). (58)

High Resolution Image
Download MS PowerPoint Slide
Canonical Cys2His2 (CCHH) ZF domains that adopt a ββα structure are nearly always tandemly repeated, “beads on a string” proteins (Figure 3B) are often nuclear and typically function by binding to a specific DNA sequence in the promoter regions of regulated genes, and thus are transcription factors in eukaryotes and all multicellular organisms. The presence of hundreds of Cys2-His2 ZF proteins in eukaryotes is the major reason why 9–10% of a typical eukaryotic proteome consists of zinc metalloproteins, while this fraction tends to be only 5–6% in bacteria. (36) It should be emphasized that the extent to which the structural integrity of these ZF sites, as major contributors to the zinc metalloproteome, is dictated by the cellular zinc bioavailability is unknown and remains an exciting and largely unresolved area of study. (38) It is known, however, that the major zinc-sensing transcription factors in eukaryotic cells, exemplified by vertebrate MTF-1 and Zap1p in Saccharomyces cerevisiae (Figure 3B) employ one or more ZF motifs as “sensors” of cellular zinc status, which results in the regulation of transcription of genes that encode enzymes that re-establish zinc homeostasis. (37,39,40)
The thermodynamic stability of structural ZF sites should also not be taken to suggest that they are unreactive. In fact, recent work reveals that the Cys residues in a canonical RNA-binding CCCH ZF protein, tristetraprolin (TTP) involved in regulating inflammation, are subject to persulfidation by hydrogen sulfide (H2S) and molecular oxygen, which may involve the intermediacy of hydrogen disulfide radical anion and superoxide anion, followed by disulfide bond formation and zinc release. (41) A cysteine within a CCHH zinc-finger structure is also known to react with the signaling molecule, nitric oxide (NO•), in a global regulatory response to a nitric oxide burst as part of the plant immune response to infection. (42) These findings follow on many years of work to develop electrophilic molecules, notably aromatic thioesters, that target the Cys residues in one of the two retroviral nucleocapsid protein ZF domains of the Gag polyprotein (Figure 3C), mediating zinc release or “zinc ejection”, as a promising antiretroviral strategy. (43−46) The extent to which this oxidative chemistry is regulatory in cells during the course of normal cellular metabolism remains to be explored but may be a means to connect the redox stress response to fluxes in cellular zinc that result from this oxidative chemistry.
Indeed, there is emerging evidence in support of the idea that what were once assumed to be “structural” zinc sites may be far richer in bioinorganic redox chemistry than previously imagined. The presence of four closely clustered metal–liganding Cys and/or His residues as in canonical ZF proteins, are often simply assumed to bind ZnII often with little empirical evidence. (47) There is now significant support to suggest that a subset of these ZF sites do not bind ZnII at all, but instead coordinate either an [2Fe-2S] or a [4Fe-4S] cluster, some of which are redox-active. (48−50) An important early example of this was found in Cleavage and Polyadenylation Specificity Factor 30 (CPSF30), which plays a role in mRNA pre-mRNA processing. (47,51) This is consistent with the general finding that some DNA and RNA metabolizing modifying enzymes incorporate an Fe–S cluster, which modulates enzyme activity, and may well mediate long-range electron transport through the nucleic acid to regulate and coordinate enzyme activity during DNA replication and repair processes. (48,52,53) Indeed, in recent work, at least two viral proteins encoded by severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), including the RNA-dependent RNA polymerase (nonstructural protein 12, nsp12) as well as the nsp13 helicase (Figure 3D), each harbor a redox-active and labile [4Fe-4S] cluster, that are potential targets for antiviral therapies. (54,55)

2.1.3. Regulatory Metal Sites

Structural metal sites, like some of the Fe–S cluster sites described above, can also play regulatory roles in the absence of metal oxidation or reduction. These sites are either allosteric, which function “at a distance” from the active site, or “competitive” with the binding of substrate or the catalytic chemistry in the active site. Identification of these regulatory sites relies on some functional discovery followed by biochemistry, and thus are typically much more difficult to identify in a complex proteome. A paradigm example of reversible modulation of enzyme activity by transition metals is the inhibition of protein tyrosine phosphatase 1B (PTP1B) by ZnII, (59,60) which contrasts sharply with that of MgII, which is an activator. (61) PTP1B is important target for the development of a potential therapy for obesity and insulin resistance. Both ZnII and MgII compete for the same site in the active site region of the enzyme, which becomes available only following formation of the phosphorylated Cys intermediate (Figure 4A). ZnII negatively impacts activation of a metal-coordinated water molecule that is required for hydrolysis of the phosphothioester linkage. (61) ZnII binds to PTP1B with a dissociation equilibrium constant (Kd) in the low pM range, consistent with its physiological impact on this enzyme in cells (see Section 2.2).

Figure 4

Figure 4. A) Active site of human PTP1B in a representative snapshot at the end of QM/MM MD simulations, bound to MgII (left) and ZnII (right). Residues C215-phosphate, D181, and water molecules coordinating the metal are shown as sticks. B) Ribbon diagram of Klebsiella pneumoniae NDM-1, with ZnII ions shown as black spheres, bound water shown as a red sphere, and metal-binding side chains shown as sticks (left). Chemical structure of aspargillomarasmine A (AMA, right), and inhibitor metallo-β-lactamases. C) Ribbon diagram of AbPBP2, with the TPase domain colored light blue, linker region cyan, and head region dark blue. Conserved active site residues and metal binding residues are shown as sticks including C384, with the active site serine, S383, labeled. Panel A was reprinted with permission from J. Am. Chem. Soc. 2018, 140, 4446–4454. Copyright 2018, American Chemical Society.

High Resolution Image
Download MS PowerPoint Slide
In other cases, the regulatory ZnII coordination site may be at or very near the protein surface where it functions at a distance as an allosteric modulator of protein function. These sites are formally analogous to the metalloregulatory metal sites in metallosensors (Section 3), except that they reversibly control enzyme activity. Regulatory sites like these are often difficult to identify a priori, and are often discovered using crystallographic methods and working backward using reverse genetic approaches, toward establishing functional significance, or through the use of coevolution- or structure-based metal site prediction algorithms, e.g., MetalNet, (62) Metal3D (63) and related approaches, (64−67) the predictions of which can be tested experimentally using biochemical methods. These approaches often rely on the conservation of the metal–ligand residues in the primary structure in a group of closely related enzymes using pattern-matching, and recent computational methods have been used to distinguish adventitious from physiologically relevant sites worthy of detailed study (65) as well as distinguish catalytic from noncatalytic structural sites (66) have recently been reported. The use of neural networks and machine learning approaches promises to impact the precision and accuracy of these predictions, which will require experimental validation using metalloproteomics approaches. (68)
2.1.3.1. Regulatory Metal Sites and Nutritional Immunity
One recent example of the discovery of a new regulatory metal site using crystallographic approaches is in a periplasmic, high molecular weight (HMW) penicillin-binding protein, PBP2, from bacterial nosocomial pathogen, Acinetobacter baumannii. (69) PBP2 is a transpeptidase that is an important component of the “elongasome” that maintains the rod-shaped morphology of the bacterium and a major target for the β-lactam antibiotics from the carbapenem family of antibiotics. These are the same antibiotics that are inactivated by bacterially encoded β-lactamases, including the metallo-β-lactamases (MBLs) that require ZnII for their hydrolytic activity (Figure 4B). Nutrient zinc starvation is a key feature of “nutritional immunity”, a potent antimicrobial strategy employed by the infected host to keep pathogens from causing infections beyond the initial site of colonization. (10,70−72) The crystallographic structure of AbPBP2 revealed an unexpected and conserved tetrahedral SNO2 ZnII coordination complex that is just 10 Å from the catalytic serine (Figure 4C). Mutagenesis of any of the metal ligands, notably the single thiolate ligand, C384 (Figure 4C) disrupts metal binding, and induces a thermal destabilization on par with metal removal by EDTA. More importantly, mutant A. baumannii strains harboring one of these mutations which are expressed at wild-type-like levels are far more susceptible to growth inhibition by two β-lactam antibiotics that do not specifically target PBP2. In addition, all four missense mutations lead to a cell-widened, spherical morphology, identical to that of the Δpbp2 strain. These compelling biological data, coupled with the conservation of this zinc-binding motif in PBP2s from many β- and γ-classes of Proteobacteria, suggest that host-derived metal restriction predisposes A. baumannii strains to antibiotic susceptibility. ZnII limitation also induces the expression of a low-zinc “paralog” of a housekeeping D,D-carboxypeptidase, ZrlA, that enhances resistance to cell wall envelope stress under the same conditions. (73) Left for future work is the mechanism by which ZnII binding to this site in AbPBP2 is truly reversible and sensitive to changes in bioavailable metal in the bacterial periplasm, and how ZnII binding protects PBP2 from antibiotic acylation.
Nonetheless, studies like these illustrate the considerable complexity of “nutritional immunity” as an important weapon in the arsenal of host defenses used to combat bacterial invaders. Further, they highlight cell wall homeostasis as an important target of zinc-limitation in these and likely other ways. Indeed, recent work reveals that MBLs, in particular those derived from subclass B1 from carbapenem-resistant organisms, including the New Delhi MBL (NDM), Verona integrin-encoded MBL (VIM), and imipenemase (IMP)-type enzymes, are a potential target of host-mediated metal restriction in the periplasm, given a binuclear ZnII active site, where the more weakly bound Zn2 metal appears more susceptible to chelation by metal chelators including a secondary metabolite of fungal origin, aspergillomarasmine A (AMA), an aminopoly(carboxylic acid) with KdZn of ≈0.2 nM (Figure 4B). (74,75) While NDM-1 β-lactamase is susceptible to inhibition by AMA, a more recently appearing allelic variant, NDM-6, is not, which is at least partly explained by ≈20-fold higher zinc affinity relative to NDM-1. Indeed, the evolution of increased zinc-affinity appears rather well-established in NDM-type MBLs, of which more than 20 new allelic variants have arisen within the past decade, all with increased zinc affinity. (76,77) Although ZnII-removal increases the rate of protein degradation in bacterial cells in an MBL-specific way, (77) recent data for NDMs suggest that this picture is far more complex, with recently arising mutations linked to an enhanced kinetic and conformational stability and therefore increased lifetime of the β-lactamase in the periplasm, i.e., the proteostasis network (Figure 1), with ZnII affinity just one, albeit important aspect of this evolution. (78)
2.1.3.2. Allosteric Copper Sites: Lipid metabolism, Cancer and Beyond
Copper homeostasis is now known to have a profound impact on many aspects of mammalian physiology and human disease, beyond serving as a catalytic cofactor in cuproenzymes. (79−81) For example, intracellular copper status, altered either pharmacologically or genetically in cultured vertebrate cells, was shown to be an important regulator of lipolysis, the catabolism of fats. The evidence suggests that CuI binding allosterically inhibits the activity of a cyclic-AMP-degrading phosphodiesterase, PDE3B, by coordinating a pair of Cys residues found only in the PDE3 subtype of phosphodiesterases (Figure 5A, left; Figure 5B). (81) This observation, building on previous work in cancer biology, has led to the general concept of “copper signaling” (80) and “metalloallostery” (82) where variations in intracellular or organellar Cu bioavailability, mediated by Cu-specific transporters that transport copper across membranes, (83) regulates enzyme activity and signal transduction via as yet largely unknown mechanisms.

Figure 5

Figure 5. Copper metalloallostery in vertebrate systems. A) Ribbon diagram of the structure of human PDE3B with the active site a strong candidate Cu binding site that negatively modulates the activity of the enzyme. B) Excess cellular Cu (pink sphere) impacts the specific activities of a number of signal transduction pathways in the various cells. STEAP is a metalloreductase that reduces CuII to CuI, (97) which is transported down a concentration gradient, picked up by a metallochaperone (MC) and trafficked to various signaling proteins, with distinct biological outcomes. (81,94)

High Resolution Image
Download MS PowerPoint Slide
This PDE3B-specific model of Cu-dependent regulation of enzyme activity in lipolysis may well be general, as Cu is known to regulate the activity of a number of protein kinases involved in various cell signaling pathways (Figure 5A). (84) These include pathways that are upregulated as part of the host immune response against bacterial infection, (85) and most prominently in cancer cells, consistent with the higher cellular demand of cancer cells for this micronutrient. (86) Enzymes that appear to be regulated by direct binding of Cu including mitogen-activated protein kinase kinase 1/2 (MEK1/2), which is required for downstream phosphorylation and activation of ERK1/2 and thus potentiation of the mitogen-activated protein (MAP) kinase pathway indeed, the MAP kinase pathway is a significant repository of oncogenic mutations and a long-standing target of cancer therapies (Figure 5B). (87,88) Subsequent studies suggested that the copper chaperone for superoxide dismutase (CCS) is responsible for chaperoning Cu to MEK1/2, (89) which would be required given the extremely limited bioavailability of CuI in cells (Section 2.2). (90) The other major Cu chaperone in mammalian cells, ATOX1, may also impact MAP kinase signaling pathway, (91,92) although the mechanistic details are not in. Finally, Cu is known to function as a positive allosteric modulator of autophagy in cancer biology, again via direct binding to a specific kinase in the P13K phosphoinositide 3-kinases)/AKT signaling pathway. (93−95) The extent to which cuprous vs cupric copper is the signaling moiety in these processes is an important mechanistic question not yet fully resolved. The recent development of a robust CuII-specific small molecule sensor strongly suggests that CuII, rather than CuI, mediates at least some of these Cu-dependent signaling processes described to date. (96) These processes are collectively coined cuproplasia, defined as regulated copper-dependent cell growth and proliferation. (94)

2.1.4. Catalytic Metal Sites

Catalytic metal sites are defined as those metal coordination sites that bind and activate substrates directly via coordinate covalent bonds. Catalytic metal sites are structurally diverse and encompass a remarkable breadth of catalytic power, as members of all seven major classes of enzymes as categorized by the Enzyme Commission (EC): oxidoreductases (EC 1), transferases (EC 2), hydrolases (EC 3), lyases (EC 4), isomerases (EC 5), ligases (EC 6), and translocases (EC 7). (66,98) Catalytic metal ions derive their catalytic power by the precise arrangement of donor atoms that define structure and geometry of the first and more distal coordination shells around the metal. Metals can be bound to proteins via a specialized and unique arrangement of coordination bonds, integrated into a metallocofactor, either one that is purely inorganic, e.g., in Fe–S clusters, as part of an organic chelate, e.g., heme or more generally Fe-porphyrin complexes and molybdopterin (Mo), (23) or as a more exotic organometallic cofactor, featuring metal–carbon bonds, and including cobalamin (Co) (99) and nickel pyridinium-3,5-dithiocarboxylic acid mononucleotide (Ni) and others. (100)
Comprehensive metalloprotein databases, more recently advanced by machine learning approaches, (20) and recent monographs on this topic reveal a number of general functional features. (22) Fe enzymes are more abundant in prokaryotes than eukaryotes, and generally come in three flavors: nonheme Fe enzymes, Fe–S cluster-containing enzymes and heme enzymes. (101) Fe enzymes play roles in electron transport and oxygen activation chemistry and many other types of oxidation–reduction (redox) chemistries. ZnII enzymes are nature’s preferred hydrolytic catalyst and are incorporated into more than 100 distinct structural superfamilies. This unique feature of ZnII metalloenzymes has motivated the deployment of de novo protein design to create highly efficient novel ZnII catalysts in a diversity of biomolecular scaffolds (102,103) in a way that extends far beyond zinc metalloenzymes themselves. (104,105) Fe and Zn are the most cell-abundant transition metals in most cells, rising to tenths of millimolar or higher, when simple estimates for total bacterial cell volumes are used, (106) and are required for virtually all life, with some exceptions. (107) MgII, a main group element and hard Lewis acid, is present at even higher concentrations than ZnII and Fe, generally in the single-digit millimolar range, and is an important cofactor of cellular energy metabolism. MgII is present in chlorophyll and also in enzymes, where it is deployed by nucleotide (ATP, GTP) binding proteins to facilitate nucleotide hydrolysis, e.g., in ATPases and phosphatases or transesterification reactions. (108)
Cu and Mn are typically approximately 10-fold lower than Zn and Fe, with cell-associated Ni and Co about 10-fold lower still. (109,110) There are bacteria, however, exemplified by the radiation-resistant bacterium Deinococcus radiodurans, that possess Mn quotas on par with those of Fe, in contrast with the general trend indicated above. These have been termed Mn-centric bacteria. (110) Cu is a redox-active metal that cycles between oxidized cupric CuII and cuprous CuI oxidation states, with CuI forming low coordination complexes that generally function in oxygen activation chemistry. (111,112) MnII plays many diverse roles in biology, (113) including that of Lewis acid in transformations that do not require a change in oxidation state, e.g., in phosphatases, arginase and a number of metabolic enzymes, but also in other contexts where MnIII and higher oxidation states are reached, including Mn-superoxide dismutase, some ribonucleotide reductases, and the oxygen evolving complex in plants. Ni and Co play far more specialized roles in biology. Although fewer than ten Ni-metalloenzymes are known, they punch far above their weight by producing gases in bacteria, algae and plants that drive the global carbon, nitrogen and oxygen cycles. (114) NiII, like ZnII and MnII, can perform Lewis acid chemistry, exemplified by two hydrolytic enzymes, the dinuclear NiII enzyme urease (Section 5.1.2) and the mononuclear glyoxylase-I, but also functions as a cofactor in [NiFe]-hydrogenase and Ni-superoxide dismutase, among others, where both the NiI and NiIII oxidation states must be accessible. (114,115) Co, on the other hand, functions prominently in the context of two enzymes dependent on the organometallic cofactor vitamin B12, or cobalamin, found in all kingdoms of life. (99) These are methionine synthase and methylmalonyl-CoA mutase (MCM) and closely related CoA mutases which catalyze a methyl transfer and an isomerization reaction, respectively, where the Co center cycles through CoI and CoIII oxidation states (Section 5.1.3).

2.2. The Metallostasis Set-Point Model

2.2.1. Overview

It is well established that the total concentration of metal that can be measured by inductively coupled plasma-mass spectrometry (ICP-MS) in a cell lysate or in a tissue, and often expressed in units of concentration (mol L–1) greatly exceeds, by many orders of magnitude, the metal concentration that is “bioavailable” (see below). (106,109,116,117) The implication of this is that cells and tissues possess an extraordinary overcapacity to chelate metal ions; however, this says nothing about the extent to which these bound metals contribute to a buffered “pool”, which for ZnII and CuII/CuI that form high affinity coordination complexes with proteins. may well be small. Only those metal sites that rapidly equilibrate with this pool will be sensitive to small perturbations in bioavailable metal concentration that result from cellular metal toxicity or metal restriction. (118) A collection of metallosensors (Section 3) are widely accepted to surveil this pool and are thus “tuned” to effect a regulatory response, over some 12 orders of magnitude of metal concentration; (119) these metal sensing sites are generally near the protein surface, characterized by fast metal association rates and dissociation rates that can be modulated by one or more components of the cytoplasmic “buffer”. (120,121) The chemical composition of this buffered pool is not known, although a handful of spectroscopic and analytical methods (122,123) have been developed to explore the coordination chemistry of this pool, or alternatively, identify those major buffering species in cells. Histidine is clearly implicated as an important component of this buffering pool. (119,124,125) We view this pool as coordinatively unsaturated, readily exchangeable, (85) “mobile” or “labile”, which we collectively term “bioavailable”. We note that metals in these pools are not technically “free”, which is rigorously and correctly defined only as a fully solvated ion; the term “bioavailable” as proxy for free avoids the obvious confusion of “less than one free metal atom per cell” (90,109,126) which obfuscates, rather than clarifies.

2.2.2. The Set-Point Model

A metallosensor is a single-component signal transduction system, in that inducer “sensing” and “regulation” of gene expression occurs within the same polypeptide chain. This contrasts with bacterial two-component systems (TCS) where sensing and regulation require two proteins and a post-translational modification as the signaling moiety. In a single-component system, cognate metal binding or some other modification “signals” to the DNA binding domain of metallosensor via some intramolecular connection, classically defined as an allosteric connection. (18,127,128) Allostery couples metal binding to DNA binding of transcriptional activation in metallosensors in distinct ways depending mainly upon the structural family to which they belong (Section 3). Metallosensors elicit a transcriptional outcome in response to a range of intracellular metal concentrations over which their cognate metals fluctuate in cells. (119,129) These metalloregulatory proteins function as metal-dependent derepressors, (130,131) metal-dependent corepressors, (132−134) and metal-dependent activators, (135) of transcription (Figure 6A). Coordination of the cognate metal thus drives transcriptional derepression (an ON switch mediated by allosteric inhibition of DNA operator binding by metal) (Figure 6A, top), transcriptional corepression (an OFF switch mediated by allosteric activation of DNA binding) of downstream genes (Figure 6A, middle), or transcriptional activation (an ON switch mediated by metal-dependent conformational changes in the bound DNA, converting a weak promotor to a strong promoter) (Figure 6A, bottom).

Figure 6

Figure 6. Metallostasis set-point model. (121,136,137) A) Schematic renderings of the three allosteric mechanisms used by metallosensors (gray) described here (Section 3) and their biological output (expression of the metallosensor regulon; ON or OFF). Ki (Kunbind, Kbind, and Kactivate) are indicated. The set-point for each sensor, KdS is the metal concentration at which these Ki = 1 (ΔG = 0). B) Generic thermodynamic cycle where β1 and β2 are overall ZnII binding affinities for the apoprotein dimer and protein–DNA complex. The allosteric coupling free energy, ΔGc, describes the overall ligand exchange reaction shown. (138)

High Resolution Image
Download MS PowerPoint Slide
Quantitative biochemical investigations of the metallosensor-DNA and metallosensor-metal binding equilibria allow one to determine the affinity and selectivity that a particular regulator has for its cognate inducer and the DNA operator. For a regulatory protein with two distinct ligands (metal, DNA), these equilibria can be cast as a coupled or “linked” equilibrium with four states from the perspective of the unliganded metallosensor (Figure 6B). This thermodynamic framework allows the quantitative determination of this allosteric connection, embodied in the allosteric coupling free energy, ΔGc. (18,138) The magnitude and sign of ΔGc quantifies the extent to which metal binding impacts DNA binding, either negatively (+ΔGc; Figure 6A, top) or positively (−ΔGc, Figure 6A, middle). This framework is particularly useful, for example, in the search for repressor mutants that uncouple metal and DNA binding to variable degrees (ΔΔGc), thus providing fundamental insights into the mechanism of allosteric connectivity. (139−141) Although this thermodynamic framework applies directly to transcriptional repressors, (18) it can also be applied to allosteric activation of transcription where inducer recognition does not necessarily affect DNA binding affinity but instead impacts RNA polymerase recruitment and/or DNA promoter architecture. (142,143) Here, Kct is simply substituted by kct, the ratio of sensor-dependent transcription initiation rates in the presence and absence of cognate metal (Figure 6B).
The currently accepted set-point model that incorporates this thermodynamic framework is that cells effectively buffer transition metal concentration in a bioavailable pool so as to ensure that a particular metalloprotein can capture its cognate metal under virtually any cellular condition. (146) This buffering capacity goes as the inverse order of the stability complexes formed a series of divalent metals, known as the Irving-Williams series (Figure 7), (144) and is collectively maintained by the cellular repertoire of metallosensors, each of which is “tuned” to a set-point, KdS, allowing each to modulate transcription of their regulons over a relatively narrow range of metal concentration (Figure 8A). (119,121) KdS is the cognate metal concentration in cells at which the equilibrium constants for each of the three equilibria shown (Figure 6A, Ki) are equal to 1.0, and defines the point at which fractional expression of gene expression is 0.5 (Figure 8A). Once all parameters derived from the thermodynamic cycle shown (Figure 6B) as well as the impact of the cellular buffer and any change in the number of sensor molecules per cell (in high and in low metal) for each metallosensor are considered, response curves centered on KdS that relate fractional transcription output to the intracellular availability of the cognate metal are obtained (Figure 8A). (121)

Figure 7

Figure 7. Schematic rendering of the Irving-Williams series of stability constants for 2+ metal ions of the first-row late 3d-block metals, (144,145) with the extent to which each is competitive with other metals (middle), which is inversely proportional to their general bioavailability is cells (top).

High Resolution Image
Download MS PowerPoint Slide

Figure 8

Figure 8. Metallostasis set-point model. (121,136,137) A) A plot of the fractional responsiveness (0–1; 0.5 is where Ki = 1, from Figure 6A) of each of the metallosensor proteins shown in an idealized bacterial (Salmonella) cytoplasm based on their experimentally determined set-points. (121,137) These set-points, KdS, quantitatively integrate negative or positive allosteric linkage of metal binding to DNA binding (Figure 6B) (18) as well as changes in sensor concentration with metal added to cells. Efflux regulation changes the expression of the regulated genes from 0–1 (ON), via transcriptional derepression (InrS, RcnR) or activation of transcription (CueR, ZntR) while uptake regulators induce a change in gene expression from 1–0 (OFF) (Zur, NikR, Fur, MntR and the MgII-sensing riboswitch). (150) The predicted range of bioavailable metal under unstressed conditions is indicated by the filled bar and is color-coded for each metal, and spans a fractional change in gene expression from 0.1–0.9 for a single sensor, or from the combined action of a 0.1 response for the uptake repressor and a 0.9 response for the efflux sensor. The total cell-associated metal and one experimental measurement of the exchangeable Zn (free Zn) are indicated above the plot. (151) B) The midpoint of a metallosensor response range (from Figure 6A) is proposed to define the free energy of metal binding to the buffering species, ΔGbuf, which is ΔGbuf=RTlnKdS to which each sensor is tuned (or represents the average KdS for an uptake and efflux sensor pair). The metalation status of two hypothetical proteins (P1, black circle; P2, gray circle) are overlaid onto this thermodynamic construct. The various Kd for metal binding to P1, KdP1 are indicated, where ΔGP1=RTlnKdP1. The relative magnitude of the change in free energy upon metalation, ΔΔG=ΔGP1–ΔGbuf thus defines the extent to which P1 will be bound (at equilibrium) to each metal in this idealized cytoplasm. In the hypothetical case of a zinc-specific metalloprotein, P1, only the ΔΔG of metalation by ZnII is favorable (−ΔΔG), despite the fact that CuI binds ≈104-fold more tightly. In the case of P2, metalation by MnII will be preferred over FeII despite the fact that each binds with a similar dissociation constants, KdP2. Note that under conditions of ZnII restriction (lower left); purple arrow), the buffered pool may fall precipitously, which makes the free energy of metalation, ΔΔG, positive for P1, leading to undermetalation of P1 and perhaps other components of the metalloproteome. Metal toxicity, on the other hand, moves the bioavailable pool dramatically to the right (to higher bioavailable [metal]), which makes metalation of noncognate sites by more strongly binding metals, notably ZnII and CuI, favorable (−ΔΔG), particularly when these concentrations exceed the capacity of the buffering molecules, thus approaching total metal (upper right).

High Resolution Image
Download MS PowerPoint Slide
Early work showed that the equilibrium affinity of a zinc efflux regulator for ZnII, Kd,Zn, was ≈10–11 M. (147) This implies a transcriptional response that is tuned to change in bioavailable or weakly chelated [Zn] of 10–12 M to 10–10 M (corresponding to one log on either side of KdS which covers 10–90% of a canonical transcriptional response) (Figure 8A). (116,129) In contrast, FeII is buffered to ≥10–6 M; as a result, an Fe-containing metalloprotein that binds FeII with Kd,Fe of 10–7 M and ZnII with Kd,Zn of 10–10 M–1, for example, will not be mismetalated with ZnII in the cell unless free ZnII is allowed to rise far above its buffered range (>10–11 M) in this example (Figure 8B). (148) This is the central feature of the set-point model of cellular metallostasis since the relationship between the metal dissociation constant and the bioavailable metal concentration defines the sign on the conditional free energy of metal loading; a + ΔΔG will be unfavorable and a – ΔΔG will be favorable (Figure 8B). This, in turn, is thought to dictate the metalation status of the metalloproteome. (121) This is so because metal sensor proteins minimally (in nearly all cases) regulate the expression of metal transporters that uptake or efflux a specific metal into or from the cytoplasm, respectively. These transporters are thought to function to quickly restore metallostasis in the case of an acute response, or maintain metallostasis in the event of chronic metal limitation or toxicity. (149)
2.2.2.1. Limitations of the Set-Point Model
A major strength of the metallostasis set-point model is its conceptual simplicity, which allows robust, quantitative predictions of the extent to which a single metalloprotein will be metalated with a spectrum of metals in a cell. (148) This has led to the development of the web-based metalation calculator tool, which allows a user to input known metal binding equilibrium constants for their protein of interest and calculates the degree to which that protein will metalated with a particular metal in an “idealized cytoplasm.” (146,152) The lack of an unbiased validation of the predictions of this model, however, remains both a significant limitation and a considerable analytical challenge, for two reasons. The first is that experimental measurements of the concentrations of the bioavailable pool for these transition metals (Figure 8B) remain incomplete, in particular for the “idealized” cytoplasm of a bacterial cell on which the calculator was created. The only measurement of the concentration of exchangeable ZnII in E. coli grown on a minimal growth medium using a carbonic anhydrase-based sensor is ≈20 pM, which is, gratifyingly, just inside the range predicted by the model (Figure 8A). Significant progress continues to be made here, with the development of dynamically highly responsive ratiometric fluorescent protein-based FRET biosensors optimized for eukaryotic cells and organelles, (153) alongside more recent biosensor designs that originate with other fluorescent modalities that do not require molecular oxygen for maturation. (154)
A second analytical challenge is that validation of the set-point model requires an independent determination of holistic metal occupancy of a protein(s) to compare with prediction. Two-dimensional liquid chromatography inductively coupled plasma mass spectrometry (2D-LC-ICP-MS) and polyacrylamide gel electrophoresis laser ablation (GE-LS). (155,156) ICP-MS approaches on cell lysates prepared under nondenaturing conditions remain the most popular ways to do this. (107,157−159) Both approaches, however, require that metal complexes, even those bound to weakly bound metals, not dissociate during workup. (160) 2D LC-ICP-MS holds considerable promise for biologically required metals discussed here, potentially capable of detecting a change in metalation status as a function of growth conditions, for example, but appears to be readily applied to a quantitative analysis of only the most abundant protein-metal complexes of high thermodynamic (CuI, ZnII) or kinetic stability. (157) On the one hand, a considerable strength of the 2D LC-ICP-MS workflow is to identify the (mis)metalation targets of metallodrugs containing abiological metals, which benefits from a very low signal background, as long as care is taken to stabilize what might be labile metallodrug–protein adducts. (160,161)

2.2.3. Mismetalation of Proteome Metal Sites

For the reasons discussed above, the cellular factors that control the metalation status of the proteome remain incompletely defined. Recent efforts based on thermodynamic principles alone have led to the concept that the relative free energies of metalation by cell-abundant metals (ΔΔG, Figure 6B) is the primary determinant as to whether a metalloprotein is metalated with the cognate metal or one of a number of other noncognate metal ions (Section 2.2.2). Thus, cells incorporate extraordinary measures to ensure that the relative concentrations of metals are buffered in a range that allows these relative free energies to manifest themselves (Figures 6-8). Simply put, if the metal affinity of a recipient metalloprotein is in the nM range, the bioavailable metal concentrations must be in the nM range or higher in order to significantly populate that metal site, as dictated by the law of mass action (Figure 6B).
Mismetalation of a metalloprotein occurs at a point in time when capacity of the cell to buffer a noncognate metal is exceeded. This is indeed a major antimicrobial weapon the infected vertebrate host will deploy against bacteria, particularly those with intracellular lifestyles, e.g., in the phagolysosome. (162−164) CuI and ZnII are often used this way given the fact that they are tight-binding metals, on the “competitive” side of the Irving-Williams series (Figure 7), and readily displace more weakly bound metals, including Fe. An important target of CuI toxicity is the Fe–S cluster biogenesis machinery, (165−167) as well as Fe–S clusters of mature enzymes, including dehydratases, where CuI displaces one or more Fe atoms on what are essentially solvent exposed clusters, (168) potentiating oxidative stress through Fenton chemistry. In addition, other work has stablished that Cu toxicity directly inhibits heme biosynthesis in the obligate human pathogen Neisseria gonorrheae. (169) In Staphylococcus aureus, Cu negatively modulates flux through the pentose phosphate pathway by direct binding to phosphoribosylpyrophosphate synthetase (Prs), which synthesizes phosphoribosyl diphosphate (PRPP), a key metabolite that is used to synthesize the purine and pyrimidine nucleobases and nicotinamide dinucleotides. Prs is generally thought to be a MgII-dependent enzyme, and the speculation is that CuI prevents MgII from binding. (170) While it is true that more weakly bound metals could in theory displace more tightly bound ones if their bioavailability becomes sufficiently high. as predicted by the relative free energies of metalation (Figure 8), examples of this kind of metal toxicity may well be rare.
Ionophores are membrane crossing channels that allow transition metals to pass, and are being developed as antimicrobial agents to induce CuI or ZnII poisoning; indeed, recent work with the zinc ionophore and 8-hydroxyquinoline PBT2 reveals that a strategy of ZnII intoxication may be particularly efficacious in “breaking” the resistance of A. baumannii to tetracycline class antibiotics, an important finding. (171) In other work long the same lines, it has recently been shown that the host actively leverages the metal promiscuity of the bacterially produced nicotianamine-like metallophore, staphylopine, (172,173) which is produced by Staphylococcus aureus during the response to “nutritional immunity” in an effort to capture metal from the extracellular milieu. (174) The host makes extracellular CuII bioavailable here, which is then brought into the bacterial cell as a CuII-metallophore complex, thus driving bacterial copper poisoning. (174) At least one of the targets of CuI toxicity is the glycolytic enzyme, glyceraldehyde-3-phosphate dehydrogenase, (174) which while not a metalloenzyme, has a pair conserved Cys residues in the active site that are proposed to coordinate CuI. (175) Indeed, there is abundant evidence to suggest in a wide range of the bacteria that Cu toxicity broadly impacts Zn homeostasis although the mechanisms may well vary, and perhaps to a lesser degree Mn homeostasis in a subset of organisms. (176)
There are a number of examples where bacterial Zn stress results in the mismetalation of the MnII-dependent enzymes. In S. pneumoniae, a Mn-centric bacterium, (110) ZnII is an inhibitor of MnII uptake by the ATP-binding cassette (ABC) transporter, PsaBCA, (106) where it forms a coordination complex that is very nearly isostructural with that of MnII, but fails to transport ZnII across the pneumococcal membrane, thus inducing cellular MnII starvation. (177,178) This may be a consequence of the very slow off-rate of bound ZnII in this inhibited complex, relative to the more weakly bound MnII which upon engagement of the transmembrane domain readily releases MnII into the transport channel. An intracellular feature of the same Mn–Zn cross-talk is the ZnII-dependent inhibition of the MnII-cofactored protein phosphatase PhpP, which functions in the regulation of cell division. Under conditions of MnII stress, this enzyme is hyperactive, leading to significantly elongated cells; under conditions of ZnII-supplemented MnII limitation, PhpP is inhibited, thus leading to shorted cells, analogous to the Mg–Zn crosstalk that characterizes PTP1B (Figure 4A). (179) This work in S. pneumoniae suggests a regulatory model where the bioavailable Mn:Zn ratio impacts MnII-dependent metabolic activity. Indeed, this hypothesis is consistent with subsequent studies that showed that the Mn:Zn ratio impacts the activity of the MnII-dependent phosphoglucomutase, which interconverts glucose-1-phosphoate and glucose-6-phosphate, and thus modulates the biosynthesis of the polysaccharide capsule. (180)
Recent work in Staphylococcus aureus, although not mismetalation per se, reveals that cells upregulate the expression of metal-independent paralogs (Section 4) of two obligate MnII-requiring glycolytic enzymes, phosphoglycerate mutase and fructose-1,6-biphosphate aldolase, under conditions of cellular MnII-starvation mediated by calprotectin. (5,6) In a striking illustration of mismetalation for “good”, the strategic deployment of a truly cambialistic enzyme that is equally active with MnII or FeII results in metalation of a cambialistic superoxide dismutase with Fe under conditions of host-mediated MnII restriction (Section 4.3.5). (181)

2.2.4. Undermetalation of the Metalloproteome

Nutrient metal starvation is thought to result in undermetalation of some or all of the proteome. This would result if the bioavailable pool of a particular metal is insufficient to metalate all cognate metalloproteome sites quantitatively, and is of course a prediction of the set-point metallostasis model, where the free energy of (mis)metalation is positive (+ΔΔG) (Figure 8B). The extent to which the metalloproteome is metalated, i.e., the fraction of metal sites that are occupied at any point in time, is unknown. However, robust global chemoproteomics strategies have been developed that are capable of quantifying a change in metal site occupancy, at least for those sites that incorporate cystine thiolate ligation. The experiment is based on the simple premise that metal coordinating thiol ligands will be protected from reaction with an electrophilic probe, and become far more reactive if the metal is lost from the coordination complex (Figure 9), and is a “metals-focused” extension of the general concept of activity-based protein profiling. (182) This workflow leverages a rich history of cysteine thiol-reactive probes, which are available in distinct isotopomers, e.g., “heavy” and “light” versions which allows implementation of an isotope dilution-based strategy to detect a change in reactivity for all peptides in a proteome in a single experiment; when coupled to a parallel experiment that quantifies a change in peptide abundance, a normalized change in cysteine reactivity to be measured.

Figure 9

Figure 9. Global chemoproteomics strategy to probe the status of the metalloproteome under conditions of transition metal restriction. (183,184) Both proteome samples are capped with thiol-specific iodoacetamide derivative that harbors an alkyne and then subjected to Cu-catalyzed azide–alkyne cycloaddition (CuAAC) with the a “heavy” or “light” enrichment moiety as indicated. The two lysates are mixed 50:50 and subjected to a bottom-up proteomics workflow and the “heavy”/”light” ratio quantified for all Cys-containing peptides by mass spectrometry. Nonmetalloproteins with surface-accessible cysteines will be labeled to the same extent with the “heavy” and “light” probes (tan protein, right); those metalloproteins that lose the metal under conditions of metal starvation will feature more light vs heavy counts in the MS1 spectrum (green protein, right).

High Resolution Image
Download MS PowerPoint Slide
This experiment has been used to track the maturation of Fe–S cluster-containing proteins across the E. coli proteome through the strategic use of a collection of deletion strains that are predicted to stall or interrupt this process at a specific stage, either early or later, and monitoring metalation status in each case. (184) In another application, an isotopic tandem orthogonal proteolysis-activity-based protein profiling (isoTOP-ABPP) workflow was used to show that the antibiotic holomycin, a dithiolopyrrolone, functions by inducing widespread metal depletion of the E. coli proteome (183,185) upon reduction by cytoplasmic low molecular weight thiols (186) to the dithiol form. Two enzymes among several, including fumarase A, which harbors an Fe–S cluster, and carbonic anhydrase II, which is a ZnII enzyme (Section 4.3.4), were projected to be undermetalated upon holomycin treatment. (183) Indeed, holomycin induces broad spectrum zinc and iron starvation, with striking parallels to that of the antimicrobial protein, calprotectin, providing support for the idea that transition metal-starvation is an evolutionarily conserved weapon to keep microbial predators at bay. This chemoproteomics workflow is of course blind to the specific metal being probed, which must be inferred from our accumulating knowledge of the metalloproteome. In addition, a cysteine must be present in the first metal coordination shell, with metal sites that lack thiolate ligation undetected in this assay. This has motivated in part the development histidine- and carboxylate-specific probes that can be used in parallel alongside of thiol-specific probes, which when jointly deployed, may well approach full coverage of the metalation status of the metalloproteome. (184,187,188)

3. Metal Sensors

Click to copy section linkSection link copied!

3.1. Historical Overview

3.1.1. Protein-Based Regulatory Proteins

The efficacy of metallostasis (Figure 1) relies on a panel of sensory macromolecules that are “tuned” such that they maintain a specific bioavailable pool of metals in the bacterial cytoplasm, thus ensuring the integrity of the metalloproteome (Figure 8). Historically, these sensory molecules drive either transcriptional repression or activation of the expression of genes that encode metal uptake or efflux transporters thus allowing cells to control intracellular metal levels. The first efflux regulators were identified during the characterization of metal and metalloid ions resistance genes found in bacteria, in both clinical isolates and in soil or water isolates from areas polluted with these “heavy” metal species. (189,190) These resistance determinants were often plasmid-borne or present in other mobile genetic elements, (191−193) in contrast to the previously characterized chromosomally encoded transport systems for phosphate (Pho) and potassium (K), which were regulated by two component systems. (194,195) The initial interest in these heavy metal resistance determinants was primarily to understand the remarkable chemistry of the biological transformation of toxic metals and metalloids into nontoxic or less toxic species. (189,190)
The idea that a site-specific DNA binding protein might regulate transcription of resistance genes by reversible binding of a metal first emerged by identifying potential metal binding sites in these DNA binding proteins. (196) Direct validation of this hypothesis quickly followed and led to the concept of a “metalloregulatory” protein. (135) The first metalloregulatory protein so-designated was the mercuric ion resistance regulator, MerR, from the transposons Tn501 from P. aeruginosa and Tn21 from Shigella flexneri R100 plasmid. (135,191,192) While this seminal work focused mostly on DNA recognition and the mechanism of transcriptional regulation, these authors noted the presence of a potential metal binding motif, later shown to be critical for imparting metal selectivity (vide infra). (135) Soon thereafter, quantitative DNA binding experiments were used to demonstrate that HgII triggered both a modest change in affinity and a significant remodeling of the DNA promoter structure, converting a weak promoter into a strong one (Figure 6A, bottom). (197) MerR thus became the founding member of the best characterized family of metal-responsive transcription activators (Section 3.2.3).
This discovery was followed by the identification of other transcriptional repressors that responded to arsenic ions (ArsR) (198) or zinc (SmtB) toxicity, (130,199) both of which became founding members of the first and most diverse superfamily of metal- and metalloid-responsive repressors (Section 3.2.2). Shortly thereafter, the discovery of the copper-inducible repressor of a copper ATPase (CopY) from Enterococcus hirae, (200) revealed that a metalloregulatory protein could be a member of a protein family better known to elicit a response to small organic molecules, including antibiotics (Section 3.2.5). (201) A tetracycline repressor (TetR) family member was then shown to be responsible for regulating ZnII efflux in S. pneumoniae. (202) Finally, CuI-responsive efflux regulation described initially in M. tuberculosis led to the discovery of CsoR (copper-sensitive operon repressor), (203) a founding member of the CsoR/RcnR (204) protein family that harnesses a unique DNA-binding fold consisting of a tetrameric antiparallel four-helix bundle architecture (Section 3.2.4).
The discovery and early biochemical characterization of metal efflux regulators occurred alongside the identification of the repressor of iron uptake, derived from the study of the molecular mechanism of regulation of siderophore-mediated iron assimilation. (132,205) The characterization of E. coli Fur (ferrous iron uptake repressor) as a negative regulator of the expression of siderophore-mediated iron assimilation under conditions of replete cellular Fe validated an earlier proposal that Fur was a metal (FeII) binding protein that would be “activated” to bind DNA upon metal coordination. (206) It was not until later that metal binding properties were studied in vitro which ultimately led to the discovery of myriad Fur family proteins that sense other transition metal ions (207) or oxidative stressors (208,209) (Section 3.2.6). The discovery of the founding members of the other two protein families of uptake regulators came with the study of the regulation diphtheria toxin expression in Corynebacterium diphtheriae (DtxR) (210−212) (Section 3.2.7) and nickel uptake regulation in E. coli (NikR) (213,214) (Section 3.2.8). Various members of another DNA-binding family (Rrf2), initially connected to iron uptake in rhizobia, (215) were later shown to regulate iron–sulfur cluster status or to sense diverse reactive oxygen, nitrogen and electrophile stressors (Section 3.2.9). (216) In the case of Rrf2, this category of metalloregulator is connected both with the regulation of iron uptake and the crucial role of metals in sensing reactive small molecules. Similarly, the observation that GntR/FadR proteins require metalation for binding organic inducers allows us to include this family as putative metalloregulators (Section 3.2.10). (217) Other uptake regulators were then described in well-characterized protein families that responded primarily to organic molecules, such as the zinc uptake regulator of S. pneumoniae (AdcR), from the multiple antibiotic resistance regulator superfamily (MarR) (218,219) (Section 3.2.11), zinc efflux activator of S. pneumoniae (SczA), from the tetracycline repressor family (TetR) (202,220) (Section 3.2.12), and the molybdenum uptake regulator of E. coli (ModE) from the lysine uptake regulator family (LysR) (221) (Section 3.2.13).

3.1.2. RNA-Based Regulation of Metallostasis

Although single pairs of protein-based efflux and uptake metallosensors are sometimes sufficient to control cellular metallostasis (Figures 6 and 8), RNA-based regulatory mechanisms continue to emerge that generally allow bacteria to expand or otherwise fine-tune the range of bioavailable metal (Section 2.2.2). The first example of this was the discovery of a small trans-acting regulatory RNA, E. coli RyhB, that significantly extends Fur-based regulation of metal starvation responses by downregulating translation of mRNA transcripts of nonessential iron-containing proteins via an imperfect base pairing model. (222,223) This canonical form of “Fe sparing” makes Fur in many organisms a global regulator.
While these RyhB-like sRNA molecules clearly function in the regulation of metallostasis, these “off”-switches do not recognize a metal ion directly, and their cellular specificity derives entirely from the metal specificity of metalloregulatory protein that controls their expression. RNA-based regulation of metallostasis expanded significantly with the discovery of metal-sensing riboswitches, termed metalloriboswitches. (150) These generally (but not always) function as “on”-switches under conditions of metal toxicity and are highly structured regulatory RNA segments found within the 5′ noncoding or untranslated region (UTR) of mRNAs that often encode metal transporters. The binding of a metal (or a metal complex) to the riboswitch aptamer alters the folding of the RNA such that transcriptional “stop” and/or translational “off” signals are disrupted, thus leading to expression of the immediately adjacent gene (an “on” switch). These riboswitch elements are considered cis-acting RNAs, since they regulate transcription or translation within the same mRNA. While the first metalloriboswitch was a MgII sensor that turned “off” the expression of a MgII uptake transporter, (150) subsequent work showed that metalloriboswitches have evolved to sense Ni/Co, (224) Mn, (225,226) and other metallocofactors such as vitamin B12. (227) Only recently was it shown that Ni/Co riboswitches respond primarily to iron bioavailability in cells. (228) This observation reinforces the idea that numerous other metals or metal complexes might be monitored in cells by harnessing the chemical selectivity of as yet undiscovered or “orphan” riboswitches to prevent the toxic accumulation of a specific analyte. (229) Although evolved to sense metals or metal-complexes directly, the set-point or sensitivity of these riboswitches to bioavailable metal may not fully align with the set-point of the analogous protein-based sensors given the complexity of the regulatory response involved (Section 3.3).

3.1.3. Biotechnological Applications

Finally, we discuss how the selectivity of some of these molecular switches has been leveraged in downstream synthetic biology and biotechnology applications, as biosensors (Section 3.4). The key component of any biosensor is the sensory module itself, tasked with selective detection of a target analyte, which is coupled to a reporter module that produces a detectable output. Among the myriad of possible sensory “parts”, the metalloregulators discussed here have emerged as the “gold standard” for metal detection, due mainly to their metal sensitivity and selectivity. (230−232) Historically, the first applications of metalloregulators as biosensors were based on engineering living bacteria, where a regulated promoter was positioned upstream of a reporter gene, with strategic modifications of the bacterial chassis allowing for the desired intracellular bioaccumulation and sensitivity. (233−235) Later, this simple genetic circuity was adapted to cell-free platforms, which permits the development of field-deployable devices. (19) Synthetic biology continues to push the engineering and discovery of novel transcription factors and metalloregulator proteins in particular, enhancing both our fundamental knowledge of these molecular switches and application of metallosensors as biotechnologically useful tools. (236)

3.2. Families of Metalloregulatory Proteins

3.2.1. Overview

Metallosensors can be functionally subdivided into two broad groups of regulatory proteins. One group mediates the up-regulation of metal efflux systems upon metal binding in response to metal toxicity generally via transcriptional derepression or activation mechanisms. The second group consists of metalloregulators that down-regulate metal uptake systems via transcriptional corepression in response to cognate metal binding (Figure 6). The affinity for different metals generally follows the Irving–Williams series of stability constants for divalent metal ions (Figure 7). (144) Metal responsiveness is dictated by the set-point, KdS (Figure 8). the ligand set and metal coordination number and geometry, where formation of the “right” coordination geometry upon metal binding is a key feature of metal-mediated allosteric inhibition or activation of DNA operator binding. (120,141,237−240)
Among the more than 20 structural classes of “one-component” (Section 2.2.2) transcriptional regulatory systems in prokaryotes, no fewer than 12 of these structural families harbor metallosensors. Most metallosensors are “helix-turn-helix” or “winged helix” superfamily proteins proposed to be monophyletic, with the prominent exception of the CsoR protein family, proposed to have evolved from a four-helix-bundle RNA-binding protein. (203,241,242) In addition, most of these 12 structural classes also harbor representatives that control the adaptive response to cellular reactive oxygen species (ROS), e.g., superoxide and peroxides, reactive nitrogen species (RNS), e.g., GSNO and NOx, reactive electrophile species (RES), e.g., formaldehyde, quinones and the recently described reactive sulfur species (RSS), e.g., thiol persulfides or more generally “supersulfides.” (243,244)
We review how sequence diversity in each structural family can be represented by clusters of closely related sequences (239,245,246) and interpreted in terms of a conserved metal binding or sensory motif, thereby defining the cognate inducer with greater confidence. Specific characteristics of the functions of genes that colocalize with a metallosensor gene (genomic neighborhood) can further improve the robustness of these predictions. This is because nature has employed a “mix-and-match” approach in the evolution of metal-sensing operons or regulons in a specific organism, likely aided by horizontal gene transfer (247) and subsequent convergent evolution. (248) This implies that a sensor derived from a particular structural family may have evolved to sense a distinct inducer (metal or reactive molecule), dictated by the specific characteristics, e.g., a specific range in metal bioavailability (Figure 8) or type of small molecule stressor specific to that organism or “chassis” encoding it. Alternatively, a collection of sensors from distinct structural families may have evolved to regulate the same gene in different organisms. For example, M. tuberculosis, E. coli and E. hirae encode essentially the same CuI-specific P1B-1-type ATPase efflux transporter (Section 5.3) whose CuI-inducible transcription is regulated by three structurally distinct CuI-sensing metallosensors, CsoR, (203) CueR (a MerR family member), (249) and CopY, (200) respectively. The common feature is that each CuI sensor employs a low-coordination number Cys-rich CuI coordination complex with a characteristic KdS (Figure 8A), but which triggers very different mechanisms of metalloregulation of transcription (Figure 6A).
On the other hand, crosstalk and mismetalation (Section 2.2.3) also play critical functions in metalloregulators, which brings into question the very idea that each regulator has evolved to sense a single cognate metal. (250) For example, metal binding to S. pneumoniae CopY elicits a distinct transcriptional response dependent upon the bound metal; here, ZnII acts as a corepressor and CuI functions as an allosteric inhibitor of DNA binding (Figure 6A) allowing the expression of CuI-specific P1B-1-type ATPase effluxer specifically under conditions of CuI toxicity. (251) In Salmonella, acute ZnII toxicity transcriptionally induces a CoII toxicity response orchestrated by the Co/Ni sensor, RcnR, while conversely, acute CoII toxicity triggers a ZnII toxicity response regulated by the ZnII efflux activator ZntR, leading also to repression of the Zur regulon. (121) Mis-metalation of Fur by MnII under conditions of MnII toxicity, in contrast, is detrimental to cell viability, and is avoided by proper tuning of the coordinated responses of the MnII-sensing repressor and Fur to distinct ranges of cellular metal activity (Figure 8A). (252) Here we review the impact of metal promiscuity and crosstalk in the context of structural and sequence determinants of metal specificity and allosteric connectivity for four classes of metal efflux regulators (ArsR, MerR, CopY, and CsoR), four families of uptake regulators (Fur, DtxR, NikR, and Rrf2) and finally, four families within which metallosensors are significantly under-represented (TetR, MarR, LysR and GntR).

3.2.2. ArsR

Members of the arsenic repressor (ArsR) superfamily sense a wide range of stimuli (Figure 10A) and feature a compact homodimeric “winged helical” DNA binding fold, characterized by a core α1-α2-α3-α4-β1-β2-α5 secondary structure, where the dimer is formed by the quartet of α2 and α5 helices from each subunit (Figure 10B). Some ArsR proteins harbor extensions on either or both N- and C-terminal sides of this core motif, generally denoted the α0 and α6 helices, respectively. (120,253) The helix-turn-helix (HTH) motif that mediates DNA-binding is located at the α3-α4 segment that engages successive DNA major grooves, with a degenerate tetrapeptide sequence in α4 that enforces specificity for a particular DNA operator sequence. (254) The β1-β2 wing extends from the periphery of the dimer and may mediate interactions with the immediately adjacent “outer” minor grooves. Early work on ArsR family proteins suggested that the >3000 distinct members of this family were mostly metal ion or metalloid-specific regulators. (255−257) This is clearly no longer the case. (120,247) A number of ArsRs have been recently characterized and shown to respond to redox-active small molecules (258−260) or lack inducer binding sites altogether (Figure 10A). (261−263) ArsR superfamily sequence relationships were recently evaluated using a sequence similarity network (SSN) analysis of ≈168,000 unique entries from the Pfam and Interpro data sets. (245,264)

Figure 10

Figure 10. ArsR family proteins. A) Abbreviated periodic table showing the metals and reactive species (shaded) known to be sensed by ArsR proteins. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of SaCzrA protein (PDB 2M30) (265) with a close-up view of the ZnII binding site. An additional close-up shows the coordination of CdII present in SaCadC. (266) Conserved primary structural features of different metalloregulator ArsR proteins with the different metal binding sites indicated in the text. C) Ribbon representation of CgArsR (6J0E) (248) with close-up view of the of the arsenic binding site. Conserved primary structural features of different arsenic-binding proteins (EcArsR, CgArsR, and AfArsR). D) Ribbon representation of persulfide (RSSH) sensing RcSqrR (6O8N) (267) with a close-up view of the polysulfide bridge formed upon reaction of the dithiol site with a thiol persulfide. Conserved primary structural features of different persulfide (RSSH) sensing ArsR. E) Ribbon representation of NoRexT (7TXN) (258) with a close-up view of the H2O2 binding site. Conserved primary structural features of different RexT homologues are shown.

High Resolution Image
Download MS PowerPoint Slide
3.2.2.1. Metal-Sensing ArsRs
The largest cluster of sequences (245,264) contains the vast majority of described metalloregulators to date, that sense either biologically relevant transition metals, e.g., CuI, ZnII, NiII or CoII and toxic heavy metals, including CdII or PbII. Collectively, these sequences share a conserved metal binding site conforming to a DxHx10Hx2(H/E) motif in the α5 helix that defines an interfacial site between the two subunits of the dimer, and an additional metal binding site harboring a CxC motif in the α3 helix (Figure 10B). Interestingly, the ZnII sensors SaCzrA, (131) BsCzrA, (255) SynZiaR (268) and TtSmtB (130) that sense ZnII with significant specificity both in vivo and In vitro, (269) share the same DxHx10Hx2(H/E) motif with the NiII sensor MtbNmtR; however, NmtR completes an octahedral coordination complex by recruiting the N-terminal His-Gly sequence. (238,270) On the other hand, CdII and PbII responsive regulators LmCadC (271) and SaCadC (272,273) rely largely upon the α3 CxC motif and two N-terminal Cys as the primary allosteric site, denoted α3N, for negative regulation of DNA binding (Figure 6A, top) since the α5 DxHx10Hx2(H/E) motif is either not present (274) or allosterically uncoupled (silent). (266,275) While this Cys-rich site retains a certain degree of selectivity for soft transition metal ions CdII and PbII over ZnII, the In vitro metal selectivity is not as high as the paradigm PbII sensors from the MerR superfamily (vide infra). (269) Other characterized ArsR family metallosensors that contain allosterically active α3N and α5 sites are known to have relaxed metal selectivity. One example here is cyanobacterial BmxR, whose regulon is inducible by both monovalent (CuI/AgI) and divalent (ZnII/CdII) ions in O. brevis. (276) Additionally, cyanobacterial AztR (277) responds to ZnII/PbII/CdII by engaging metal donors in a modified α3N site only, while mycobacterial KmtR (278) responds to ZnII/CoII/NiII by binding metal to a His-rich site that resembles, but is not identical to, the α5 DxHx10Hx2(H/E) motif found in a canonical ZnII sensor.
3.2.2.2. Other Non-α5 or α3N Metal-Sensing ArsRs
There are a number of other metal-sensing repressors from the ArsR family that differ from those described above. The regulon of the ZnII-responsive efflux repressor mycobacterial SmtB includes zur (Section 3.2.6) which suggests a mutually reinforcing response to zinc toxicity (Section 5.3.3). (279,280) The CdII sensor CmtR from Mycobacterium tuberculosis and Streptomyces coelicolor employ a distinct metal binding site containing two metal coordinating cysteines derived from the DNA binding helix α4 and a Cys from the C-terminal flexible tail (Figure 10C). (281,282) The NiII sensor SrnR from Streptomyces griseus (263) is representative of a group of candidate NiII-specific regulators that depend on a NiII-binding partner protein (SrnQ) to elicit a transcriptional response. (283,284) The HgII resistance regulator from the Gram-positive bacterium Streptomyces lividans (285) is present in a divergently transcribed operon that has been replaced by a canonical MerR family regulator in other bacteria. (286) This sensor has more than six cysteine residues in different structural elements and several other putative metal binding sites; however, the bioinorganic chemistry of this system has not yet been examined.
3.2.2.3. AsIII and Organoarsenical Sensors
The canonical AsIII sensors harbor a C-(V/A)-C motif in the α3 helix coincident to that described above, which donates two of the three AsIII ligands, adopting a trigonal planar coordination geometry (Figure 10D). This is the dominant AsIII-sensing regulator, with an additional Cys present in either close proximity to this motif (287) or in the C-terminal region. (255) Two distinct AsIII sensing sites described for AsIII sensors are represented by Acidithiobacillus ferrooxidans ArsR (288) and Corynebacterium glutamicum ArsR (Figure 10D). (289) AfArsR and another ArsR in Agrobacterium tumefaciens (290,291) feature a CCx5C motif in the α5 helix that defines a C-terminal AsIII binding site. (292,293) The AsIII selectivity over HgII of this protein has been recently interrogated with a combination of different spectroscopic techniques that confirm that AsIII binds in a rigid trigonal pyramidal geometry compared to an rather undefined geometry with HgII. (294) On the other hand, the CgArsR AsIII binding site engages a Cys from the canonical C-(V/A)-C motif and two N-terminal Cys from the other subunit within the dimer (Figure 10D). (289) The common feature is that each of these three AsIII sensing ArsRs utilizes three cysteine residues at spatially distinct locations within the same three-dimensional structure, suggesting that the ArsRs must have arisen independently during evolution. (266,288,289) It is important to note that while the coordination of AsIII by the canonical As-sensing ArsR has only been characterized by EXAFS, (295) the other two AsIII sites have been resolved by crystallography. (248)
3.2.2.4. Redox-Active Small Molecule ArsR Sensors
The remainder of the biochemically characterized ArsR family sensors are not metal ion or metalloid-sensing regulators, but instead use cysteine thiolate chemistry enabled by a precisely tuned pKa, nucleophilicity and topology to specifically sense either RSS or ROS (Figure 10E). (245,267) All known RSS or supersulfide sensors are characterized by a pair of cysteines in the α2 and α5 helices which form an intraprotomer polysulfide bridge when incubated with a sulfane sulfur (sulfur-bonded sulfur) donor. (245,267,296,297) The paradigmatic RSS-sensing ArsR is SqrR from the purple bacterium Rhodobacter capsulatus. (259) X-ray crystal structures of four distinct oxidation states reveal that the two sensing Cys lie quite far apart from one other, forming a cavity that restricts solvent accessibility and disulfide bond formation, which would occur with an oxygen-derived ROS. (267) This in turn promotes the formation of a polysulfide bridge linking the two cysteines. RSS-specificity was shown not only in vitro but also in cell culture, (298) as in the case of Vibrio cholerae HlyU, an SqrR homologue, which regulates the expression of exotoxins necessary for bacterial colonization of the gut in response to endogenously generated persulfides. (245) SqrR homologues have also been found in other human pathogens, (296,299) plant pathogens (260) and sulfur-oxidizing bacteria. (300) On the other hand, cyanobacterial RexT is H2O2-specific sensor, which reacts instead with two vicinal Cys residues in the α3 helix (Figure 10F). (258) The formation of disulfide bridges between the vicinal Cys residues is the mechanism by which transcription of downstream H2O2-detoxifying genes is allosterically regulated. It remains unclear, however, if crosstalk with other oxidants occurs in RexT. Overall, nonmetal sensing ArsR proteins appear to exhibit exquisite chemical inducer specificity afforded by the distinct structural environment of the dithiol pair.
3.2.2.5. Allostery
The picture that emerges from a gallery of different ArsR family structures is that the sensing sites are widely distributed throughout this simple molecular scaffold, which in turns triggers a change in structure and/or conformational dynamics that inhibits DNA binding by several orders of magnitude (ΔGc ≈+3–6 kcal/mol; Figure 6A-B). (265−267,270,301) This large allosteric coupling free energy comes from distinct metal, metalloid and reactive species-specific sensing sites, which implies that each of these sites must be connected in some way to the DNA-binding helices to drive allosteric inhibition of DNA binding. This remarkable feature is unique to ArsR family sensors among all of the metalloregulatory protein families discussed here. The allosteric plasticity of this simple molecular scaffold has obvious implications for evolvability of new biological outcomes, which in one case, the ZnII sensor SaCzrA, appears to arise from a redistribution of internal dynamics important for DNA operator binding. (139) The extent to which this allosteric mechanism is a general feature of ArsR family regulators is unknown, but we believe likely, given the sheer number of structurally distinct sensing sites that have emerged on this simple homodimeric scaffold. (139)

3.2.3. MerR

MerR proteins, like ArsR proteins, sense a wide range of metal and chemical inducers (Figure 11A). They are homodimers organized as two symmetry-related signaling and DNA-binding domains linked by a long α-helical coiled coil derived from the α5 helix (Figure 11B). (302,303) Each DNA-binding domain is formed by the N-terminal region of each protomer which conforms to a winged helix-turn-helix motif with an α1-α2-β1-β2-α3-α4 topology (Figure 11B). The inducer binding signaling domain typically involves residues from the C-terminal region of one protomer and the N-terminal region of the α5 helix of the other protomer. (304−306) The diversity of signals collectively detected by MerR sensors may well be broader than that of the ArsR family as they not only form coordination complexes with monovalent and divalent metal ions, e.g., mercury (Hg), (135,307,308) cadmium (Cd), (309,310) lead, (237,311) cobalt, (312) zinc, (313) iron, (314) nickel, (315) gold (AuI) (304,316) and copper, (317) but also respond to oxidative stressors, (305,318) carbonyl and nitrosative stressors, (319,320) and lipophilic drug molecules (306,321) (Figure 11A). In general, the affinity of MerR sensors for their target operators is not significantly changed by binding the cognate inducer, i.e., the apoprotein may function as a weak repressor of a weak promoter. On the contrary, inducer binding drives a concerted change in both the conformation of the homodimer and DNA promoter to which it is bound, eliciting a transcriptional activation response via recruitment of RNA polymerase to what is now a strong promoter. (253,321,322) Canonical MerR activators bind to their DNA operators embedded in a suboptimal promoter with the −35 and −10 promoter elements separated by 19–20 bp instead of the optimal ≈17 bp distance. (197,313) Signal perception is linked to DNA underwinding, thus moving the −35 and −10 promoter elements to same side of the DNA duplex and allowing productive engagement of RNA polymerase, and transcription initiation (Figure 6A, bottom).

Figure 11

Figure 11. MerR family proteins. A) Abbreviated periodic table showing the metals and reactive species (shaded) known to be sensed by MerR proteins. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of CadR protein (6JGX) (309) with close up view of the Cd metal binding sites, denoted site 1 and site 2. Conserved primary structural features of different metalloregulator MerR proteins with metal binding sites with Cd binding preference. C) Ribbon representation RmPbrR (5GPE) (237) with a close-up view of the of the PbII binding site. Conserved primary structural features of different MerR proteins with Pb and Cd metal binding preference. D) Ribbon representation EcZntrR (1Q08) (317) with a close-up view of the of the Zn binding site showing two Zn metal ions. Conserved primary structural features of different MerR proteins with Zn metal binding preference. E) Ribbon representation PaMerR (5CRL) (308) with a close-up view of the Hg binding site. Conserved primary structural features of different MerR proteins with Hg metal binding preference. F) Ribbon representation EcCueR (PDB 1Q05) (317) an activation complex with operator DNA with close up view of the of the CuI binding site. Conserved primary structural features of different MerR proteins with CuI metal binding preference. G) Ribbon representation of the peroxide sensing ExSoxR (2ZHG) (305) in complex with operator DNA with close up view of the [2Fe-2S] site. Conserved primary structural features of different ROS sensitive MerR proteins.

High Resolution Image
Download MS PowerPoint Slide
3.2.3.1. PbII- and CdII-Sensing MerRs
Early studies of a number of canonical metal-sensing MerRs revealed that a major determinant of metal selectivity was the specific arrangement of residues (usually Cys or His) in the inducer-binding sites largely localized to a metal binding loop (MBL), (323) with a preferred coordination geometry for a monovalent vs a divalent metal ion. (149) Insights that have emerged more recently suggest a far more nuanced picture. For example, while most MerRs contain a single metal binding site in the MBL, the CdII efflux regulator from Pseudomonas putida CadR features two CdII binding sites, as inferred from the crystal structure of the different metalated forms (Figure 11B). (309) One site is formed by Cys in the MBL and a Cys from the N-terminal region of the dimerization (α5) helix, an arrangement shared with other divalent metal ion MerR sensors, with the second site a histidine-rich site positioned between the dimerization helix, the H-T-H motif and the C-terminal tail (Figure 11B). (309) CdII binding to the MBL site facilitates the assembly of His-rich site and the presence of this second site in turn enhances the metal affinity of the MBL site significantly, i.e., the binding of CdII in CadR is positively cooperative. CadR is the most specific, naturally occurring CdII-sensor reported to date, as the cadR promoter is induced only by cadmium, and this may be traced to this two-site intraprotomer cooperativity. (324) Other CdII-sensing proteins, such as CadR from P. aeruginosa and E. coli, (310) do not conserve this second CdII site, and instead rely exclusively on a P. putida-like Cys-thiolate-rich MBL site for regulation.
The PbII sensor PbrR from Cupriavidus metallidurans is representative of a group of MerRs that form a trigonal-pyramidal coordination complex with three conserved Cys, with two derived from the MBL and one from a Cys in the N-terminal region of the dimerization helix (Figure 11C). (311) The crystal structure of PbrR suggests that the PbII-selectivity of this switch may be traced to the coordination geometry itself, where fourth coordination position is occupied by the stereochemically active lone pair of electrons on the PbII ion. CmPbrR, like CdII-sensing CadRs, harbors the same asparagine residue (N81; Figure 11B) that functions as the fourth metal ligand in CadRs, (309,310) but is excluded from the first coordination shell in CmPbrR by the PbII lone pair. These characteristics precisely parallel previous findings for the ArsR family PbII/CdII sensor CadC. (325,326) It is interesting to note that although this three-coordinate PbII-pyramidyl structure may well be a common feature of PbII-responsive regulators, the presence of four Cys residues in the metal binding site of CadC negatively impacts PbII specificity over CdII or ZnII.
3.2.3.2. MerR Family ZnII Sensor
The canonical ZnII efflux regulator (Figure 6) EcZntR is characterized by a conserved CCx3Hx4C motif in the MBL and a Cys in the N-terminal region of the α5 helix, which collectively create a binuclear ZnII cluster in which the two ZnII ions adopt approximately tetrahedral coordination geometries (Figure 11D). (317) EcZntR responds to intracellular spikes of ZnII in the nanomolar concentration range (327) but also responds to acute phase CdII and PbII toxicity as well. (328) This suggests that the zinc selectivity of ZntR may be quite modest in E. coli. (329) In Salmonella, however, ZntR appears quite selective, at least relative to other biologically required metals, MnII, FeII, CoII, NiII and CuI. (121) ZntR plays a key role specifically in Brucella abortus ZnII homeostasis. (330) On the other hand, A. tumefaciens ZntR responds primarily to CdII toxicity, eliciting a comparable response to ZnII and CoII, (331) while in Serratia marcescens, ZntR appears to regulate a PbII toxicity response. (332) Thus, the true cytoplasmic metal inducer of a ZntR may well be dependent upon the composition of the intracellular “buffering” milieu, which thus impacts its set-point KdS for a specific metal (Figure 8), perhaps in response to its environmental niche in which an organism resides. This degree of metal-sensing promiscuity across a collection of closely related ZntRs in a cell is not only anticipated but may well be beneficial, since it is well documented that the P1B-2 ATPase, ZntA, can efficiently transport a range of divalent soft transition metal ions, including ZnII, CdII, PbII and HgII by adopting a range of coordination geometries in the transmembrane metal binding site (Section 5.3.1). (250) In fact, this modest selectivity of EcZntR has been exploited in the selection of engineered ZntR mutants capable of detecting xenobiotics (329,333,334) that are subsequently leveraged in the design of whole cell biosensors (Section 3.4).
3.2.3.3. HgII-Selective MerRs
The founding member of the MerR superfamily is plasmid-derived EcMerR (135) and all share three conserved cysteines, two in the MBL and one in the N-terminal region of the α5 dimerization helix (Figure 11E). The HgII site is trigonal planar as revealed by 199Hg NMR spectroscopy (335,336) and EXAFS, (337) and later confirmed by X-ray crystallography of both Gram negative (308) and Gram positive (307) HgII-sensing MerRs. This tricoordinate HgS3 site in MerR remains the only naturally occurring HgII complex of its kind, beyond de novo designed three-helix bundle peptides. (338) The selectivity and sensitivity of this HgII specific switch is noteworthy, with ≈100-fold discrimination of HgII over other metal ions. (197,339,340) This property may well be traced to several His and Cys residues beyond the first coordination shell that have been implicated in driving a significant structural rearrangement in the homodimer which buries the tris-thiolate HgII site. (308) However, these ideas are difficult to rationalize on the basis of early extensive mutagenesis studies. (341) In any case, directed evolution of HgII-sensing MerR to create a CdII-specific sensor has been achieved through the introduction of mutations distributed throughout the molecular scaffold. (342)
3.2.3.4. The Monovalent “Coinage Metal” Sensors
The coinage metal sensors of the MerR family are exemplified by the CuI sensor E. coli CueR, which provides resistance against copper toxicity, and the gold resistance regulator from Salmonella, GolS (Figure 11F). CuI coordination by CueR involves formation of a linear bis-thiolato coordination complex formed by two Cys from the MBL. Selectivity for CuI is mediated by a negative design element, a conserved Ser in the α5 helix, required to exclude divalent metal ions, (317,343,344) as well as the somewhat shorter length of the MBL relative to divalent metal sensors (Figure 11F). (317) For example, the number of residues between the two Cys ligands in the MBL appear to impact discrimination between HgII and CuI. (307) GolS binds AuI with the same coordination geometry as CuI-sensing CueR (316) and can be phylogenetically distinguished from CueR only on the basis of differences in the N-terminal region that appears to confer AuI specificity (316) while modulating regulator-DNA operator specificity. (345,346) Thus, it is not surprising that EcCueR is responsive to CuI but can also detect AgI and AuI in vivo and in vitro. (302,317,339)
3.2.3.5. Unconventional Metal-Sensing MerR Proteins
There are several MerR family regulators that do not appear to conform precisely to this classical paradigm. (303) For example, Synechocystis CoaR harbors an additional domain that is required for CoaR to respond to intermediates of the B12 biosynthetic pathway. (312) MliR from the marine Bacteroidetes Bizionia argentinensis has been reported to function as both an iron-dependent activator and repressor of genes linked to bacterial survival under Fe-replete conditions so as to avoid Fe-overload. (314) Similarly, the Haemophilus influenza nickel-responsive MerR-like regulator (NimR) is the only described MerR that responds to NiII and regulates the expression of genes encoding a NiII uptake system. (315) This is in contrast to most MerR regulators, which generally elicit a detoxification response to a specific inducer.
3.2.3.6. Redox-Sensing MerRs
Three phylogenetically distinct groups of MerR sensors respond to redox-active small molecules. The superoxide sensor, E. coli SoxR, is among the best characterized of these subfamilies. (347−349) Each SoxR protomer harbors a CIGCGCLSxxC motif in the MBL that coordinates a sensory [2Fe–2S] cluster and mediates a response to superoxide anion, nitric oxide and redox cycling agents (Figure 11G). (305) Oxidation or nitrosation of the Fe–S centers induces a conformational change in the homodimer that renders SoxR-regulated promoters permissive for open complex formation in front of its only target gene, soxS, which encodes a global redox resistance regulator. It is interesting to note that in contrast to most other metal-binding MerRs the [2Fe-2S] clusters are completely solvent exposed, (305) a finding consistent with a role in cellular monitoring of small redox-active compounds. SoxR is an excellent illustration of evolutionary “plasticity” of the MBL in MerR proteins and how that plasticity can be tuned to perform a distinct function. Unlike E. coli and related enteric bacteria, however, the majority of SoxR regulons in P. aeruginosa and Streptomyces coelicolor lack the genes typically involved in O2– detoxification. (350,351) Thus, it is not yet clear that all SoxRs are superoxide anion sensors, which suggests that other redox active small molecules could have a more significant role to play in these organisms.
NmlR, first identified in Neisseria gonorrheae, (352) also activates transcription in response to both oxidative and nitrosative stresses, with the structure and mechanism of regulation described for the H. influenza homologue. (319) Oxidation of a conserved Cys pair in the MBL changes the protein conformation allowing transcriptional derepression and/or activation of the expression of a genes encoding alcohol dehydrogenase family enzymes. (319) A recent report suggests Streptococcus pneumoniae could also harness this chemistry to respond to strong electrophiles such as HClO. (353) AdhR, on the other hand, is a Bacillus subtilis aldehyde-activated sensor that is representative of the third group of redox active regulators that regulates the expression of genes encoding enzymes that reduce aldehydes or repair aldehyde derived posttranslational modifications. (354) More recently, it has been shown that polysulfane compounds induce AdhR-regulated genes. (355) Altogether these observations suggest that AdhRs and NmlR proteins may function as promiscuous redox or electrophile sensors in cells.

3.2.4. CsoR

CsoR family proteins exhibit high pairwise sequence similarities and adopt a disc-shaped, dimer-of-dimers D2-symmetric tetrameric architecture consisting of four α-helical ≈90-residue protomers (Figure 12B). (356) Members of the CsoR family collectively detect a wide range of transition metal- and small molecule inducers, with an inducer footprint that is at present far smaller than that of the ArsR and MerR families (Figure 12A). The founding CsoR family members are the CuI sensor (357) from Mycobacterium tuberculosis (203) and the NiII/CoII sensor RcnR from E. coli. (358) Since that time, other CsoR family members have been reported, including a second mycobacterial CuI sensor, RicR, (359) the NiII sensor InrS from Synechocystis, (360) the formaldehyde-sensing repressor FrmR from E. coli, (361,362) and the thiol persulfide sensor CstR, now functionally characterized in S. pneumoniae, (239) S. mitis, (239) Staphylococcus aureus (363) and Enterococcus faecalis. (364) Inducer specificity by CsoR-family repressors was originally described in the context of a simple “W-X-Y-Z” motif using the MtCsoR CuI chelate as a template, (203) where the first coordination shell ligands around a subunit-bridging, trigonally coordinated CuI define the X (Cys36), Y (His61’) and Z (Cys65’) residue positions (Figure 12B). (204) This conceptually simple perspective on inducer specificity has proven remarkably resilient, but cannot fully explain how inducer promiscuity is avoided, particularly when considering specificity toward reactive species such as persulfides (239) and formaldehyde. (362) A recently published SSN analysis significantly extends these initial ideas, providing new insights into inducer specificity in this very large superfamily of transcriptional repressors. (239)

Figure 12

Figure 12. CsoR family proteins. A) Abbreviated periodic table showing the metals and reactive species (shaded) known to be sensed by CsoR proteins. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of MtCsoR tetramer protein (2HH7) (203) with close up view of the Cu metal binding site. Conserved primary structural features of CsoR homologues proteins denoted in the text. C) Ribbon representation of the formaldehyde sensitive SeFrmR protein (5LCY) (362) with close up view of the metal binding site. Conserved primary structural features of FmrR homologues. D) Ribbon representation of the nickel activated SyInrS protein (5FMN) (365) with a close-up view of the metal binding site. Conserved primary structural features of InrS proteins are shown. E) Ribbon representation of the persulfide (RSSH) sensitive SpCstR protein (7MQ2) (239) with close-up view of the reduced cysteines of the RSS-reactive site. Conserved primary structural features of different CstR like proteins.

High Resolution Image
Download MS PowerPoint Slide
3.2.4.1. CuI Sensors
There are three distinct types of CuI-sensing CsoR-family repressors exemplified by BsCsoR, (366) MtCsoR, (203) and TtCsoR (Figure 12B). (367) The reported SSN (239) coupled with a phylogenetic analysis suggests that copper specificity within this family of repressors may have arisen multiple times during evolution. (366) Copper specificity has generally been assessed by determining an affinity ≥10–18 M–1 using CuI specific chelators as competitor ligands, (203,357,368−370) although the selectivity for CuI over other transition metal ions has been determined in only for a few cases. (203,371) Nearly all functionally characterized bona fide CsoR-like CuI sensors are related to BsCsoR (357,359,370) which conserve the three cognate CuI binding residues that form a C–H–C (Cys-His-Cys) “X-Y-Z” motif and lack an N-terminal metal binding residue “W”, thus denoted x–C–H-C. Two structures of this group are SlCsoR (357,372) and GtCsoR, (366) obtained in the apo and CuI-bound states, respectively. A comparison between the apo- and CuI-bound states suggests a mechanism of allosteric connectivity where the metal binding event induces a kink in the α2 helix, which leads to a rearrangement of surface electrostatic potential distribution, thus driving DNA dissociation and transcriptional derepression of CuI-resistance genes. (372,373)
The x–C–H-C motif is also present in MtCsoR which exhibits high specificity for CuI. (203) MtCsoR and the related sequences only present in actinomycetes are characterized by a ≈30-residue, poorly conserved C-terminal tail not found in other CsoRs. (203,374) The only structure available for this subgroup is the CuI-bound form of MtCsoR. (203) While structural details of the apo form are missing for this groups of CuI sensors, ion mobility mass spectrometry studies on this protein suggest that metal binding subtly alters the shape of the tetramer to a more hydrodynamically compact structure (375) which may lead to changes in the surface electrostatic potential distribution similar to that observed for SlCsoR. (372,373) Moreover, the ε-face of CuI-coordinated His ligand interacts with residues in the second coordination shell including nearly invariant Tyr and Glu residues that are energetically linked to CuI-mediated allosteric inhibition of DNA binding, suggesting a path for allosteric connectivity. (203)
In T. thermophilus CsoR one of the CuI-coordinating Cys is replaced by a His, thus defining a x–C–H-H metal-binding motif (Figure 12B). (367) Although the CuI specificity of this switch has not yet been unequivocally evaluated, residues in the second coordination shell are conserved, as are the neighboring genes found in all three CuI-specific clusters (a P1B-1 CopA efflux pump and a CopZ metallochaperone); these features are consistent with a CuI-specific sensing function. (239) Studies designed to examine the impact of this Cys-to-His substitution concluded that the histidine positively impacts DNA-binding affinity and CuI-responsiveness of this sensor. (367) Metal displacement studies on different SlCsoR mutants suggest that a His in this metal-ligating position may impact the kinetics of NiII displacement of bound CuI ions. (371) Thus, substitution of the second Cys may enhance the promiscuity of this subgroup of CuI sensors.
3.2.4.2. NiII/CoII Sensors and the Formaldehyde Sensor
The NiII/CoII-sensing repressor RcnR is readily distinguished from CuI-sensing CsoRs proteins by conservation of a H–C–H-H “W-X-Y-Z” motif. (204,376) RcnRs are most closely related to the formaldehyde (HCHO)-sensing FrmR, (362,377) which harbor a P (Pro)–C-H-x “W-X-Y-Z” motif, as exemplified by EcFrmR and SeFrmR (Figure 12C). A very high level of pairwise similarity is consistent with the relative ease with which inducer specificity can be relaxed or switched in an FrmR by simple reintroduction of the Z-position His residue found in RcnRs. (362) The Ni/Co specific transcriptional response of RcnR has been extensively characterized in lacZ-reporter assays and EXAFS studies to identify subtle differences in NiII vs CoII coordination modes in RcnR. (376,378) For example, the N-terminal His donor appears critical for CoII-induced DNA dissociation but not for NiII. (376) While detailed spectroscopic studies have led to atomic-level models of the different holo RcnR species that define residues beyond the “W-X-Y-Z” motif, the lack of a structure hinders our understanding of the allosteric mechanism of transcriptional derepression by RcnR. (204,376,379) On the other hand, allostery in FrmR is driven by a methylene bridge formed between the Cys (C) and the Pro (P) in the “P–C–H-x” motif after the Cys-thiolate attacks HCHO; this chemistry appears to lead to a rearrangement of the surface electrostatic potential distribution that impairs DNA recognition. (361)
Two additional groups of NiII/CoII specific clusters are exemplified by cyanobacterial InrS (119) and Leptospirillum ferriphilum NcrB. (380) The Ni/Co specificity of InrS over CuI and ZnII has been evaluated both in vivo and In vitro, revealing that the affinity of InrS for these other transition metal ions is sufficiently low, i.e., an insufficiently negative ΔΔG relative to NiII (Figure 8), which prevents induction of the expression of genes encoding NiII transporters as a result of mismetalation (Figure 12D). (381) Unlike RcnR, the NiII coordination complex in InrS may well be four-coordinate square planar, which would provide some intrinsic specificity for NiII (Figure 12D). Moreover, InrS is the only Ni/Co specific sensor crystallized to date and is thus far unique among CsoR-family members in that it has been shown to repress the expression of genes encoding NiII efflux transporters, while promoting the expression of NiII uptake genes. (119) These functional characteristics of InrS permitted a detailed study of the relationship between intracellular metal bioavailability and metal sensor metal affinity in Synechocystis cells that provides critical support for our current understanding of the metallostasis set point model (Figure 8A). (119) The NiII/CoII specificity of NcrB and NreA has been inferred from their regulation of nickel resistance determinants; (380,382) however, detailed biochemical and structural studies are needed to evaluate the extent and molecular basis of metal selectivity in these repressors, and what features distinguish them from an RcnR or an InrS.
3.2.4.3. RSS Sensors
Candidate thiol persulfide (RSSH) sensors constitute the largest group of sequences from the CsoR superfamily. The most well-studied groups of CstR (CsoR-like sulfurtransferase repressor) proteins is represented by the founding member S. aureus CstR found largely in Gram-positive organisms (Firmicutes). (363,383) The structure of pneumococcal CstR reveals that the two Cys found on opposite subunits within the dimer unit are characterized by long intersubunit Sγ-Sγ distances (7–9 Å), mediated in part by a “wedge’ residue N55, which defines a x-C-N-C “W-X-Y-Z” motif (Figure 12E). This structure provides a rationale for the observed low CuI binding affinity and enhanced nucleophilicity that prevents crosstalk with the copper sensor CsoR in the same S. aureus cytoplasm. (239) A mass spectrometry-based kinetic profiling method performed with a variety of oxidants, including cysteine persulfide (CSSH), (384) reveals a striking asymmetry of trans-persulfidation within each CstR dimer unit within the tetramer, with one side of the dimer reacting and ultimately closing to a cross-linked product far faster than the opposite side of the dimer. (239) Although reactivity profiles of even closely related CstRs are distinct from one another, a per- or polysulfidated monomer is formed rapidly, on pathway, in all cases. This suggests that the key feature that controls RSS specificity is the enhanced nucleophilicity of the attacking (N-terminal) Cys, which is distinct from persulfide sensors of the ArsR family, e.g., SqrR, which Cys nucleophilicity is lower (Section 3.2.2.4). CstR, unlike SqrR, reacts rapidly with H2O2 In vitro to form the disulfide-cross-linked species, but sluggishly with GSSG. (239)
The lack of a characterized H2O2-specific CsoR-family sensor prevents a detailed side-by-side evaluation of the oxidant specificity of a bona fide RSS and ROS sensor, limiting our understanding of inducer specificity in this structural class of repressors. It is interesting to note, however, that these regulators can function cooperatively with other transcriptional regulators that respond to related sulfur-containing species. For example, it was shown BlYrkD that is encoded in an operon alongside a two-component system that has been proposed to respond to extracellular H2S. (383) The largely uncharacterized S. coelicolor CsoR (CstR) (385) and a mycobacterial protein, MtRv1766 (203) also harbor two conserved Cys of a CstR, adopting a x-C-x-C “W-X-Y-Z” motif; a functional role in RSS sensing is consistent with a genomic neighborhood that encodes for distinct sulfide or sulfite detoxification enzymes. (239) These sensors do not conserve the wedge Asn (N) residue, which suggests that these proteins may well perform oxidative chemistry with a distinct profile of sulfur-containing donors, as yet unknown.

3.2.5. CopY

The copper-specific metallosensor CopY (Figure 13A) is a member of the methicillin and β-lactam repressor (MecI/BlaI) family as a result of its high sequence similarity in the N-terminal DNA binding domain and an identical cognate DNA operator sequence. (386) The DNA-binding domain adopts a canonical winged-helical domain like that found in ArsR, MerR and MarR repressor families, as determined by NMR spectroscopy. (387) The C-terminal domain functions as both the regulatory and the dimerization domain and is of unknown structure. (388) CopYs differ from MecI and BlaI by the presence of a C-terminal unstructured extension appended to the C-terminal domain that harbors a conserved CXCX4CXC motif, or less frequently, a single CXC motif that is known to coordinate CuI or ZnII in distinct coordination geometries and stoichiometries (Figure 13B). (389) CopYs regulate CuI efflux via P1B-1 ATPase CopA and the oxidative stress response in a small number of bacterial pathogens from closely related Firmicutes, derived from the Lactococcus, Streptococcus and Enterococcus spp. (200,390−392) CopY-type regulators are experimentally validated copper sensors in E. hirae, (200) Enterococcus faecium, (393) L. lactis IL1403, (391) Streptococcus mutans, (386) Streptococcus pneumoniae, (390) Streptococcus pyogenes, (176) Streptococcus suis (394) and Streptococcus gordonii. (395)

Figure 13

Figure 13. CopY family proteins. A) Abbreviated periodic table showing the metals and reactive species (shaded) known to be sensed by CopY proteins. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of antibiotic sensing SaMecI dimer protein (1OKR) (397) with close up view of an Alphafold2-modeled CuI metal binding site from CopY. Conserved primary structural features of CopY metal binding homologues are shown.

High Resolution Image
Download MS PowerPoint Slide
Remarkably, CopYs bind both CuI and ZnII with opposite functional outputs. A single ZnII per dimer is coordinated tetrahedrally to the CxC-containing CopY, and is an allosteric activator of DNA binding required for full repression of the cop operon in the absence of CuI stress; as CuI levels in the cell rise, two CuI displace the single ZnII to form an S4–Cu2 cluster that impairs DNA binding. (251) Ion mobility mass spectrometry and SAXS of SpCopY reveals that S4–Cu2 cluster formation leads to conformational instability and protein aggregation, possibly favoring DNA dissociation. (251,396) The distinct functional output of ZnII vs CuI is a unique feature of CopYs, which allow the organism to exploit repressor metal promiscuity to facilitate bacterial survival to copper and zinc toxicity. While the impact of Cu regulation on the crosstalk between Zn and Mn has been investigated in other CopY-encoding bacteria, (176) it remains unclear whether the mechanistic proposal described for SpCopY characterizes other members of this group of repressors.

3.2.6. Fur

The ubiquitous Fur family proteins have evolved to sense all first-row, late d-block divalent metal ions with the exception of CuII, and one specific reactive oxygen species, hydrogen peroxide (Figure 14A). FeII-sensing Fur and namesake of the family is a global regulator that integrates the control of metallostasis of MnII vs Fe, in particular, with other oxidative stressors. This is because FeII in the presence of even small amounts of H2O2 results in the generation of the biomolecule-damaging hydroxyl radical (OH•) via Fenton chemistry which becomes catalytic in the presence of cellular reductants, and catastrophic for cellular function. All Fur superfamily proteins are homodimers (some are tetrameric) where each protomer is characterized by an N-terminal winged helix-turn-helix DNA-binding domain and a C-terminal dimerization domain that are connected by a flexible hinge region. This “hinge” occurs just N-terminal to a conserved HHxHx2Cx2C motif formally located in the C-terminal domain in the vicinity of the β2 and β3 strands (Figure 14B). (398,399) In general, one or more His in the HHxH motif are ligands to the metalloregulatory sites (sites 2 and 3; site 3 is not present in all Furs), with the Cx2C motif forming the N-terminal half of a tetrathiolate (S4) site 1, paired with another Cx2C sequence at or very close to the C-terminus of each protomer (Figure 14B-D). Site 1 is a structural site which is nearly always occupied by ZnII, whose integrity stabilizes the dimer. Although site 1 can form a [2Fe-2S] cluster that is reported to play a role in iron sensing in cells, the significance or generality of this finding among Fur-family proteins is not yet clear. (400)

Figure 14

Figure 14. Fur family proteins. A) Abbreviated periodic table showing the metals and reactive species (shaded) known to be sensed by Fur proteins. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of VcFur (2W57) (416) with close-up view of the metal binding sites. Conserved primary structural features of different Fur proteins with the different metal binding sites indicated in brackets are shown. C) Ribbon representation of ScZur (3MWM) (414) with a close-up view of the metal binding sites. Conserved primary structural features of different Zur proteins with the different metal binding sites indicated in brackets. D) Ribbon representation of BsPerR (2FE3) (417) with a close-up view of the structural metal binding site only. Conserved primary structural features of different PerR proteins with the different metal binding sites indicated in brackets.

High Resolution Image
Download MS PowerPoint Slide
Fur family members harbor one (site 2) or two (sites 2, 3) metalloregulatory sites (Figure 14B) whose affinity for cognate metal is tuned so as to be maximally responsive to intracellular changes in bioavailable metal (Figure 8). When site 3 is present, this allows a “graded” or fine-tuned response to cellular metal status, which generally expands the range of metal concentrations over which the sensor is responsive. (401,402) Fur proteins in general bind to a palindromic A/T-rich sequence found in the promoters of regulated genes using a metal cofactor as a corepressor (Figure 6A, middle). (205) Metal coordination in the regulatory site causes a change in the structure and dynamics of the repressor that promotes and stabilizes its interaction with DNA, (403) although the details if allostery in most Fur proteins remain unclear. (399,404)
The Fur family, like the MerR and CsoR families, provides an excellent illustration of how transition metal selectivity can be tuned by evolving a single primary metal sensing site (site 2) to adopt distinct coordination numbers and geometries optimized to mediate a metal-selective allosteric response (Figure 14B-D). Indeed, the presence of several Fur family proteins in the same organism is rather common, consistent with the idea that they can function independently of one another, while possessing a sufficient degree of specificity for their cognate coregulatory metal as predicated by the set-point model (Figure 8). However, crosstalk among Fur paralogs has been observed, for example in the biosynthesis, efflux and uptake of the broad-spectrum metallophore staphylopine in S. aureus, which are cooperatively regulated by both Fur and Zur. (405,406) Similarly, there are several examples of partially overlapping regulons, (208,399,407) suggesting a tight interconnectivity between metal balance and the resistance to oxidative stress. Indeed, Fur mismetalation may allow Fur to function as a signaling “hub” that globally integrates control of metallostasis and a response to other stresses. (399,401,408−411) Indeed, a regulatory function for apo-Fur has been demonstrated (412) reinforcing the idea that every differentially ligated state of Fur may drive a distinct regulatory output. This is in fact a central tenet of the ability of Fur proteins to orchestrate a graded response to increasing degrees of cellular metal limitation, where differentially metal-ligated states differentially repress the expression of downstream genes as metal levels incrementally fall in the cell. (252,413−415)
3.2.6.1. FeII-Sensing Fur
Fur proteins that function as FeII sensors have been described in many pathogenic and nonpathogenic Gram-positive (418) and Gram-negative bacteria. (409,410,419−425) Early structural work on PaFur (426) and VcFur (416) revealed that FeII and other divalent metals could allosterically activate DNA binding by altering the metalation status of one or both of the two regulatory sites observed (Figure 14B). These two Fur proteins, like MgFur, appear to lack what is now regarded as the canonical S4 structural site 1, like that found in CjFur and BsFur, or that site if present, is simply devoid of metal (Figure 14B). It is the case that the double Cx2C sequence motif of a canonical S4 site 1 is present in most Fur sequences, which when metalated with ZnII, behaves as an apoprotein (not activated to bind DNA). (252) Regardless, the conservation and interdigitated nature of the residues that form coordination bonds to one or the other regulatory site (site 1 in these structures corresponds to site 3 in other Furs with the S4 site) allows for a significant degree of plasticity in the coordination chemistry. Indeed, the identity of the donor ligands and the coordination geometry in each site differs among the X-ray structures solved with the same or different divalent metals, e.g., ZnII in PaFur, (426) CjFur (404) and VcFur; (416) and MnII in MgFur. (425) Moreover, the regulatory impact of two observed metal sites may well differ among individual FeII-sensing Furs. For example, in VcFur, metal occupancy of site 2 only by MnII, ZnII and FeII, (416) but not NiII, (427) promotes DNA binding regardless the identity of the metal in site 1 (site 3). In contrast, MgFur site 1 (site 3) has been shown to be essential for DNA binding and site 2 was proposed to be only fine-tune DNA binding affinity. (425) How distinct FeII-sensing Furs are wired for distinct allosteric connectivities is not yet understood but remains an interesting unresolved question.
3.2.6.2. ZnII-Sensing Zur Proteins
The zinc uptake regulator (Zur) is a ubiquitous Fur-family metalloregulator that represses transcription of the genes that encode proteins involved in ZnII homeostasis, including high affinity transporters that mediate zinc uptake, metal allocation, low zinc paralogs and other physiological processes, depending upon the organism (Sections 4-5). (428−430) EcZur, (431) BsZur (413) and AbZur (432) harbor only two ZnII sites per protomer corresponding to the canonical S4 structural site 1 and a regulatory site 2, while MtZur (433) and ScZur (414) each contain three sites per protomer, adding site 3, the presence of which is proposed to allow for the “graded” zinc deficiency response. (413,414) It is remarkable in fact, that the central stretch of four consecutive His residues in MtZur (433) and ScZur (414) alternate by forming coordination bonds to site 2 and site 3, on opposite sides of the same β-strand. In all cases, the ZnII ions are tetrahedrally coordinated (Figure 14C). Zur proteins are readily distinguished from Fur proteins by the presence of a Cys in the N-terminal region of the β2 strand, which roughly defines the hinge point between the N- and C-terminal domains which plays a key role in allostery; (432) however, it remains unclear the extent to which these differences in sequence can easily explain the differences in metal binding preferences, particularly in those regulatory sites that impact DNA binding. While some Zur proteins bind DNA only in the presence of ZnII ions, (330,434,435) others showed Zur–DNA binding in the presence of ZnII, MnII, and CdII, which again suggests that the cellular milieu enforces a metal specificity profile (Figure 8) that cannot be recapitulated by simple In vitro experiments. (436)
Recent work reveals that some Zur proteins harbor a pair of regulatory ZnII sites distinct from the central HHxHx2Cx2C motif (vide supra) that stabilizes an oligomeric assembly of three or more Zur dimers on the promoters of genes encoding cognate metal efflux transporters, leading to RNA polymerase recruitment and activation of transcription when metal levels become very high. (437−440) This regulatory mechanism thus integrates repression of uptake transporters and up-regulation of efflux transporters over a very large range of metal activity using a single regulator, allowing for a coordinated cellular response to extreme metal toxicity.
3.2.6.3. Hydrogen Peroxide Sensor PerR
PerR is the master regulator of peroxide stress in many Gram-positive organisms and is functionally equivalent to the thiol-containing OxyR (Section 3.2.13) encoded by most Gram-negative bacteria, with only a few exceptions. (441,442) PerR is an FeII binding protein that forms a six-coordinate FeII structure with five protein-derived ligands (site 2) and an open coordination site that accommodates the binding of H2O2. The first structure of a PerR protein solved by X-ray crystallography corresponded to apo-PerR-Zn from B. subtilis, in which the structural S4 site 1 was occupied by ZnII (Figure 14D). (417) When peroxide binds to the regulatory FeII complex, metal-catalyzed oxidation triggers the incorporation of one atom of oxygen into one of the His in site 2 (either H36 of H91 in BsPerR sequence, (443) labeled with a star in Figure 14D). (444−447) Histidine oxidation weakens the affinity of PerR for FeII and thus dissociates from the DNA with the loss of its corepressor. The structural changes induced in PerR upon loss of the regulatory remain unresolved, although it seems unlikely that the loss of the metal creates a “splayed out, open” conformation (Figure 14D). (417) The PerR regulatory site binds FeII and MnII and both function as corepressors; however, only FeII can carry out the oxidative chemistry. Thus, the functional responsiveness of PerR to peroxide stress is tuned to the cellular MnII:FeII ratio, which is in part governed by FeII-sensing Fur. (443) The mechanism underlying the peroxide selectivity by PerR relative to Fur remains incompletely understood, but replacement of E108 in β2 by Asp as found in PerRs to create E108D BsFur makes it sensitive to H2O2. (448)
3.2.6.4. Other Fur Regulatory Proteins
There are a number of Fur family regulators that integrate responses to other transition metal ions, including Mur (449−451) and Nur, (452,453) both of which appear to sense MnII, and Irr, which is a heme-based sensor. (454,455) Although microbial physiology studies suggest metal or metallocofactor-specific responses, there are significant data that show that metalation with various metals can promote DNA binding In vitro. These observations reinforce the idea that our understanding of metal specificity in Fur proteins remains limited, and must be developed beyond the simple model of a single cognate metal corepressor.

3.2.7. DtxR

DtxR (Diphtheria toxin repressor) family proteins form a phylogenetically conserved group of FeII-and MnII-regulated repressors prevalent largely in GC-rich Gram-positive bacteria and archaea, evolutionarily distinct from, but functionally analogous to, Fur family proteins (Figure 15A). (456,457) DtxR family of proteins are characterized by an N-terminal DNA-binding domain, an interfacial domain which contains two metal-binding sites, and a third C-terminal SH3 (Src homology 3)-like domain not present in all DtxR family crystal structures. (458,459) The second domain is comprised of three helices, organized as a pair of three-helix bundles from each protomer and is thus responsible for homodimerization and regulatory metal site binding. SH3-like domains are structurally like eukaryotic SH3 domains, which mediate protein–protein interactions in many signaling proteins, (460,461) but have no detectable sequence similarity with them.

Figure 15

Figure 15. DtxR family proteins. A) Abbreviated periodic table showing the metals and reactive species known to be sensed by DtxR family are colored. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of SeIdeR protein (7B1 V) (459) with a close-up view of the primary metal binding site and auxiliary site. Conserved primary structural features of different IdeR proteins with the different metal binding sites indicated in brackets. C) Ribbon representation of BsMntR (1ON1) (462) with a close-up view of the primary and auxiliary metal binding sites shown. Conserved primary structural features of different MntR proteins with the different metal binding sites indicated in brackets.

High Resolution Image
Download MS PowerPoint Slide
3.2.7.1. FeII-Sensing IdeR/DtxR
IdeR (Iron-dependent repressor) regulates the expression of genes that encode proteins involved in the acquisition and storage of iron. These include siderophore production and transport, iron transporters, ferritins and bacterioferritins, as well as proteins involved in iron–sulfur cluster assembly and metabolic enzymes that harbor Fe cofactors. (211,458,459,463−465) While FeII–IdeR represses iron uptake, it activates the expression of iron storage genes. (463) IdeR/DtxR proteins play other roles beyond controlling iron homeostasis and in M. tuberculosis and C. diphtheriae are required for full virulence; (456,463) in the actinomycete Streptomyces avermitilis, IdeR regulates morphology and cellular metabolism. (465) While several IdeR homologues are known to respond specifically to FeII, (465−467) CdDtxR responded equally well to FeII and MnII when expressed in Bacillus subtilis (an organism that employs Fur to sense iron). (468) Reconciling in vivo selectivity for FeII with in vitro promiscuity remains a challenge. (212,459,469) For example, the binding of DtxR to the Tox operator In vitro is activated by a range of divalent metal ions, including CoII, FeII, MnII and NiII, whereas ZnII shows only weak activation. (470) Each IdeR subunit contains two metal-binding sites, each adopting octahedral coordination geometries, with the so-designated “primary” site characterized by a larger impact on IdeR activity; however, both primary and “auxiliary” sites must be occupied for full IdeR activity, and the auxiliary site has higher metal binding affinity (Figure 15B). (469,471−473) In both metal binding sites, the metal coordination sphere is completed by a water ligand (not visible in all structures, but likely always present). (458) To date only one structure of a bacterial IdeR protein activated by its physiological effector FeII is available. (459) However, structures of fully metalated, full-length bacterial DtxR/IdeR proteins show the same metal-binding sites and coordination geometry regardless whether a cognate or noncognate metal is used. (458,459,474) Selectivity for FeII over MnII may be enhanced by the presence of two sulfur ligands (M10 and C102) in the primary FeII site, since the corresponding MntR MnII-sensing site contains an aspartate or asparagine in place of M10 and a glutamate in place of C102 (Figure 15C). Introducing the IdeR-specific metal ligands to an MntR protein and vice versa switches or broadens the metal responsiveness of the respective variant in cells. (468) While IdeR can bind MnII it does not strongly activate DNA binding, as it is likely not bound to the primary site. (469,473) On the other hand, crosstalk with ZnII homeostasis has been noted. (475) It has been proposed that due to the high ZnII binding affinity to the primary site, only the auxiliary site would be filled with FeII in vivo with the other bound by a noncognate metal, most likely ZnII; if true, this scenario is precisely analogous to what has been described for sites 1 and 2 in Fur proteins (vide supra). (458)
3.2.7.2. MnII-Sensing MntR
Structural and functional studies of MnII-specific DtxR family sensors are more limited that of the FeII responsive DtxR family proteins. Most of what we know about the metal selectivity of a MnII-sensing DtxR protein comes from BsMntR. BsMntR forms noncognate metal complexes with FeII, ZnII and CoII, that are mononuclear, rather than binuclear MnII complex, and do not allosterically activate DNA operator binding (Figure 15C). (462,476) Early structural work on BsMntR revealed that the metal-binding pocket was substantially distinct from that of DtxR, with the two metal binding sites physically closer, sharing two bridging ligands, and including two water molecules, allow for heptacoordination of the MnII ions (Figure 15C). (462) Other MnII sensing DtxR family proteins such as S. pneumoniae PsaR are also known employ two regulatory metal sites to activate DNA binding, with ZnII in the primary site and MnII in the auxiliary site. (477) A more recent study of the Deinococcus radiodurans MnII sensor DR2539 revealed additional metal binding sites and new insights into DNA operator recognition by these MnII sensors. (478) Other MnII-dependent transcriptional repressors from the DtxR family have been described to have an impact on pathogen virulence (Figure 15A). (456,477,479−482) This is an exciting area of future study since MnII-specific DtxRs are found broadly in Gram-positive and Gram-negative bacteria. (458)

3.2.8. NikR

NikR proteins are obligate homotetramers with a central C-terminal metal-binding domain (MBD) connected by flexible linkers to two peripheral N-terminal DNA-binding domains (DBDs) that adopt a ribbon-helix-helix fold. (483) NikR proteins sense the bioavailability of nickel only, which allosterically activates high affinity DNA binding and complex formation with the promoters of regulated genes (Figure 16A). While H. pylori NikR has been shown to act as both a repressor and as an activator of diverse genes that share a weak consensus in their promotor sequences, (484) E. coli and Pyrococcus horikoshii NikRs are known only to repress NiII-specific ABC-type importers under NiII-replete conditions. (213,485) The nearly 30 high resolution crystal structures of NikRs in various allosteric states reveal four symmetry-related conserved square planar His3Cys NiII coordination complexes per tetramer that promotes a conformational change to a closed conformation competent for DNA binding (Figure 16B; Figure 6A). (134,214,485−490) This coordination geometry allows for picomolar affinity for NiII as measured by chelator competition assays. (491,492) Since NikR also binds other divalent metals, the biological NiII selectivity arises here from a precise control of allosteric coupling to DNA binding (Figure 8A) and KdS (Figure 8B). Noncognate metal ions ZnII and CoII bind with different affinities and stoichiometries, and form non-native coordination geometries and ligand sets that do not support a metal-mediated allosteric response. (491,493) A recent study suggests that NiII-specific binding couples an allosteric network of residues at an interface between the DNA binding and metal binding domains (Figure 16B), which in turn, appears to result in an increase in protein rigidity, locking NikR in a DNA binding-active conformation. (490)

Figure 16

Figure 16. NikR family proteins. A) Abbreviated periodic table showing the metals and reactive species known to be sensed by NikR (shaded). B) Ribbon representation of EcNikR (2HZV) (134) with close-up view of the NiII coordination complex. Conserved primary structural features of different NikR proteins with the primary metal binding site indicated by the brackets.

High Resolution Image
Download MS PowerPoint Slide
NikR thus provides an exceptional illustration of the importance of metal coordination geometry as a key feature of cognate metal sensing by a metalloregulatory protein, since if mis-metalation occurs, it does so without triggering a transcriptional response. It is interesting to note that while there is no evidence of regulatory crosstalk with other metal ions, HpNikR also responds to changes in pH which is consistent with its role as a global regulator in that organism. Here, NikR binds DNA with high affinity in the acidic environment H. pylori encounters in the human stomach, even in the absence of nickel. (494,495) Both nickel and lower pH activate HpNikR to bind to the ureA promoter with similar affinities and locations on the DNA, (495) but the mechanism of allosteric regulation of DNA binding by acidity remains unknown.

3.2.9. Rrf2

Rrf2 repressors are dimers consisting of an N-terminal DNA-binding “winged helical” domain composed of five helices and two β-strands, and a C-terminal helical dimerization domain dominated by a long helix, and thus superficially resemble MerR-family activators (Section 3.2.3). (216,496−498) The two domains are connected by an unstructured linker where the inducers and cofactors are bound. Rrf2 family repressors regulate sulfate reduction (Rrf2), (499) iron–sulfur cluster biogenesis (IscR), (216,500) cysteine metabolism (CymR), (496,501,502) iron homeostasis (RirA), (215,503,504) catechol iron uptake-coupled quinone detoxification (SifR), (246) redox balance/oxidative stress (RsrR, (497,505) HypR, (506) SaiR (507)), and the sensing of NO and reactive nitrogen species via a [4Fe-4S] cluster (NsrR) (Figure 17A). (498,508,509)

Figure 17

Figure 17. Rrf2 family proteins A) Abbreviated periodic table showing the metals and reactive species known to be sensed by Rrf2 family are colored. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of SvRsrR (6Y42), with a close-up view of the [2Fe-2S] cluster site. (505) Conserved primary structural features of different RsrR-family proteins with the primary iron sulfur cluster site indicated by the brackets. C) Ribbon representation of ScNsrR (5N07), with a close-up view of the [4Fe-4S] cluster site. (498) Conserved primary structural features of different NsrR-family proteins with the primary iron sulfur cluster site indicated by the brackets. D) Ribbon representation of SaCymR (3T8T), with a close-up view of the reactive cysteine. (515) Conserved primary structural features of different CymR-family proteins that are also thought to function as redox sensors.

High Resolution Image
Download MS PowerPoint Slide
Functionally characterized members of the Rrf2 family can be broadly divided into two groups. One group utilizes an Fe–S cluster as the sensory module, (510) while the other contains one or a pair of Cys residues that sense reactive oxygen or reactive electrophile species. (507) The prototypical Rrf2-family repressor is proteobacterial IscR, an Fe–S cluster regulator that contains a 2Fe–2S cluster ligated by three Cys and one His and controls the biogenesis of Fe–S clusters in cells. (216,500,511) The chemical nature of the Fe–S cluster differs depending on the small molecule inducer. While RsrR harnesses the same IscR-like 2Fe–2S cluster as a sensory module for redox sensing (Figure 17B), (509) RirA (512) and NsrR (513) utilize a 4Fe–4S cluster to detect iron limitation and nitric oxide, respectively (Figure 17C). The cluster coordination chemistry appears to be remarkably variable. EcIscR is unique, (500) while RsrR uses one His, one Glu, and two Cys as cluster ligands (Figure 17B). (497) RirA utilizes three Cys as cluster ligands, (503) while NsrR employs three Cys and one Asp as cluster ligands (Figure 17C). (498) Studies on TpIscR reveal that different homologues of IscR employ different ligand sets, despite what is likely the same function in surveilling the Fe–S status of the organism (Figure 17B). It is important to note that while many biochemical studies provide support for the disparate structure of these Fe–S clusters in Rrf2 proteins, crystallographic structures have only been reported for holo-NsrR (498) and RsrR. (497)
The diversity in allosteric mechanisms of the Fe–S-containing members of Rrf2 family proteins is also of high interest. While IscR binds DNA in the apo-form and activates transcription, (514) Fe–S cluster formation in IscR, RsrR, NsrR and RirA leads to high-affinity DNA binding and repression, with subsequent chemistry leading to DNA operator dissociation and transcriptional derepression. (216,497,503,509,513) For example, in RirA, oxidation of the cluster by molecular oxygen leads to cluster disassembly and DNA dissociation, which allows RirA to function as a redox sensor. (504) In contrast, the one electron oxidation of the Fe–S cluster in RsrR leads to a ring-flip of the conserved Trp9 that appears necessary and sufficient to induce DNA dissociation; this allows RsrR to elicit specific transcriptional responses to cellular redox imbalance (Figure 17B). (497) Similarly, NsrR utilizes the 4Fe–4S cluster as a NO-sensing module; here, NO displaces Asp8 as a cluster ligand, destabilizing the cluster and thus leading to DNA dissociation (Figure 17C). (498,513)
The chemistry and allosteric mechanisms of Rrf2 proteins that do not coordinate an Fe–S cluster is comparatively less well understood, complicated by the fact the degree of Cys conservation in the sensory module is rather low. However, most characterized thiol-reactive proteins harbor at least one reactive Cys, modification of which modulates DNA binding affinity (Figure 17D). In the case of SaCymR, the reactive Cys is in the wHTH domain itself, and not in the sensory module; here, H2O2-promoted sulfenylation has been proposed to trigger DNA dissociation (Figure 17D). (515) This Cys however is not conserved among other CymR proteins that are also thought to be redox sensors. (496,501,502) BaSaiR also responds to H2O2 likely through a similar mechanism, (507) while SaHypR appears to react selectively with hypochlorous acid (HClO) and does not react with H2O2. (506) The most recent member of this thiol-containing subgroup of Rrf2 proteins to be characterized is the streptococcal IscR-like family transcriptional repressor (SifR). SifR uses a single conserved Cys to react with quinones which are potent cellular electrophiles. (246) These quinones are thought to originate with the oxidation of simple catechols brought into the cell as chemically diverse coordinately unsaturated FeIII–catecholate complexes through PiuBCDA. S. pneumoniae does not encode Fur or other global Fe regulator due to relatively modest Fe quota; the SifR-regulon seems to allow S. pneumoniae to meet its needs for nutritional Fe while avoiding quinone toxicity. (516)

3.2.10. GntR

Many GntR superfamily proteins regulate the expression of enzymes involved in the metabolism of sugar acids, (517−519) fatty acids, amino acids and cofactors, generally orchestrating a response to the nutrient and morphological status of the organism. (520,521) They harbor an N-terminal winged helix-turn-helix DNA-binding domain and a C-terminal effector-binding and oligomerization domain often denoted FCD (FadR C-terminal domain); (518) the C-terminal domain is the differentiating feature of well-characterized bacterial regulators, including FadR, HutC, MocR, YtrA, AraR, and PlmA subfamilies. (518,522) A small number of GntR family proteins have been shown to be metal-responsive regulators, notably M. smegmatis GfcR, a global regulator that responds directly or indirectly to external copper stress by regulating the expression of a number of membrane-associated transporters (Figure 18A). (523) It has been proposed that ≈70% of GntR proteins maintain a conserved, high binding site for ZnII. (217) This metal binding site is buried at the bottom of a solvent accessible cavity, where it facilitates the binding of small organic acids that harbor a carboxylate group in a way that is required for transcriptional regulation (Figure 18B). (217,524−526) The small molecule effector that induces allosteric regulation of DNA binding has only been identified for a handful of sugar-sensing GntR proteins, namely GntR from B. subtilis (D-gluconate), LldR from Corynebacterium glutamicum (L-lactate), (525) GguR from Polaromonas sp. (5-keto-4-deoxy-D-glucarate/galactarate), (527) and DgoR (D-galactonate), (528) ExuR (D-galacturonate and D-glucuronate) and UxuR (D-glucuronate and D-fructuronate) from E. coli. (529) The extent to which metal binding is regulatory in this family of proteins or functions simply a structural cofactor that enables molecular recognition of the cognate small molecule effector remains an open question, and a subject for future studies given the ubiquitous nature of this protein superfamily in bacteria. (519)

Figure 18

Figure 18. GntR family proteins. A) Abbreviated periodic table showing the metals and reactive species known to be sensed by GntR family are colored. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of CgLldR protein (2DI3) (525) with a close-up view of the primary ZnII binding site. Conserved primary structural features of different GntR proteins with the disinct metal sites indicated in the text.

High Resolution Image
Download MS PowerPoint Slide

3.2.11. MarR

MarR proteins are triangularly shaped homodimers that share a winged-helical DNA-binding domain connected to a DNA-distal all-α-helical dimerization domain. Members of this protein superfamily share the same winged-helical DNA-binding fold of ArsRs but harbor an additional C-terminal (α6) helix positioned in the dimer interface that dramatically alters the quaternary structure of the homodimer relative to ArsR family repressors (Figure 19A-B). (530−532) Canonical MarRs that repress the expression of drug efflux pumps may bind antibiotics or other small organic molecules leading to the transcriptional depression of the genes associated with an adaptive response to antibiotic stress, e.g., efflux systems. (533) Individual MarR members have been shown to bind a diverse range of ligands at a cleft between the dimerization and DNA-binding domain, (534−543) as well as to different sites on the dimer; (537,544−546) likewise, oxidation-sensing cysteine residues are also found in distinct positions within the dimer (Figure 19A). (547−556) This functional diversity of biological outcome and inducer recognition among MarR proteins stands in striking contrast to a relatively common derepression mechanism where either ligand binding or cysteine oxidation leads to DNA dissociation (i.e., allosteric inhibition, ΔGc > 0; Figure 6B). (532,557−559) It has been argued that the ancestral role of MarR family proteins was to maintain physiological homeostasis by sensing and derepressing the expression of genes encoding efflux pumps for both toxic metabolites as well as unnatural xenobiotics. (560) The structural versatility required to accommodate binding of a broad range of ligands to different sites coupled with the wide distribution of oxidation-sensing Cys residues and low overall sequence similarity among MarR regulators, collectively suggest a model where ligand binding induces a conformational change in the homodimer that cannot be explained by a single sensing site and accompanying pathway of allosteric communication. MarR proteins are thus reminiscent of ArsR-family repressors in this regard (vide supra). (139)

Figure 19

Figure 19. MarR family proteins. A) Abbreviated periodic table showing the metals and reactive species (shaded) known to be sensed by MarR proteins. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. Ribbon representation and conserved primary structural features of B) SpAdcR protein (3TGN) (218) with a close-up view of the two ZnII binding sites indicated by brackets; C) XcOhrR protein (2PFB) (553) with a close-up view of the disulfide cross-link that results from ROS sensing.

High Resolution Image
Download MS PowerPoint Slide
3.2.11.1. ZnII-Sensing MarRs
ZnII-sensing MarRs are the only characterized metalloregulatory proteins from this very large superfamily, while also constituting an exception to the rule of allosteric inhibition of DNA binding as a general feature of MarR proteins. MarR family metallosensors are thus far restricted to regulators of ZnII uptake, (561−563) exemplified by AdcR (adhesin competence regulator) from S. pneumoniae and closely related Streptococcus and Lactococcus sp. exemplified by L. lactis ZitR (Figure 19B). (141,563−569) AdcR and ZitR possess two physically proximate pseudotetrahedral ZnII binding sites that bind ZnII with very different affinities. (133,218) A detailed structural study of LlZitR showed that the binding to ZnII occurs sequentially with a graded effect on the DNA binding affinity. (568) ZnII is an allosteric activator of DNA operator binding, in strong contrast to all other members of the MarR superfamily, consistent with its biological function as uptake repressor at high intracellular ZnII. It is interesting to note that ZnII coordination sites employ ligands from an unstructured region between the first two α-helices that for all other MarR regulators is part of a longer α2 helix (Figure 19B–C). All characterized metalloregulators of this family have been shown to bind DNA only in the metal-bound form and they all share this unstructured region, which is characterized by enhanced flexibility in the subns time scale in AdcR in the apo-state. (141,218,566−569) It has been proposed that introduction of this dynamic element during the course of MarR protein family evolution may have enabled allosteric activation in this molecular scaffold where metals bind as a corepressors (Figure 6A, middle). (141) The extent to which this allosteric activation model characterizes other metal-sensing MarR proteins yet to be characterized is unknown. (570)
3.2.11.2. Oxidation-Sensing MarRs
In many MarRs, cysteine oxidation is known to occur in distinct sites on the molecular scaffold, dependent on the nature of the stressor (Figure 19C). While SarZ, MgrA, AbfR and OhrR share a conserved cysteine in the N-terminal DNA binding domain that can become sulfenylated, (547,549,551,553,556) the identity of the nonconserved resolving Cys, as well as the Cys pair for other cysteine-based regulators from the MarR family differ. (534,571) Alternatively a resolving Cys may not be present at all, with S-thiolation or mixed disulfide formation with an abundant low molecule weight thiol the regulatory switch for DNA binding. (550) In any case, while the degree of crosstalk among different reactive species that signal through MarRs has not been fully elucidated, it is interesting to note that OhrR and MexR have been shown to perform chemistry with sulfane sulfur donors, including thiol persulfides in P. aureuginosa. (552,572) Furthermore, EcMarR has been shown to respond to increased levels of CuII, (534,573) which results in Cys oxidation and disulfide bond formation to create tetramers that have lost the ability to bind DNA. The extent to which this CuII-dependent regulatory modelas proxy for localized bursts of oxidative stress, functions in other MarRsrequires further study.

3.2.12. TetR

The tetracycline repressor (TetR) family, like the MarR family described above, is widely associated with antibiotic resistance and the regulation of genes encoding small molecule exporters. (574) TetRs, however, play a far more widespread regulatory role in metabolism, controlling the expression of genes involved in central carbon metabolism, antibiotic production, quorum sensing, and many other aspects of prokaryotic physiology. There are more than 700,000 proteins identified as TetRs to date and more than 500 high-resolution structures have been deposited in the Protein Databank (PDB). Canonical TetRs are all-α-helical proteins that function as ligand-mediated, homodimeric repressors, where each subunit harbors an N-terminal, three-helix bundle (α1−α3) within which is the helix-turn-helix (HTH) DNA binding motif. The HTH motif is followed by a large C-terminal regulatory domain (α4−α9) that incorporates a number of independently evolved ligand binding cavities that are readily accessible from the side, front, or top of the homodimer. (574,575)
While many TetRs function as repressors that provide cellular resistance against hydrophobic antibiotics and detergents, the few metalloregulator proteins and inorganic sensors described to date have distinct allosteric mechanisms of transcriptional control (Figure 20A). On the one hand, the large hydrophobic inducer-binding pocket in a canonical TetR regulator has been repurposed by the heme-regulated transporter regulator from L. lactis (HrtR) to bind heme tightly and induce DNA dissociation leading to transcriptional derepression of a heme-efflux system. (576,577) On the other hand, Cys thiol-based TetR sensors have been described that react with a wide variety of small molecule electrophiles. For example, E. coli N-ethylmaleimide reductase repressor (NemR) utilizes a cysteine-lysine sulfenamide thiol switch to sense HClO; (578) furthermore, a recent report suggests that an Enterobacter sp. protein may be induced by binding of organoarsenic compounds and chromium to a three Cys site. (579) S. aureus glucose-induced biofilm accessory protein A (GbaA) utilizes a dithiol pair to respond to oxidative and electrophile stress, however, the identity of the in vivo inducer is as yet unknown. (580,581) While there are no TetRs reported to date that can be induced by persulfides, a recent report suggests that a TetR encoded in Gordonia sp. regulates biodesulfurization processes in the presence of organosulfur compounds. (582) While the objective of this singular study was to evaluate this sensor in an eventual application for the removal of sulfur in petroleum fractions, the reported results suggest that this regulator may react with sulfane sulfur-containing oxidants.

Figure 20

Figure 20. TetR family proteins. A) Abbreviated periodic table showing the metals and reactive species known to be sensed by TetR family are colored. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of SaSczA protein (3KKC) with a close-up view of the primary ZnII metal binding site region. (220) Conserved primary structural features of different TetR proteins with distinct metal sites.

High Resolution Image
Download MS PowerPoint Slide
The clearest example of a TetR-family metallosensor is the transcriptional activator of the S. pneumoniae ZnII efflux system, SczA, and related activators from other streptcocci (Figure 20B). (583) SczA activates the expression of the gene encoding the ZnII efflux transporter, CzcD, upon metal binding to two pairs of metal sites per dimer, one with a tetrahedral coordination geometry and one with an octahedral geometry. (220) It is interesting to note that the crystal structure available from an SczA homologue was crystallized with NiII (likely obtained from the purification buffer). The identification of the true cognate metal inducer requires detailed studies of metal affinities (584) and a thorough characterization of the transcriptional response in cells. (220) Although the mechanism of transcriptional activation of SczA is not yet fully understood, this likely involves recruitment of RNA polymerase to SczA-regulated promoters in a ZnII-dependent fashion. In any case, the community has taken full advantage of the ease with which metal-dependent transcriptional activation can be obtained with SczA, in both synthetic biology and routine molecular biology applications. (585) The footprint of TetR family inorganic sensors might not be restricted to heme and ZnII, as a putative Cu-responsive TetR has been identified as ComR in Escherichia coli. (586,587) Although ComR is involved in the Cu stress response in some way, direct CuI or CuII binding by ComR has not yet been reported.

3.2.13. LysR

The LysR-type transcriptional regulators (LTTR) are named for the lysine repressor, LysR, which regulates lysine biosynthesis and is representative of the largest family of transcriptional regulators in prokaryotes. (588) LTTRs are involved in the transcriptional regulation of genes involved in central metabolism, amino acid biosynthesis, cell division, fermentation, photosynthesis, nitrogen fixation, oxidative stress, quorum sensing, motility, transport of NaI, symbiosis, and virulence. (588) Members of this family possess the same two domain-based molecular architecture of other regulators discussed here, i.e., GntR and TetR, an N-terminal all-helical DNA binding domain and C-terminal globular ligand binding domain. While most characterized LysR members respond to effector molecules such as amino acids or intermediate metabolites, some members can be described as metalloregulators, since they are involved in regulation of inorganic species uptake (ModE) (589,590) or harness thiol chemistry analogous to that described above for dithiol sensors that sense H2O2 (OxyR) (221,591) and/or persulfides (LasR) (592) (Figure 21A).

Figure 21

Figure 21. LysR family proteins. A) Abbreviated periodic table showing the metals and reactive species (shaded) known to be sensed by LysR proteins. The four-letter designations for individual proteins that perform the function listed in the nearby box are shown. B) Ribbon representation of CgOxyR (6G4R) with a close up view of the H2O2 binding site. Conserved primary structural features of different regulator LysR proteins with the different ligand binding sites indicated in the text. (596) C) Ribbon representation of EcModE (107L) with a close-up view of the molybdate binding site shown. (221)

High Resolution Image
Download MS PowerPoint Slide
E. coli OxyR is the prototypical redox-responsive LTTR that mediates transcriptional activation of an oxidative stress regulon in response to nanomolar hydrogen peroxide. (593−595) OxyR exploits cysteine sulfenylation chemistry to affect transcriptional activation of the expression of target genes. OxyR can also respond to superoxide and nitrosative stresses, although these functions are presumed to be ancillary to its primary role as H2O2 sensor. (589,590) Crystallographic evidence for the H2O2-sensing mechanism of Corynebacterium glutamicum OxyR (Figure 21B) identified key residues for the peroxidatic reduction of H2O2 and the formation of a disulfide bond that triggers a transcriptional response by altering the OxyR interacton for the promoter region. (596) Similarly, the quorum sensing master regulator from Pseudomonas aeruginosa PAO1 LasR binds to its target DNA and activates transcription upon modification by a sulfur-based oxidant, e.g., a thiol persulfide, which first forms a C188 persulfide or polysulfide moiety, ultimately resulting in a pentasulfide linkage between Cys201 and Cys203. (592) However, in this case, installation of this modification does not impact the affinity of LasR for its target DNA operator, but makes it several-fold more effective in activating transcription in both In vitro and in vivo assays. Since the identities of the reactive Cys residues in the peroxide sensor OxyR and the sulfane sulfur sensor LasR are similar, the structural mechanism of oxidant specificity remains unresolved.
Metallosensors identified for this family are limited thus far to the sensing of the oxyanion molybdate (MoVI) by ModE, which represses the transcription of MoVI uptake system in E. coli and other bacteria. (221,591) Crystal structures of both apo- and molybdate-bound forms of ModE (597) reveal that the molybdate binding to the C-terminal domain alters the relative orientation of the N-terminal DNA binding domains from each protomer, allosterically activating the ModE homodimer to bind DNA (Figure 21C). The C-terminal regulatory domain of ModE is actually two subdomains, each corresponding to a Mop sequence similarity region. The first Mop subdomain harbors the molybdate binding site, where R128, S126 (SARN segment), S166, and K183, with the assistance of T163 and A184, form the binding pocket for the oxyanion. ModE shares the conservation of two cysteines separated by 8–13 residues in the C-terminal domain with OxyR/LasR, which enable the formation of a disulfide bridge in response to oxidants in these regulators.
While no LTTR is as yet known to bind transition metals directly, (588) the remarkable diversity of the family, the difficulties identifying the cognate effector and the clear role of dithiol chemistry suggest that it is possible that metalloregulatory LysR activators may yet be discovered or evolved for the development of novel biosensors for metal ion monitoring.

3.3. RNA-Based Metalloregulators

3.3.1. Overview

Metalloriboswitches, (598) are highly structured, chameleon-like regulatory RNA domains found within the 5′ noncoding or untranslated region (UTR) of protein-coding mRNAs. The binding of a ligand, a specific metal, metal complex or other inorganic anion, to the aptamer domain alters the folding of the RNA, and thus expression of a downstream gene. More than 55 distinct classes of natural riboswitches have been discovered in the past 20 years. The discoveries of inorganic element-sensing riboswitches have revealed that RNAs harness the power of supramolecular chemistry, functional dynamics and coordinate-covalent bonding to achieve a metal selectivity profile that rivals that of metalloregulatory proteins, albeit sometimes with lower sensitivities (KdS; Figure 8). (229,599) While some extant riboswitches might be direct descendants of the RNA-based sensors and switches that may have been present in ancient organisms before the emergence of proteins, (229) in existing bacteria they are often paired with a metalloregulatory protein, endowing an organism with the ability to expand the range of metal bioavailability that is sensed, as dictated by their often weaker affinities (Figure 8A). More work is needed to provide support this conclusion, however.
Inorganic riboswitches thus far described are known to sense transition metal ions, (226,228,600−602) fluoride anion (603) and other ions including the alkali metals LiI and NaI, and the alkaline earth metals MgII and CaII (Figure 22A). (150,226,604) Here we aim to discuss molecular insights into metalloriboswitch metal selectivity from atomic structures, and how the physiological “division of labor” between these RNA elements and metalloregulatory proteins might be orchestrated. It should be noted that the metalloriboswitches thus far characterized represent only a small fraction of the ≈10,000 riboswitches that have been proposed to exist in modern organisms; (229,599) thus, this picture of metal selectivity and specificity may well change as new insights emerge. An interesting observation is that small organic molecule-sensing riboswitch pairs are capable of functioning as canonical logic gates in cells; (605) the extent to which metalloriboswitches function in this way, perhaps in concert with protein partners, is not yet known.

Figure 22

Figure 22. RNA-based metalloregulators. A) Abbreviated periodic table showing the metals and other elements (shaded) known to be sensed by inorganic riboswitches. The designations for individual RNAs that perform the function are listed near the box shown. B) Structure of cobalamin riboswitch in marine metagenome with close up view of the aquo-adenosylcobalamin binding site (4GMA). (227) C) Structure of MnII-sensing riboswitch from Lactococcus lactis (Ll ykoY) bound to MgII (site A) and the MnII (site B) (4Y1I). (606) D) Structure of NiCo riboswitch from Erysipelotrichaceae bacterium with a closeup view of the four Co sites (4RUM). (224) Metal site 4 was later shown to play no regulatory role and is not occupied in solution studies. (607)

High Resolution Image
Download MS PowerPoint Slide

3.3.2. B12- and Moco-Sensing Riboswitches

The adenosylcobalamin (AdoCbl) riboswitch represents the first high-resolution structure of a transition metal or organometallic cofactor bound to a riboswitch. (227) Here, AdoCbl recognition is accomplished largely through shape complementarity between the RNA and the ligand, thus highlighting the role that supramolecular chemistry plays in molecular recognition (Figure 22B). More recent studies show that the cobalamin-sensing riboswitch RNAs as a class is capable of sensing and responding to multiple derivatives of vitamin B12, collectively termed corrinoids. (608) The exquisite corrinoid specificity of these RNAs is achieved by shape complementarity that is highly sensitive to distinct conformational features of individual B12 variants.
The mechanism of metallocofactor recognition of corrinoids shares similarities to Moco (molybdenum cofactor)-sensing RNA motifs known to respond to Mo- and W (tungsten)-containing compounds broadly identified in γ-Proteobacteria, δ-Proteobacteria, Clostridia, Actinobacteria, Deinococcus-Thermus species and environmental DNA samples. (600) Interestingly, the Moco-sensing RNA motifs are present downstream of the operator-promoter regions of genes regulated by molybdate-bound ModE (Figure 21), including the high affinity molybdate transporter, modABCD. (221) Both protein-based and RNA-based regulatory mechanism collectively ensure sufficient molybdate supply to biosynthesize Moco, an “expensive” cofactor. This highly conserved RNA, of unknown structure, effectively discriminates against closely related analogues of Moco, which is also triggered by the analogous tungsten cofactor (Tuco). Although clearly connected to cellular metallostasis, corrinoid and Moco sensing is dictated by specific interactions with the organic cofactor rather than specific contacts with the metal ion itself. This mode of molecular recognition also contrasts with most metalloriboswitches, which combine the influence of supramolecular chemistry and “soft” ligand (nitrogen) coordination chemistry to create coordination complexes that bind transition metals selectively while outcompeting cell-abundant MgII, required to fold these complex polyanionic and globular RNA structures. This is highlighted by the crystal structures of the MnII-sensing (Figure 22C) (609) and NiII/CoII-sensing riboswitches (Figure 22D). (224)

3.3.3. MnII/CaII-Sensing Riboswitches

The discovery of a MnII-sensing riboswitch was driven by genetic experiments first carried out in Salmonella, (225) but the chemical selectivity remained speculative until the structure of the E. coli yybP-ykoY orphan riboswitch in complex with both MgII and MnII was reported in 2015 (Figure 22C). (609) The MnII coordination chemistry in this structure suggests that that metal selectivity is achieved largely by a single “soft” coordination bond to the N7 of an invariant adenosine (A41), as the only inner-sphere MnII contact (Figure 22C). Indeed, when A41 is substituted by U41, the structure contains only MgII ions. On the other hand, the slow (millisecond) conformational dynamics of the MnII-sensing yybP-ykoY aptamer RNA revealed by single-molecule fluorescence energy transfer (smFRET) spectroscopy suggests that conformational selection, i.e., where MnII “selects” a MgII-stabilized “closed” conformation of the RNA, is an important component of metal ion selectivity. (610,611) These findings reveal how these highly structured and complex RNA motifs harness the conformational ensemble to stabilize distinct ligand-bound forms of the RNA, thus enhancing the affinity of what would otherwise be weak coordinate-covalent bonds.
A follow-up study of distinct yybP-ykoY mRNA sequences posited that the selectively of this RNA for MnII could be traced to the structure of the MnII chelate itself, which resembles the heptacoordinated MnII complex in the MnII-sensing protein repressor, B. subtilis MntR (Figure 15C). (612) However, the long coordination bonds and relatively modest resolution of the MnII-bound yybP-ykoY riboswitch structure casts doubt on this conclusion. What is more significant from this work is that the yybP-ykoY riboswitch MnII binding affinity (in a background of physiological MgII) is at least 10-fold weaker than the reported affinity of the MnII-sensing repressor MntR, a finding that parallels analogous experiments in S. pneumoniae with the MnII-sensing repressor PsaR and the yybP-ykoY riboswitch RNA. (226,477) These physiological studies are consistent with a model where riboswitches sense higher bioavailable concentration of MnII than is sensed by their counterpart protein metalloregulators, and thus likely function as a fail-safe “on” signal to prevent MnII toxicity by high cellular MnII. (226)
Interestingly, the S. pneumoniae yybP-ykoY riboswitch RNA that regulates mgtA expression binds and responds to CaII in the presence of millimolar MgII, stabilizing a compact state that elicits the same transcriptional response in vitro with E. coli RNA polymerase. (226) However, CaII is not as strong an inducer of mgtA transcription in vivo as MnII, and the extent to which CaII efflux by MgtA is beneficial for this respiratory pathogen during an infection is not known. Nonetheless, MgtA is most closely related to the P2A-type ATPase, human SPCA1, known to efflux both MnII and CaII (Section 5.3.1.1) with the coordination structures of these two metal complexes now known (see Figure 54, below). Thus, while the metal selectivity of the riboswitch appears to match that of the regulated efflux pump, the functional significance of CaII binding by this riboswitch in bacteria is not yet fully understood.

3.3.4. NiII/CoII- or FeII-Sensing Riboswitches

The NiCo riboswitch was originally discovered and functionally characterized in Clostridium botulinum and Erysipelotrichaceae bacterium (Eba) cultured in the presence of CoII. (224) This riboswitch is found in the 5′ UTR of the czcABC mRNA in these organisms, which encodes a periplasm-spanning metal efflux “gun” in Gram-negative bacteria, and is generally considered to mediate resistance against CdII, ZnII and CoII (czc) toxicity. (613,614) Thus, these initial findings were broadly consistent with a regulatory scenario that is analogous to the yybP-ykoY riboswitch described above (Section 3.3.3).
However, one puzzling feature of this initial work is that the binding affinity for NiII and CoII ions are in the low micromolar range, (224) which is significantly less sensitive than the corresponding NiII- and CoII-sensing metalloregulatory proteins that have affinities in the (sub)-nanomolar range of bioavailable metal (Figure 8A). (204,270,615,616) Indeed, more recent studies reveal that this NiCo class of metal-responsive riboswitches is overrepresented in obligate anaerobes and has recently been shown to exhibit a robust response to FeII (but not FeIII), a conclusion reached through use of a fluorescent-sensing derivative of the NiCo riboswitch. (228) Iron responsiveness was initially missed because cells were cultured aerobically, where FeII is rapidly oxidized to insoluble FeIII. (224) An RNA-based fluorescent sensor based on the NiCo RNA aptamer to detect bioavailable Fe levels in E. coli cells suggests that the Kd is actually tuned to known bioavailable FeII pools buffered in the low μM range (Figure 8A). (228) It has therefore been proposed that czcD riboswitches primarily respond to physiological FeII to mitigate iron toxicity, rather than CoII and NiII. (228,602) The transition metal selectivity of the NiCo riboswitch remains unsettled and more work is clearly warranted here.
In any case, the structure of the CoII-bound NiCo riboswitch reveals four CoII binding sites that likely nucleate formation of a four-way junction stabilized by a network of contacts that may explain how cooperative binding of CoII ions to adjacent sites might be achieved (Figure 22D). (224,607) This structure reveals at least seven inner sphere CoII coordination bonds with nucleobase nitrogen atoms, while the supramolecular chemistry of the folded RNA drives a cooperative metal binding that must contribute to fine-tuning of the regulatory response to these metals in cells. More comprehensive metal-binding studies of NiCo riboswithces from Eba and Listeria monocytogenes reveal a metal binding stoichiometry of three in solution, with metal site 4 (Figure 22D) playing no regulatory role, and likely an artifact of the high CoII concentration used in the crystallographic experiments. (607)

3.3.5. Other Inorganic Metalloriboswitches

The first crystallographic structure of a metal-sensing RNA riboswitch of any kind was the B. subtilis MgII-sensing riboswitch positioned upstream of a MgII uptake transporter. (150) This ubiquitous MgII-sensing “off” riboswitch is thought to make a major contribution to cellular MgII homeostasis in most bacterial cells. (121) Consistent with this, the metal binding affinity for MgII and noncognate cations are all weak, in the millimolar range, which ensures a MgII-specific response as this is the only metal in the cell that reaches millimolar bioavailability (Figure 8A). This is the case despite that fact that the folding of this RNA could be induced In vitro by many different divalent metal cations. (150)
The ligand-bound structure of the Thermotoga petrophila fluoride-sensing riboswitch reveals that inorganic riboswitches are not necessarily restricted to positively charged metal ions. Here, the bound fluoride is encapsulated within a junctional architecture, anchored through direct coordination to three MgII ions, which in turn are octahedrally coordinated to water molecules and five inwardly pointing backbone phosphates. (617) The selectivity of this riboswitch is unparalleled as it readily discriminates against other larger halides; as such, it has been used to develop a highly specific F– biosensor for water analysis. (618)
More recently, a new riboswitch class, exemplified by the DUF1646 motif, was characterized to selectively sense NaI and regulate the downstream expression of sodium and calcium transporters among other genes relevant to sodium biology. (604) Interestingly, these NaI riboswitches exhibit dissociation constants in the low mM range, and show no response to any other alkali and alkaline earth group metal. It has been proposed that bacterial cells exploit the specificity of these NaI-specific riboswitches alongside the c-di-AMP-sensing riboswitches to coordinate gene expression in order to mount an osmotic stress response. The structure of this riboswitch is as yet unknown but promises many surprises.

3.4. Biotechnological Applications of Metalloregulatory Systems

3.4.1. Overview

The fast-growing field of synthetic biology sees biological systems as toolboxes to be deconstructed into “parts” for subsequent repurposing. In that sense, the capacity of a regulatory “circuit” that leverages the ability of the “one-component” transcriptional regulators discussed here (Section 3.2) to sense a specific metal stimulus makes them ideal sources of diverse parts. At its most basic level, a biosensor is composed of a sensory module, which selectively detects a target analyte of interest, and a reporter module that produces a detectable output (Figure 23A). Among the myriad of possible sensory parts (functional DNA, RNA, transcription factors, enzymes, antibodies) the metalloregulatory systems discussed here have emerged as the “gold standard” for rapid and sensitive metal detection, as a result of their desirable sensitivity and selectivity profiles. (230−232)

Figure 23

Figure 23. Transcription and transcription-translation based biosensors. A) Components of the sensor and reporter modules, these parts can be employed in both WCB and cell-free contexts. The sensor module can comprise metalloregulators from any transcription factor family that act as repressors or activators of transcription. Promotor design is an important factor in biosensor performance, correctly assembling a regulatory circuit requires defining an operator sequence (blue) and a reporter gene (green). In the case of riboswitches, circuit design must incorporate the correct features to allow for correct regulation. The reporter gene encodes can encode for a transcriptional reporter such as fluorescent aptamer or reporter proteins and enzymes in the reporter module. B) Whole cell biosensor (WCB) construction based on a two-plasmid design; one plasmid constitutively expresses the regulatory element with the reporter plasmid encoding for a reporter regulated by the metalloregulator. The inducer (orange) concentration inside the cell depends on efflux/influx systems of the selected chassis. C) Cell-free biosensors that employ the regulatory function of TFs, their minimal construction is reflected in ROSALIND biosensors that are based on In vitro transcription reactions. Among the advantages of cell-free biosensors are that the sensory molecule is directly exposed to the total concentration of inducer and the ease of lyophilization and reconstitution on-site of the reactions. (269)

High Resolution Image
Download MS PowerPoint Slide

3.4.2. Utility and General Features of an Allosteric Transcription Factor

A major application of metalloregulator-based biosensors is to assess water quality, as the chemical contamination of drinking water remains a challenging analytical problem to which these biosensors might provide useful solutions. Securing global access to clean and safe drinking water is one of the main technological challenges today. (619) Benchmark technology currently used to assess water quality must be performed by experts with expensive equipment in specialized laboratories, limiting the geographical reach of water surveillance campaigns worldwide. In the last decades of the 20th century (233) biosensors have been proposed as an inexpensive, easy to use, field-deployable, and highly sensitive alternative. The emerging enthusiasm for adopting metalloregulator-based biosensors derives from their innate sensibility and selectivity, and their compatibility with inexpensive, point-of-care devices compliant with current water quality regulations. Notably, the effective metal concentrations to which metalloregulators are tuned to activate a transcriptional response inside a cell (Figure 8) are in fact similar to required limits of detection in the field, even if a design is limited by an underperforming, but easily optimized, output or signal detection strategy. (620,621) This makes applications of whole cell biosensors to biological engineering and industrial biotechnology rather straightforward. On the other hand, a major challenge of cell-free approaches is that metal selectivities of the metallosensors tend to be reinforced by the cellular milieu, i.e., the buffering molecules in cells (Section 2.2; Figure 8). This has to be taken into account in the design of such cell-free devices. (19)
Although well-characterized metalloregulators have the advantage of a well-defined metal selectivity profile, metal promiscuity (250) if present can also be leveraged to create sensors for arrays of metals where the limit of detection of each metal ion matches the desired application. (622) Moreover, an attractive feature of working with bacterial transcription factors is access to naturally occurring sensors for potentially any natural product of interest through genome mining approaches (Figure 24A, below). Improved bioinformatics tools have more robust annotation of candidate allosteric transcription factors (aTFs) in genomes (623) and have recently been used to predict a cognate DNA operator(s) suitable for circuit design. (624) Multiple databases are available to explore well-characterized transcription factors and their regulons, including regulonDB, (625) SwissRegulon (626) and GroovDB. (627) Now, thanks to decades of research into the components that make them, metal-sensing biosensor platforms are closer than ever to mainstream application in the field. (628)

Figure 24

Figure 24. Transcription factor sourcing and tuning. A) Novel genome mining tools assisted with operator and inducer predictions software complement the existing transcription factor databases to extract novel TFs for novel biosensor development. B) Rational design of metalloregulator mutants is based on prior knowledge of its regulation mechanism. In the case of the MerR family, in depth studies of its metal preference led to different design strategies such as exchanging metal binding domains and points mutations in key residues. C) Example of an established directed evolution scheme (SELIS) with dual selection optimized for transcription factor evolution, starting from a library of random TF mutants assessing their repression and induction capabilities. (629)

High Resolution Image
Download MS PowerPoint Slide

3.4.3. Whole Cell Biosensors

The simplest whole cell biosensor (WCB) functions as a traditional reporter assay (Figure 23B). Here, a synthetic regulon or “circuit” is constructed with a regulatory factor, e.g., a metalloregulator, that controls the expression of a reporter gene positioned downstream of a DNA operator/promoter region. Growth of this strain containing the gene circuit in the presence of an inducer activates transcription of the reporter gene, producing a detectable signal, as long as the inducer exceed a specified concentration, which defines the limit of detection (LOD) or sensitivity of the circuit. WCBs offer many advantages, including the low cost of growing small quantities of bacterial culture and access to a range of configurable and orthogonal output signal detection modalities, including luminescence, fluorescence and colorimetric approaches. Additionally, these circuits can be tailored for any specific application by expanding the toolbox to include amplification circuits, logic computations, signal integration (630,631) and rapid allosteric transcription factor library screening.
The development of WCBs specific for metal ions and metalloids began almost immediately with the description of the regulatory element of the ars operon, (198,632,633) and indeed, biosensors employing ArsR continue to emerge. (235,634,635) A typical ars operon contains a regulatory gene, arsR, and genes encoding for an arsenate reductase and an arsenite efflux pump (namely, asrA, arsB and arsC) that collectively mediate AsIII resistance (Section 3.2.2). A number of improvements to arsenic biosensors have been reported that highlight advantages of employing WCBs. On the one hand, the choice of strain or “chassis” was demonstrated to be a major factor in determining AsIII sensitivity (636) because the WCB signal ultimately depends on the bioavailability of the inducer in that cell, and thus how it is “tuned” to detect the analyte. (637) On the other hand, strategic choice of reporter (638) and gene circuitry design (621,639) has permitted significant gains in performance relative to previous designs. The first generation As sensors were based on ArsRs that harbor the canonical AsIII α3 coordination site (Section 3.2.2); however, this biosensor is limited to the detection of soluble AsIII species and cannot detect other As speciation. To expand the repertoire of detection of As species, biosensors based on ArsRs containing other AsIII sensing sites, e.g., AfArsR and CgArsR, were found to specifically enhance detection of organoarsenicals for example, (640) highly relevant due to their widespread use as herbicides. In an effort to detect arsenate (AsO4; AsV) directly, a Paracoccus sp. ArsRC fusion protein containing an arsenate reduction domain linked to an AsIII-sensing ArsR regulatory domain resulted in improved detection of AsV. (641) Metal promiscuity can also be leveraged, as in SaCadC, (642) while aTFs from different metallosensor families can be employed in parallel to construct biosensor arrays suitable for quantitation of many heavy metal contaminants simultaneously. (230,622) Moreover, these sensor arrays combined with DNA logic circuits may also permit the design of tools able to compensate for the lack of selectivity of a given metal if other sensors predictably respond to interferences. (19) MerR family activators (Section 3.2.3) are particularly relevant in this regard, since some exhibit a relatively high selectivity for a specific metal analyte, e.g., for PbII with CmPbrR, (643) CuI with PpCueR, (644) and AuI with StGolS. (645)
A significant strength of the WCB approach is that the anticipated biosensor metal selectivity profile is well predicted by extensive studies that provide the foundation for the set-point model, which aims to reflect metal sensitivity in cells (Section 2.2). A major limitation of WCBs is that bacterial cultures must be maintained and continuously cultured over multiple generations under antibiotic selection, which could result in the accumulation of mutations that ultimately disrupt WCB performance. Field deployment is further limited by the need to transport sophisticated equipment to the field and the use of biocontainment protocols to prevent the release of engineered microorganisms into the environment. (230,646)

3.4.4. Cell-Free Biosensors

In vitro or cell-free biosensor systems do not suffer from the same limitations of a WCB and have emerged in the past decade as a means to circumvent the use of living bacterial cells. While cell-free reactions with metalloregulatory proteins were developed soon after their discovery, (340) the output signal of initial designs proved insufficient. A number of strategies have been introduced to transduce, and more importantly, amplify output signals from transcription factors In vitro, which has catalyzed further development of these platforms. (647−649) Some leverage structural changes that occur in the metalloregulator upon ligand binding or couple DNA binding to an electrochemical signal. (650) Other strategies couple transcription factors with the wealth of established DNA signal detection platforms such as strand displacement amplification (SDA) (648) or nicked DNA template-assisted signal transduction (NAST). (647) These emerging strategies have largely been developed with nonmetal-sensing transcription factors and have been thoroughly reviewed elsewhere. (231) Exploring these strategies with metalloregulatory proteins represents a promising avenue for the future development of novel metal biosensing platforms.
In vitro cell-free transcription systems, like WCBs, leverage metalloregulators as metal detectors that elicit a transcriptional response that leads to an amplified signal. The ROSALIND platform (269) employs apo-repressors as transcription factors from different structural families (Section 3.2) to sense a wide array of metal and organic contaminants in water (Figure 23C). The platform consists of only a few purified components. These include a linear DNA template with a bacteriophage T7 RNA polymerase promoter and a DNA operator sequence that binds the transcription factor of choice, each separated by a spacer. These elements regulate the transcription of an RNA aptamer (three-way junction dimeric broccoli, 3WJdB) that when bound to an organic dye, produces a fluorescent signal. This work showed that with appropriate promoter design an aTF can regulate transcription of even the highly processive phage T7 RNA polymerase. The modularity of ROSALIND has permitted the development of transcriptional circuits controlled not only by metallosensors, including SmtB (ZnII), CsoR (Cu/ZnII) and CadC (PbII/CdII), but also those capable of detecting small organic molecules, such as TetR (tetracycline), MphR (erythromycin) and QacR (benzalkonium chloride). (269)
While ROSALIND is limited to the use of well-characterized apo-repressors, cell extract-based In vitro transcription-translation (Tx-Tl) systems permit the development of biosensors with the whole array of possible aTFs (622,651,652) and orthogonal regulatory elements, including riboswitches (Section 3.4), (653) since both protein- and RNA-based “parts” make use of the same endogenous RNA polymerase. The use of riboswitches in this way has thus far has been limited to Tx-Tl systems, and as yet only for the detection of nonmetal analytes such as fluoride anion. (653) We believe that riboswitches are a particularly unexplored area for metal biosensor development given our considerable knowledge base (598,602,654) and the comparative ease with which they can be engineered relative to protein-based factors. (655) Cell extracts also allow the use of metabolic engineering to integrate the signal from a target molecule for which a transcription factor is currently unknown. (656,657) Major limitations of the use of cell extracts arises from a generally slow “turn-on” response as a result of inconsistent manufacturing (658,659) and storage (660) of the extracts that does not allow for robust calibration required by a quantitative biosensor.

3.4.5. Engineering and Evolution of aTFs for Biosensor Applications

Currently, the construction of a biosensor based on a metalloregulator begins by defining a target metal contaminant of interest for detection. Then, the choice of a metalloregulator is made by consulting available databases that are populated with well-characterized and annotated transcriptional regulators. These databases are continuously updated, allowing the user to create a “short list” of strong candidate metalloregulators with the desired inducer specificity profile and operator DNA sequence. Candidate combinations are then evaluated using homologues of well-characterized metalloregulators to identify one that is best suited for the desired application. (661) Even then, metal binding site promiscuity and a lack of sensitivity can limit the use of naturally occurring metalloregulators for a particular application.
A key aspect of transcription factors is their tunability. (662) Rational (targeted) engineering of transcription factors has been explored as a means to improve sensibility, metal specificity and/or expand the repertoire of analytes to be detected. MerR family members have been the subject of many engineering efforts from different perspectives, facilitated by the fact that these are largely “single-site” regulators, in which metal specificity is determined, at least part, by the specific nature of the metal binding loop (MBL) (Section 3.2.3). For example, the mutation of a highly conserved serine residue in the MBL of monovalent metal specific regulators to a cysteine, based on its presence in divalent metal regulators, resulted in variants capable of sensing both HgII as well as the monovalent coinage metals (Figure 24B). (304) In another study, EcZntR metal promiscuity was further enhanced by introduction of targeted mutations in the MBL, resulting in variants with and ability to sense both HgII and CdII. (333,663,664) Other studies that simply “swapped” MBLs among MerRs with different metal selectivity profiles, however, have met with limited or unpredictable success, (665) which reinforces the idea that transition metal selectivity in MerRs is not solely determined by the first coordination shell, but also by the physical presentation of the MBL in the context of the protein scaffold. This conclusion has been reinforced by findings with a dual selection, directed evolution campaign carried out with CmPbrR. (666) Two very distinct PbrR variants were obtained with improved selectivity of PbII over ZnII. One variant harbored a C134R substitution in the MBL, while the other variant featured two substitutions (N64A/L68S) far from the MBL. These findings are consistent with the conclusion that predicting selectivity from sequence information alone remains a challenge in the biosensor design field (Figure 24B).
More recently, other selection schemes have been implemented to improve analyte inducibility and thus the biosensor response using EcArsR as a model. (639,652,667) Other robust dual selection schemes easily tailored to sensor evolution, e.g., SELIS (Figure 24C), (629,668,669) have been used to introduce novel organic ligand selectivity into a previously ligand-promiscuous transcription factor. This scheme has not yet been applied to metalloregulators and represents a promising avenue to evolve metalloregulators with virtually any desired metal selectivity profile. Although metalloregulator engineering and rational or targeted mutagenesis campaigns geared toward biosensor applications have provided valuable new information on how transition metal selectivity emerges, we believe nonguided design paths such as directed evolution or more recent “ligand-aware” artificial intelligence tools may ultimately emerge as the new standard approach for biosensor development. (670−672)

4. Metalloproteome Paralogs

Click to copy section linkSection link copied!

4.1. Overview

Metal sensors control the expression of the genes that encode proteins that allow an organism to maintain homeostasis while mounting an adaptive response to either metal limitation or metal toxicity. In the following two sections, we consider two specific adaptive responses to perturbation of the extracellular metal availability that highlight the remarkable versatility of the adaptive response in prokaryotes. These include the expression and characterization of metalloenzyme paralogs and cambialistic metalloenzymes recently described (Section 4) and specific mechanistic features, recently uncovered, of metal delivery to downstream protein clients and trafficking of metals across bacterial membranes, including intracellular organelles (see Section 5). (673,674)

4.1.1. Metalloenzyme Paralogs vs Cambialism

The general concept of a metalloenzyme paralog derives from a comparative genomics analysis of distinct species, (675) where a particular metabolic pathway, while clearly present in two organisms, appears to lack one or more enzymes in that pathway, e.g., a “pathway hole”, and yet the pathway remains fully functional. (676) Thus, there must be a workaround. In one case, a distinct enzyme has evolved via convergent evolution, or “nonorthologous displacement,” that catalyzes exactly the same reaction. (677,678) In another case, an enzyme has evolved that catalyzes a different reaction that eliminates the requirement for one or more other reactions within the same pathway, establishing an “alternative route.” (679) Metalloenzyme paralogs are a specialized type of nonorthologous displacement where an enzyme evolves to catalyze the same reaction as an obligatory ZnII-dependent enzyme, for example, but is characterized by a relaxed metal cofactor specificity profile or uses no metal at all. If distinct genes encoding the conventional or “housekeeping” vs paralogous enzymes are present in the same genome, they may well be differentially regulated by metal nutrient status of the growth conditions. An early example of this was in fact GTP cyclohydrolase-I (Section 4.3.3). Known enzyme paralogs often catalyze what might be considered “permissive” chemistry, where the metal cofactor does not undergo a change in oxidation state, e.g., in hydrolytic enzymes or other kinds of Lewis acid catalysts.
Cambialism, in contrast, defines the specialized case where the same metalloenzyme can be metalated with distinct metal cofactors without a significant loss of catalytic activity. (680) This might occur in response, at least in part, to nutrient metal availability, (7,681,682) a strategy that may well be more widespread than previously anticipated. (683) Remarkably, the work summarized here on the superoxide dismutases (SODs) and ribonucleotide reductases (RNRs) reveal that cambialism and more generally metalloenzyme paralogs have evolved in enzymes that catalyze more “restrictive” chemistry, where a change in oxidation state and even access to high-valent intermediates, is involved. Here, MnII/FeII cambialism is superimposed on paralogism, where one member of a series of paralogous enzymes exhibits the distinct property of cambialism.

4.1.2. Expression of Metalloprotein Paralogs in the Metal Limitation Response

Although Zn and Fe are essential nutrients for growth, bacteria often colonize Zn- and Fe-deplete environments. (684,685) Pathogenic bacteria are particularly susceptible to transition metal starvation by the action of metal scavengers including calprotectin (CP) and related S100 proteins. CP often induces components of the Zur and Fur regulons, which is hypothesized to result in collateral undermetalation of the proteome (Section 3.2.6; Figure 8B). (10,686) In the specific case of zinc starvation, bacterial cells upregulate the expression of protein paralogs of obligatory ZnII-dependent enzymes that function in key physiological pathways. (687,688) These paralogs include ribosomal proteins, global transcriptional regulators, enzymes of central carbon metabolism, one-carbon metabolism and carbon fixation, and perform tRNA modification and aminoacylation, nucleotide (purine, pyrimidine and pyridine) biosynthesis, heme biosynthesis, amino acid biosynthesis and bacterial cell wall remodeling. (689) In the case of the ZnII enzymes, the regulation of the expression of these paralogs is believed to be a form of “Zn sparing” that mobilizes ZnII for other functions, (687) conserving a severely restricted nutrient for essential processes.
We classify the functional characteristics of currently known paralogs into two distinct groups (Figure 25A): (1) paralogs that have lost metal coordinating residues and may or may not dispense of the need for metal altogether (Category 1), and (2) paralogs that are metal-promiscuous and require a metal for catalysis but may or may not retain function with ZnII (Category 2). As discussed above, enzyme paralogs can arise as a result of convergent evolution, in which structurally distinct solutions to the same reaction chemistry evolve (Figure 25B). Each category exhibits distinct mechanistic features in adaptation to low Zn so that metabolic flux through a major biosynthetic pathway can be preserved. Here, we exploit one genomic region of the P. aeruginosa Zur regulon to illustrate these general principles (Figure 25C). This region of the genome harbors two distinct Zur-boxes that regulate the expression of different sets of genes (an operon). The first divergently transcribed operon encodes for two proteins of unknown function, two candidate COG0523-family putative Zn metallochaperones, the DksA2 paralog, and a TonB transporter component that may be involved in Zn uptake across the outer membrane in this Gram-negative bacterium, through the outer membrane porin ZnuD. The second divergently transcribed operon encodes the paralogs FolEIB, PyrC2, AmiA, and a γ-class carbonic anhydrase. Finally, the intriguing cases of two cambialistic enzymes, superoxide dismutase (SOD) and ribonucleotide reductase Ib (RNRs), are discussed.

Figure 25

Figure 25. Conceptual framework defining the different categories of Zur-regulated paralogs of Zn-requiring enzymes that catalyze the same reaction, as an example. A) Gene duplication and divergent evolution leads to extant Zn-requiring and non-Zn-requiring enzymes. At higher intracellular Zn (gray spheres) concentrations, the canonical enzyme (yellow) can acquire Zn while the paralog (cyan) is repressed by Zn-bound Zur. When Zn levels drop in the cell and the canonical enzyme is inactivated via loss of Zn, a Zur-regulated paralog is expressed where it functions in the absence of any metal (category 1), or with an alternative metal to Zn (purple sphere, category 2). B) Enzyme paralogs can also arise as a result of convergent evolution in which structurally distinct solutions to the same reaction chemistry evolve, where the Zn enzyme (cyan) might be subject to inactivation via loss of Zn; the others may be regulated by Zur (not shown). C) Part of the P. aeruginosa Zur regulon comprised of two divergently transcribed operons is shown, with locus tags (gene IDs; PAxxxx) and inferred functional assignments. Two known Zur boxes are indicated. (690) Genes that encode paralogs as defined here are shaded cyan, COG0523-family candidate Zn metallochaperones shaded red, (691) and genes of unknown function are shaded gray.

High Resolution Image
Download MS PowerPoint Slide

4.2. Category 1: Metal-Independent Paralogs

4.2.1. Ribosomal C– Paralogs

A classic example of Zur-regulated paralogs comes from the modification of a structural, not a catalytic, Zn binding site. (692,693) This is exemplified in the Zn sparing phenomenon of Zur-regulated, nonenzymatic ribosomal protein paralogs. These paralogs lack 4 conserved cysteines that form a pair of CxxC Zn binding motifs (called “C–” ribosomal proteins) present in their Zn binding counterparts (“C+,”). (687,692) The model organism E. coli encodes two C+ ribosomal proteins, L31 and L36, encoded by the genes rpmE and rpmJ, respectively, and two C– paralogs L31B and L36B, encoded by the genes ykgM and ykgO (Figure 26A). (694) E. coli Zur achieves upregulation of L31 through a “repression of a repressor” mechanism. Zur regulated L31B, expressed in zinc-deficient conditions, in turn represses L31 expression by binding to its mRNA 5′ UTR. In the presence of sufficient zinc, the C+ ribosomal protein is expressed while the C– paralogue is repressed. (695) This allows the organism to deploy this significant new intracellular pool of bioavailable Zn for other crucial processes as ribosomes are among the most abundant molecular assemblies in the cell. (694) Across the bacterial kingdom there exists a large diversity of C– paralogs, with seven identified thus far in Zur regulons; some organisms encode for no C– paralogs while others possess up to six. (430) These paralogs are sprinkled throughout the mature ribosome, spanning both large and small subunits inside and out, and are generally incorporated into both intermediate and late 50S ribosomal assembly intermediates, but not early ones; (696) the significance of this finding is unknown.

Figure 26

Figure 26. C– ribosomal protein paralogs. A) Ribbon representations of M. smegmatis L31 and the low-zinc paralog, L31B. (697,698) Conserved features of C+ vsC– subunits with Zn-binding Cys residues labeled highlighted with a yellow sphere. B) Space-fill model of the M. smegmatis 70S ribosome shaded gray with the indicated ribosomal subunits highlighted in yellow and cyan (PDB: 5O61 C+ ribosome and 6DZI C– ribosome (697,698)). The five C+ subunits (yellow) associated with Zur-regulated C– paralogs (cyan) are shown.

High Resolution Image
Download MS PowerPoint Slide
Structurally, the C– proteins not only lack the Zn binding Cys residues, considered essential for their C+ counterpart function, but also contain insertions, extensions, and divergent sequences. Both C+ and C– ribosomes from M. smegmatis have been structurally characterized (697,698) and these ribosomes have been evaluated for functional differences (Figure 26B). Specific C– insertions protect the ribosome active site by inhibiting the efficacy of antibiotics. (699) C– and C+ ribosomes appear to have similar elongation rates and error rates, and may be more active under Zn-limiting conditions. (700) However, C– ribosomes exhibit distinct translation profiles and are characterized by slower translation initiation rates relative to C+ ribosomes. (700,701) The extent to which C– subunits confer additional features on the ribosome remains to be fully evaluated, but an exciting hypothesis is that ribosome remodeling during Zn limitation may selectively upregulate the translation of cellular transcripts that enable physiological adaptation to nutrient limited conditions.

4.2.2. DksA/DksA2

The stringent response is induced by general nutrient starvation resulting in hyperphosphorylation of nucleotides which signals an alleviation mechanism(s) that involves a global change in gene expression patterns ultimately leading to stasis or quiescence. (702) Major signaling molecules in the bacterial stringent response are the alarmones guanosine pentaphosphate (guanosine 5′-triphosphate, 3′-diphosphate; pppGpp), and guanosine tetraphosphate (guanosine 5′-diphosphate, 3′-diphosphate; ppGpp), collectively referred to as (p)ppGpp. These are synthesized from GTP or GDP, respectively, and ATP, as the source of 3′-pyrophosphate moiety. These alarmones severely deplete cellular GTP pools, and function either by inhibiting enzymatic activity directly by binding to a GTP or GDP binding site in competition with this substrate or “signal” within the transcriptional machinery itself. (703,704) The stringent response is also dependent on specialized proteins, including DksA. (705) DksA is a broadly conserved RNA polymerase regulatory factor and is present in most bacterial species, (706) with the prominent exception of Firmicutes. DksA2 is a paralog that is found in Zur regulons of β- and γ-proteobacteria (688) and has been most extensively characterized in P. aeruginosa. PaDksA2 shares 34% sequence identity with PaDksA, and both proteins adopt nearly identical folds (Figure 27A). Moreover, during the evolution of Salmonella from an environmental bacterium to a vertebrate enteric pathogen, a number of virulence genes were genomically colocalized into an ancestral DksA regulon, highlighting an important role of DksA as a pathogenicity regulator. (707)

Figure 27

Figure 27. DksA transcription factor. A) Ribbon representation of the RNAP–DksA binary complex (PBD: 5W1T). Motifs of RNAP (domains in white and σ70 in black) binding (p)ppGpp are indicated with an arrow. (709) Right, ribbon structures of E. coli DksA (yellow, PDB: 1TI1) (710) and P. aeruginosa DksA2 (cyan, PDB: 4IJJ), (715) and the DksA Zn ion shown as a gray sphere. The globular (G), coiled-coil (CC), and C-terminal helix (CTH) domains are highlighted. Tetrathiolate Zn binding site of E. coli DksA with Zn coordinating Cys residues labeled and shown in stick, with the Zn ion shown as a gray sphere. Analogous metal binding site region of P. aeruginosa DksA2, with the two conserved Cys that from a disulfide bridge labeled and shown in stick representation. B) Conserved primary structural features of the global transcriptional regulator DksA and DksA2, centered only on the DksA metal binding region. DksA Zn-binding residues are labeled in yellow and highlighted with a yellow sphere and the DksA2 disulfide bridge is labeled with a bracket.

High Resolution Image
Download MS PowerPoint Slide
Although DksA is constitutively expressed, it is known to function only when (p)ppGpp and/or ROS/RNS are abundant in cells. (708,709) Under these conditions, DksA binds directly to RNA polymerase and differentially regulates transcription from selected promoters of genes encoding proteins involved in translation or amino acid biosynthesis and uptake. (706,710,711) The combination of DksA and (p)ppGpp in the stringent response inhibits ribosome production, thereby decreasing amino acid consumption by translation and increasing amino acid biosynthesis, as global efforts to restore nutrient balance in the organism. In fact, RelA, the synthase that synthesizes (p)ppGpp binds to the ribosome directly, particularly those ribosomes that harbor an unacylated tRNA in the A-site, which occurs as a result of amino acid starvation, further stimulating the cellular production of (p)ppGpp. The DksA-dependent regulatory profile that is independent of (p)ppGpp and connected to ROS/RNS stress down-regulates both the ribosome biogenesis machinery and amino acid biosynthesis, while also promoting redox buffering and repair of biomolecules, thus resulting in broad cellular resistance to these conditions. (708,712)
DksA is a small protein with an N-terminal globular domain (NTD), an extended coiled-coil (CC) domain, and a C-terminal α-helix (CTH) (Figure 27A, right). (710) The NTD is linked to the CTH by a tetrathiolate, tetrahedral ZnII binding site. (713,714) Substitution of the ZnII coordinating Cys residues disrupts DksA activity in cells. (711,715,716) The four Zn-coordinating Cys residues have been shown to be susceptible to oxidation by various reactive oxygen and reactive nitrogen species (ROS and RNS) which enhances global DksA-mediated transcriptional inhibition. (708,716) Reaction with ROS and RNS species results in the release of Zn, and treatment with reducing agent after oxidation results in nearly full restoration of DksA secondary structure. These findings suggest that the primary role of Zn is not that of a structural cofactor, but rather as a sensor of redox stress. This oxidation event also leads to potent transcriptional inhibition in the absence of (p)ppGpp, consistent with its assignment as a regulatory metal site that tunes ROS and RNS sensing by DksA. (712)
PaDksA2 possesses only two cysteines of the four Zn-coordinating Cys present in DksA (C96 and C117), which are precisely analogous to the redox-sensing cysteines C114 and C135 in EcDksA (Figure 27B). Loss of two or three Zn coordinating cysteines (some DksA2 paralogs possess only one of the four Cys observed in EcDksA) would of course abolish Zn binding in the cell, and thus would effectively “spare” this Zn for another purpose. (715) In P. aeruginosa the ΔdksA strain exhibits a growth phenotype on Zn-replete minimal media since Zur will repress the expression of dksA2 under these conditions. (716) As expected, this phenotype is suppressed under low Zn, where dksA2 is upregulated. Indeed, DksA2 has been shown to functionally complement DksA in vivo and In vitro to varying degrees. (708,716) Additionally, a P. aeruginosa strain that expresses only DksA (a ΔdksA2 strain) exhibits inhibited growth on Zn-deplete minimal media, clearly indicative of a disruption of DksA function specifically under conditions of low Zn.
A consensus model holds that DksA becomes undermetalated under zinc limiting conditions which results in an increased sensitivity to ROS/RNS and an impaired alarmone-dependent adaptive response. Under these conditions DksA2 is upregulated, thus restoring the stringent response. Zn coordination by DksA may attenuate its redox-sensing activity, reserving the redox-sensing Cys pair to respond only under conditions of chronic ROS stress, where cellular reductant pools become depleted. (708) Consistent with this, the Zn-free DksA2 is 50-fold more reactive toward H2O2 In vitro and oxidation-induced transcriptional repression is far more sensitive to low H2O2 in cells expressing only DksA2 vs cells expressing only DksA. (708) Kinetics studies reveal that the Zn-bound DksA requires significantly higher ROS to enhance transcriptional repression since the major cellular reducing thiol glutathione (GSH) preferentially scavenges ROS under these conditions; in striking contrast, DksA2 is a far better ROS scavenger than GSH itself. (708) This results in reduced survival in macrophage models of infection of bacterial cells that express only DksA2 compared to those expressing only DksA, as (p)ppGpp-independent transcription inhibition occurs at lower ROS levels. Taken together, dispensing of the Zn binding site in DksA2 is a remarkably complex evolutionary adaptation that appears to integrate the bacterial adaptive response to Zn restriction and chronic ROS exposure that may well be coincident in the infected host.

4.3. Category 2: Metal-Promiscuous Paralogs

4.3.1. QueD/QueD2

Transfer RNA (tRNA) modifications are widespread in all kingdoms of life and are regarded as global regulators of translational capacity and important markers of nutritional status. (717,718) Queuosine (Q) is a hypermodified, 7-deazaguanosine incorporated into the 34th or “wobble” position of the anticodon loop of four specific tRNAs (Asn, Asp, Tyr and His) that decode synonymous NAC and NAU (N = A, U, C or G) codons. (719) Q-tRNAs bind more tightly to the translating ribosome, although the effects vary widely. The Q modification in eukaryotic cells impacts the decoding of NAU codons while decreasing amino acid misincorporation, thus affecting translation accuracy and speed. (720) The loss of Q results in weaker codon-anticodon pairing of NAU codons (called Q-decoded codons) which can also lead to ribosome stalling in these cells. (720−726) The effects of Q loss in bacteria may well be more complex, and strongly growth condition-dependent, but can lead to biased translation and proteome remodeling, particularly under the combined effects of nutritional stress and antibiotic stress. (727,728) This has led to the concept of “modification-tunable transcripts (MoTTs)” (717,727) which drives a global adaptive response in bacteria, potentially enhancing virulence or endowing an organism some other growth advantage in the infected host. (727)
Only bacteria possess the ability to de novo synthesize queuosine-tRNA from GTP, with higher eukaryotes dependent on the salvage of queuine (the nucleobase) from the microbiota. (92) The Q-tRNA biosynthesis pathway is complex and depends on no fewer than seven enzymes that require transition metals as cofactors. (729−735) Nutrient metal starvation negatively impacts flux through the queuosine-tRNA biosynthetic pathway in Acinetobacter baumannii while giving rise to a more pleiotropic response in E. coli. (9,736) The committed step in the Q-tRNA biosynthesis pathway is performed by a 6-carboxytetrahydropterin (CPH4) synthase, QueD (737) or a QueD paralog termed QueD2. (688) Bacteria either encode a housekeeping QueD or a QueD2 that becomes cell-abundant under conditions of ZnII restriction, (738) or alternatively encode both a constitutively expressed QueD and a Zur-regulated QueD2. Mutant strains of different organisms lacking the housekeeping enzyme could be complemented by overexpression of a paralog, revealing a conserved function. (9,739)
Both QueD and QueD2 are obligate metalloenzymes, with QueD copurifying and structurally characterized with tightly bound ZnII. (729,740) Recent work reveals that both QueD and QueD2 are rather metal-promiscuous, with each possessing comparable specific activities with MnII, FeII and ZnII. Each is a member of the tunnel fold (T-fold) superfamily characterized by an oligomeric assembly state that forms a central β-barrel that resembles a tunnel. This family consists of two β2nαn rings (n ≥ 2) that join in a head-to-head fashion, with the symmetry-related active sites bridging the two rings. QueD possesses a conserved Cx3HGH motif containing the essential catalytic nucleophile, Cys27 in EcQueD and two His (His31 and His33) that coordinate the ZnII (Figure 28A). (740) An additional conserved residue, His16, contributes the third protein-derived ligand, with two water molecules or two oxygen atoms from the substrate, intermediates or product completing the five-coordinate trigonal bipyramidal ZnII site.

Figure 28

Figure 28. 6-carboxy-5,6,7,8-tetrahydropterin (CPH4) synthase QueD. A) Top-down (left) and side (right) views of the ribbon structure of the hexameric E. coli QueD (yellow, PDB: 4NTM). (740) A single protomer is colored, the remaining protomers are shaded gray, with the Zn ions (black) shown. The product 6-carboxy-5,6,7,8-tetrahydropterin (CPH4) binding site is highlighted. Close up view of the E. coli QueD Zn binding site with conserved catalytic and metal binding residues labeled and shown in stick (highlighted by the yellow circles), and the Zn ion represented by a black sphere. The catalytic nucleophile, C27, is highlighted with a red circle. The sepiapterin-like intermediate is shown as green CPK stick. B) Top-down (top) and side (middle) views of the ribbon structure the octameric A. baumannii QueD2 (cyan, PDB: 7 V0F). (9) Close up view of the A. baumannii QueD2 Zn binding sites 1 and 2 (bottom) with conserved catalytic (C23, red; H68′, blue circle) and metal binding residues labeled (blue circles) and shown in stick, with water molecules shown as small red spheres coordinated to Zn1, and the Zn ion represented by a black sphere. C18 (green circle) is uniquely conserved in QueD2 genes as either a Cys or a His.

High Resolution Image
Download MS PowerPoint Slide
The QueD2 paralog exhibits very low (<20%) sequence identity relative to the housekeeping QueD and harbors a one-residue insertion in the catalytic Cx3HGH motif of QueD to create the Cx4HGH motif of QueD2. A recent structure of QueD2 from Acinetobacter baumannii reveals a distinct octameric assembly state (two stacked tetrameric rings) relative to the E. coli QueD hexamer. (9) In contrast to QueD, there are two high occupancy ZnII binding sites in the structure of AbQueD2 (Figure 28B). One corresponds roughly to the active site of EcQueD (site 1), with the second site (site 2) more peripheral and solvent exposed, some ≈14 Å away and coordinated by the active site nucleophile, Cys23, an invariant Cys specific to QueD2s, Cys18 (this is a His in the other major cluster of QueD2 sequences), and two other poorly conserved residues derived from an adjacent subunit within a tetrameric ring. The inserted amino acid in the QueD2 Cx4HGH motif displaces the Cys nucleophile far from the catalytic site, while shifting the metal coordinating HGH motif (His28 and His30 in AbQueD2) further into the β2 strand relative to EcQueD. (9) In addition, His13 in AbQueD2 (analogous to metal ligand His16 in EcQueD) does not coordinate the ZnII and instead incorporates His 68’’, analogous to His71’’ in EcQueD, one of the catalytic bases in the second half of the reaction. (740) Activity assays reveal that as expected, ZnII binding to site 2 inhibits turnover by QueD2, which suggests that the active site 1 may well be conformationally “plastic” or dynamically heterogeneous and must “fold” to accommodate substrate, in striking contrast to the housekeeping enzyme QueD. (9)
What then is the function of metal site 2 in this low-zinc paralog QueD2? KZn for this site is quite weak and it is unlikely to be occupied under conditions of zinc starvation or even under zinc-replete conditions (+ΔΔG for metalation; Figure 8B). The conserved site 2 ligand, Cys18 in AbQueD2, inhibits the rate of ZnII dissociation, consistent with a function as a “trap” of exchange-labile metal in coordinately plastic first coordination shell. (9) Even more significantly, AbQueD2 possesses an ≈100-fold higher catalytic activity when bound to FeII in a way that is not dependent upon Cys18. How this high activity with FeII is accomplished is unknown, but suggests a significant alteration in the active site structure that specifically facilitates the second step of this two-step reaction, the release of acetaldehyde. (9) In any case, these features make it possible for an organism expressing a QueD2 to adapt quickly to niches where ZnII is deplete and FeII remains replete. This complex, multifaceted evolutionary adaptation to metal restriction, from metal “retention” to classical Zn sparing, represents a response that can be tuned to any flavor of metal starvation. (9)

4.3.2. PyrC/PyrC2

All members of the amidohydrolase superfamily share similar catalytic functions, enabling the reversible ring formation/opening of cyclic amides. (741) PyrC is a dihydroorotase (DHOase) that catalyzes the reversible interconversion of carbamoyl aspartate (CAA) and dihydroorotate (DHO) and functions in the third step of de novo pyrimidine biosynthesis, just downstream of the committed step catalyzed by another Zn-metalloenzyme, aspartate transcarbamoylase. (742) As such, PyrC is an essential enzyme. (743−745) Historically, bacterial DHOases can be grouped into two classes and a representative member of each has been structurally characterized. These are the 45 kDa class I PyrC enzymes largely found in Gram-positive bacteria, e.g., Bacillus anthracis (Figure 29A) (746,747) while the 38 kDa class II enzymes are generally found in Gram-negative bacteria, including the prototypical enzyme from Escherichia coli (Figure 29B). (748,749) They require a divalent metal ion for enzymatic activity, thought to be ZnII in the cell; ZnII and CoII typically exhibit maximal activity while other divalent metals result in inhibition. (749−752) While both enzyme classes harbor a nearly isostructural binuclear ZnII site where each ZnII exhibits a trigonal bipyramidal geometry containing four His and an Asp, the sixth ligand is an Asp in class I PyrC and a post-translationally modified carboxylated Lys in class II DHOases. (746−749)

Figure 29

Figure 29. Binuclear zinc metalloenzyme dihydroorotase PyrC. A) Ribbon representation of the structure of B. anthracis PyrC (4YIW) (747) with the dimerization interface indicated by the dashed line, and the CAA-bound Zn binding site (close up) with Zn binding residues labeled and shown in stick, the CAA substrate shown in green stick, and Zn ions shown as black spheres. B) Ribbon representation of the structure of E. coli PyrC (1XGE) (754) with the dimerization interface indicated by the dashed line, and the DHO-bound Zn binding site (close up) with Zn binding residues labeled and shown in stick, the DHO substrate shown in green stick, water molecules shown as red spheres, and Zn ions shown as black spheres. C) Ribbon representation of the structure of P. gingivalis PyrC (PDB: 2GWN) with the putative dimerization interface based on homology to B. subtilis PyrC indicated by the dashed line, and the Zn binding site (close up) with Zn binding residues labeled and shown as sticks, water molecules shown as red spheres, and Zn ions shown as black spheres. D) Conserved primary structural features of PyrCs belonging to class I (represented by B. anthracis enzyme), class II (represented by E. coli enzyme), and class III (represented by P. gingivalis enzyme), jointly with PyrC2 (represented by P. aeruginosa) comparing the different Zn coordinating residues in each class. Class I Zn coordinating residues are highlighted in red, class II Zn binding residues are highlighted in yellow, and class III residues are highlighted in cyan, with the ligands corresponding to metal sites 1 and 2 indicated.

High Resolution Image
Download MS PowerPoint Slide
The Zur-regulated paralog PyrC2 is structurally uncharacterized, but is found in Zur regulons of Gram-negative bacteria, e.g., P. aeruginosa (Figure 25C) and B. cenocepacia. In these organisms, the housekeeping PyrC is a class II enzyme that is not regulated by Zur or Zn starvation. P. aeruginosa PyrC2 is distinct from P. aeruginosa PyrC, with just 10% sequence identity, but notably exhibits 30% sequence identity to the class I B. anthracis PyrC. Furthermore, P. aeruginosa PyrC2 has lost the most N-terminal of the Zn1-coordinating His residues, which is replaced by a Gln that is conserved in other Zur-regulated PyrCs (Figure 29D). (688) Interestingly, the structure of Gram-negative Porphyromonas gingivalis PyrC, a third DHOase denoted class III (Figure 29C), reveals features of both class I (sequence similarity and an N-terminal extension) and class II (a carboxylated Lys as a bridging ligand, instead of an Asp) enzymes. We view this structure as an excellent model for Zur-regulated PyrC2 given the high (50%) sequence identity and the presence of the same site 1 Gln substitution that is found in the P. aeruginosa PyrC2 paralog. This suggests that class III PyrCs may well contain other paralogs that may function better under Zn-limited conditions, as yet unknown. PyrC may be undermetalated under conditions of low Zn, thereby necessitating the upregulation of PyrC2 to maintain flux through this pathway; however, the ability of a ΔpyrC2 strain to grow in Zn-limited conditions has not yet been evaluated. The difference between class I and II coordination spheres might relax the metal preference of PyrC2 allowing MnII to bind, for example, (753) or alter Zn affinities or kinetic labilities. Alternatively, the upregulation of PyrC2 simply forces metabolite flux toward pyrimidine biosynthesis by mass action in this reversible enzyme.

4.3.3. FolE1A/FolE1B

GTP cyclohydrolase-IA (GCYH-IA; FolEIA or simply FolE) is a ZnII metalloenzyme that catalyzes the conversion of GTP to 7,8-dihydroneopterin triphosphate (H2NTP). H2NTP is the first intermediate in the de novo biosynthesis of tetrahydrofolate (THF, folate) in all kingdoms of life; in bacteria, H2NTP is a true metabolic branch point, serving as the substrate for QueD/QueD2 in the biosynthesis of the queuosine-tRNA (Section 4.3.2). (755) FolEIB (GCYH-IB; also known as FolE2), in contrast, has been described as a metal-cambialistic paralog that can catalyze the same reactions when metalated with ZnII or another divalent metal. (677,678) Genes encoding FolEIB, like QueD2, are not exclusive to Zur regulons and in some organisms, e.g., Staphylococcus aureus, only folEIB is present. Like QueD/QueD2, both FolEIA and FolEIB are members of the T-fold superfamily, (678) and are characterized by very low pairwise sequence identity and distinct oligomeric assembly states, i.e, dimer of pentamers vs a dimer of dimers double ring-shaped architectures in FolEIA and FolEIB, respectively (Figure 30A,B). As apoproteins, both FolEIA and FolEIB are inactive and activity can only be observed by metalation of the active site as revealed by mutagenesis and enzymatic activity assays. (678,756)

Figure 30

Figure 30. GTP cyclohydrolase I, FolE. A) Top-view (left) and side view (middle) of the ribbon representation of the decameric T. thermophilus FolE (1WUR) (764) with a single protomer shaded in yellow ribbon, with the remaining protomers shaded gray. Active site region (close up) with Zn binding residues labeled and highlighted in yellow stick, the catalytic intermediate mimic 8-oxo-GTP* shown in green stick, and the Zn ions shown as black spheres. B) Top-view (left) and the side view (middle) of the ribbon representation of the tetrameric N. gonorrheae FolE2 (3D2O) with a single protomer shaded in cyan ribbon, with the remaining protomers shaded gray. Active site region (close up) with Zn binding residues labeled and highlighted in cyan stick, the catalytic intermediate mimic 8-oxo-GTP is labeled as GTP* and shown in cyan stick, and the Zn ions shown as black spheres. S-nitrosated C149 is also shown in NgFolE2. C) Conserved primary structural features of FolE and FolE2 in the active site region. The active site acid (which donates a proton) is labeled Z, while X and Y are metal ligands along with a conserved Cys. Note that the FolE2 Y (not shown in this region) and X coordinating ligands (panel B) and the Z functional groups in the primary structure of T. thermophilus FolE and N. gonorrheae FolE2 are significantly displaced from one another. (4)

High Resolution Image
Download MS PowerPoint Slide
Both enzymes coordinate ZnII in a distorted tetrahedral geometry as revealed by the structures of E. coli FolEIA and the Neisseria gonorrheae FolEIB (Figure 30A,B), with three protein-derived ligands and a mimic of the first catalytic intermediate, 8-oxo-GTP (GTP*) coordinated to the metal. (756,757) This open coordination site in the absence of substrate analog is filled by OH–, which is poised to attack the electrophilic C8, which itself is activated by a nearby His (Z residue; Figure 30C). FolEIA features a Cys2His-OH– coordination site while in FolEIB, one of the Cys (C179 in FolEIA; the Y ligand) is replaced with a glutamate (E201’) from a protomer in the opposite dimer in this dimer-of-dimers architecture, thus creating a “harder” coordination site. (4) While FolEIA is thought to be an obligatory ZnII enzyme and clearly copurifies with ZnII, this has not been rigorously established. In contrast, FolEIB enzymes from B. subtilis, Cupriavidus metallidurans and N. gonorrheae generally have lower activity when metalated by ZnII, significant activity with MnII, FeII and CoII and low activity with NiII, all trends consistent with the observed change in the first coordination shell in FolEIB vs FolEIA. (678,758) Interestingly, some FolEIB enzymes are reported to be active with MgII, to an extent on par with that of CoII, which is puzzling given the presence of clearly “softer” ligands in both FolEIB coordination shells.
Both FolEIA and FolEIB catalysis feature classic Lewis acid chemistry and proton transfer-catalyzed ring-opening and ring-closing rearrangements, involving several pentacoordinate-metal intermediates, with the release of formate and H2NTP. Interestingly, inspection of the active site regions of both FolEIA and FolEIB suggests the possibility that posttranslational modification of one or more residues might have a regulatory role. Indeed, NgFolEIB, as purified, features additional electron density on the nonliganding cysteines, C149 (Figure 30B) assigned as an S-nitrosated Cys essentially positioned “on top” of the active site. (759) Mutation of this Cys to Ala or Ser largely inactivates the enzyme, which suggests that S-nitrosation may be a protective mechanism that preserves FolEIB activity under conditions of nitrosative stress. The evolution of FolEIB in bacteria may be reflective of a GCYH that evolved under the influence of dual stressors, oxidative/nitrosative stress and ZnII limitation, reminiscent of DskA2 vs DksA (Section 4.2.2). (715)
In any case, the significant activity of FolEIB with divalent metal ions other than ZnII uncovers a clear evolutionary adaptation for bacterial survival under conditions of Zn starvation, given the universality of one-carbon (1C) metabolism in all cells. (760,761) Indeed, this adaptation may well have several regulatory layers, since in B. subtilis, the housekeeping enzyme FolEIA is reported to be a target of the COG0523 superfamily metallochaperone, ZagA (Section 5.1.4). (761) ZagA is reported to sustain folate biosynthesis under conditions where folEIB is not expressed, as part of the graded response to increasing degrees of cellular zinc limitation. (761,762) A putative protein–protein interaction between ZagA and FolEIA appears to be stimulated by the alarmone and purine derivative ZTP (5′-aminoimidazole-4-carboxamide-1-β-D-ribofuranosyl-5′-triphosphate; AICAR triphosphate), (763) which accumulates in cells under conditions of folate deficiency. (758) Mechanistically how ZagA impacts FolE catalytic activity is not yet known since FolEIA binds ZnII essentially irreversibly and may well be kinetically trapped, since it is difficult to remove with chelators and thiol modifying agents (Section 5.1.4). (758)

4.3.4. Carbonic Anhydrases

Carbonic anhydrase (CA) is the prototypical zinc metalloenzyme that catalyzes the reversible hydration of gaseous CO2 to bicarbonate anion, HCO3–, and a proton (H+). CAs therefore play a crucial role in equilibrating CO2 and bicarbonate inside the cell. The central importance of CAs in global carbon capture, pH regulation, pyrimidine biosynthesis and in providing a universally employed substrate for all biotin-dependent carboxylation reactions (765) underscores the remarkable evolutionary diversity of CAs. Indeed, there are now eight known gene families that encode structurally divergent classes of CAs: the α-, β-, γ-, δ-, ζ-, η-, θ-, and ι-CAs. (766,767) All human CAs the α-class enzymes, where at least 12 active isozymes have been described that differ in oligomeric assembly state, subcellular localization, tissue expression, and kinetic properties while all sharing ZnII as an essential cofactor. (768) In bacteria, the major structural classes are the α, β and γ enzymes while the ι-CA is present in marine diatoms. (769) Bacterial α-CAs are generally are generally periplasmic (in Gram-negative organisms) and function as monomers or dimers, while employing ZnII as the Lewis acid cofactor in a His3-H2O motif like the mammalian CAs (Figure 31A). β-CAs are typically dimers, tetramers or octamers that are also obligate ZnII metalloenzymes that coordinate ZnII with two Cys, one His and an open coordination site (Figure 31B). β-CAs often exhibit strongly pH-dependent activity in which an Asp is recruited into the open coordination site at high pH, thus abolishing turnover. γ-CAs are adopt a trimeric left-handed β-helical architecture and utilize the same His3-H2O motif as α-CAs (Figure 31C), but are reported to be FeII-dependent enzymes that are also active with ZnII and CoII. One ι-CA has been reported to prefer MnII over ZnII, although the extent to which this is broadly the case is not known. (770)

Figure 31

Figure 31. Carbonic anhydrases likely arose via convergent evolution. A) Ribbon representation of the homodimeric α-carbonic anhydrase (4X5S) (773) with the highlighted monomer labeled in red and other protomer colored gray. α-carbonic anhydrase active site (close up, left) with Zn binding residues shown as red sticks, Zn ion as a black sphere, and the acetazolamide (AZM) inhibitor shown as green sticks. B) Ribbon representation of the tetrameric β-carbonic anhydrase (2A5V) (774) with the highlighted monomer labeled in yellow and additional protomers colored gray. β-carbonic anhydrase active site (close up; left) with Zn binding residues shown as yellow sticks, Zn ion as a black sphere, and SCN– molecule shown as green sticks. C) Ribbon representation of the trimeric γ-carbonic anhydrase (3KWC) (775) with the highlighted monomer labeled in cyan and additional protomers colored gray. γ-carbonic anhydrase active site (close up; left) with Zn binding residues shown as cyan sticks, Zn ion as a black sphere, and water molecules as red spheres.

High Resolution Image
Download MS PowerPoint Slide
Most bacteria encode more than one CA, often but not always derived from all three major classes (cytoplasmically localized β- and γ-class enzymes feature prominently in clinically relevant pathogens), with some regulated by nutrient stressors. In the few instances where it has been documented that cellular expression of a CA changes with ZnII restriction (738) and/or represents a bona fide Zur target (as in P. aeruginosa) these enzymes tend to belong to the β- and γ-classes, with bacterial pathogens often encoding multiple β-CAs with apparently nonredundant functions. (690,738,769) We believe that Zur-regulated expression of a γ-CA may represent a case of Zn-sparing, since this enzyme is reported to be metal-promiscuous with FeII the preferred metal cofactor when prepared under anaerobic conditions. (771,772) An in depth study of metal preferences across the different families of CAs and, in particular, in the case of Zur-regulated γ-CAs is clearly warranted to correctly identify how this essential metabolism has adapted to microenvironmental niches poor in nutrient metal availability.

4.3.5. Cambialistic Superoxide Dismutases

The superoxide dismutases (SODs) detoxify superoxide, a reactive oxygen anion radical produced in all organisms exposed to oxygen particularly during aerobic metabolism. Superoxide is also employed by the mammalian immune system in response to infection. (776) The ubiquity of SODs across the tree of life illustrates their biological importance in an oxygenated world, and are particularly important as part of the host–pathogen interface. (777) Typically, SODs are classified in families according to their metal cofactor, and are grouped into three different structural classes: the copper and zinc or Cu-only SODs, the nickel SODs, and the manganese (Mn), iron (Fe) or cambialistic Fe/Mn SODs. (778) Each metal cofactor results in a metal active site with a distinct redox potential and architecture. SOD subfamilies also diverge in distribution, subcellular location, quaternary structure and specific mechanism. (778) Nonetheless, in general they achieve catalysis via a ping-pong mechanism, in which the metal center oscillates between oxidized and reduced states that differ by one electron. (779) As such, the general SOD mechanism consists of two half-reactions, where substrate oxidation produces O2 and a reduced metal site, and substrate reduction produces H2O2 and the oxidized metal site.
The differential expression of SODs with different metal requirements by the same organism can be considered a canonical metal sparing mechanism to ensure organism survival during metal nutrient limitation and the oxidative burst. (777) A highly conserved regulatory mechanism, originally described in E. coli, consists of a global regulator, the FeII sensor Fur, a small RNA, RyhB, and two genes encoding distinct Mn- and Fe-SODs, sodA and sodB, respectively. When Fe is replete, Fur represses transcription of sodA and ryhB, which lowers the stability of the sodB mRNA, thus leading to the production of Fe-SOD. Under conditions of Fe starvation, Fur dissociates from its operator, leading to transcriptional derepression of sodA and ryhB, allowing for high cellular levels of MnII–SOD. (780) This mode of regulation has also been described in Azotobacter vinelandii, (781) Bacillus subtilis (782) and Helicobacter pylori (783) with global regulators from the same regulator family and SODs with distinct metal preferences.
Another regulatory mechanism first described in Streptomyces coelicolor involves the inverse regulation of the expression of sodF (Fe-containing) and sodN (Ni-containing) by the nickel-specific repressor Nur, a Fur-family regulator. (784) In nickel-replete conditions, Nur represses sodF transcription, while inducing sodN expression indirectly. The sodF gene produces a stable small-sized RNA species (s-SodF) of ≈90 nucleotides that harbors an anti-sodN sequence complementary to the 5′ end of sodN mRNA through the ribosome binding site. Therefore, Nur activates sodN expression by inhibiting the synthesis of the sodF mRNA, from which inhibitory s-SodF RNA is generated forgoing the need for another small RNA to acts as a mediator. (784) SODs of different metal selectivity are thus differentially expressed according to cellular metal nutrient bioavailability.
In contrast to metal-specific SODs, there are other SODs that are genuinely cambialistic (camSOD) (785) or metal-interchangeable. Here, a SOD may be characterized by distinct metal binding affinities for Fe vs Mn yet possess identical specific activities when cofactored by either metal. (7) These SODs, denoted SodFMs, will be differentially metalated with Fe or Mn depending on which metal is present and bioavailable. Notably, S. aureus camSOD (Figure 32A) allows superoxide detoxification during host-imposed Mn starvation by calprotectin (CP), which functions in place of the Mn-dependent SodA (Figure 32B) and is presumably not effectively metalated under these conditions (Section 2.2.2). (181,786) This is an excellent illustration of MnII-sparing, also found for at least one glycolytic enzyme in the same organism under the conditions of CP-mediated nutrient limitation. (6)

Figure 32

Figure 32. Superoxide Dismutases. A) Ribbon representation of the dimeric cambialistic superoxide dismutase SodM (camSOD) from S. aureus (6EX4) (792) with the highlighted monomer shaded cyan. The active site (close-up) with identical metal binding residues shown as sticks and highlighted in yellow. Fe is a gray sphere. A cyan star highlights an identified residue responsible for determining metal preference. B) Ribbon representation of the dimeric MnII-dependent superoxide dismutase SodA (MnSOD) from S. aureus (5N56) (792) with the highlighted monomer labeled in yellow. The active site (close-up) shows the Mn binding residues in sticks and highlighted in yellow. Mn is a black sphere. A cyan star highlights an identified residue responsible for determining metal preference. C) Conserved features of SODs that encompass both metal selective and cambialistic enzymes. While each has a high degree of identity and highly conserved metal binding residues (highlighted in yellow), certain key residue positions have been identified as determinants in metal preference (highlighted with cyan stars). Their identity can modulate its metal preference from Mn- or Fe-selective to cambialistic.

High Resolution Image
Download MS PowerPoint Slide
The evolutionary diversity of extant SODs is thought to arise in part from a series of gene duplication events, later selected during the earth’s geological changes, e.g., the great oxygenation event, or as an adaptation of pathogens to host-imposed stresses, including nutritional immunity and ROS exposure. (181,787) Determining the metal preference of a metalloprotein in general remains a difficult task, and in the case of SODs, these preferences have been shown to depend not only on the first coordination shell of metal binding residues but is also impacted by the identity of specific second sphere residues (Figure 32C). (787) In addition, discoveries such as the Cu-only SODs from an opportunistic fungal pathogen that dispose of the requirement for Zn altogether within a canonical Cu/Zn-dependent SOD fold, and which are strongly induced under conditions of Fe-starvation, highlights a potential of still uncovered regulatory and structural diversity within the superoxide dismutases. (788−790) Recently, a comprehensive sequence and phylogenetic analysis has been shown to improve the ability to predict the preferred metal cofactor in a SOD, which enhances our understanding of the catalytic diversity present in this family. (680) This study proposes that redox tuning (791) allowed the SodFM subfamily to exquisitely fine-tune their metal preference along a continuum of Mn:Fe preferences as a result of very subtle structural changes in the active site. Indeed, cambialism may not simply be a feature that has evolved multiple independent times, but may also be under positive selection in nature to allow a bacterial pathogen to circumvent host-derived nutritional immunity during an infection. (181,680)

4.3.6. Ribonucleotide Reductases

The ribonucleotide reductases (RNRs) catalyze the 2′-dehydroxylation of a ribonucleotide substrate thus providing the only de novo route to 2′-deoxyribonucleotide building blocks for DNA synthesis. (8,793) RNRs are typically composed of two types of subunits: a catalytic R1 subunit and a small radical-generating subunit R2. R2 comprises a core four-helix bundle that harbors the metal cofactor, and four additional auxiliary helices that participate in the interaction with the other subunits (Figure 33A). Ribonucleotide reduction typically requires a metal cofactor and an oxidant in the R2 subunit (encoded by nrdF) that ultimately oxidizes an active site cysteine side chain in the R1 subunit to create a thiyl radical; this allows for radical activation of an otherwise inert C–H bond in the ribose ring of the substrate bound to R1. (8) The mechanism by which the metal cofactor generates the radical to initiate this chemically challenging reduction reaction varies substantially among the three classes of RNRs. While class II requires an adenosylcobalamin cofactor and class III, a [4Fe-4S]-activating enzyme, all class I enzymes harbor a bimetallic metal site defined in the core four-helix bundle that also encodes for an aromatic residue which participates in electron transfer (Y105 in E. coli NrdF) (Figure 33). (8) A metal-free class Ie enzyme constitutes a possible exception within class 1 which lacks the metal-ligating glutamates (highlighted in blue in Figure 33B) and instead requires a specialized NrdI activating enzyme. (8,794) The metal requirement and metal specificities of the R2 subunits have been shown to differ among the class I subclasses, with class Ib of particular note here, since class Ib R2 enzyme is truly cambialistic, forming active enzymes with both dinuclear Fe/Fe and Mn/Mn cofactors (Figure 33). (795)

Figure 33

Figure 33. Ribonucleotide reductases. A) Ribbon representation of the R2 subunit of the ribonucleotide reductase NrdF from E. coli in the Mn2-form in cyan (3N37) and Fe2-form in yellow (3N38). (795) A close-up view of the active site with metal binding residues shown as sticks. Fe is a black sphere, Mn is a gray sphere, and the red spheres represent water molecules. An orange star highlights the tyrosine residue that will become a tyrosyl radical upon metal oxidation. The interface with the catalytic R1 subunit is indicated with a red dashed line. C) Conserved features of class Ib R2 subunit that encompass metal binding residues highlighted with circles, while the tyrosyl radical is highlighted with a star. Residues conserved among all class I enzymes are highlighted in yellow while the residues absent in the metal free class Ie activase are highlighted in cyan. The residue that has been shown to impact metal preference in B. anthracis R2 is indicated with a black arrow.

High Resolution Image
Download MS PowerPoint Slide
The physiological conditions that drive iron or manganese loading in class Ib enzymes remain long-standing questions in the field. (683,796) One possibility is that metalation is primarily determined by the specific characteristics of the growth environment where the bioavailability of FeII, which can be significantly modulated by collateral oxidative stress, vs MnII may well dictate metal occupancy of a single expressed class Ib enzyme, thus enhancing bacterial survival. (7) It has been proposed that use of MnII by bacterial subclass Ib, Ic and Id RNRs, which contrasts with the exclusive use of iron by the eukaryotic Ia enzymes, may help to protect certain pathogens against iron deprivation imposed by their hosts. In E. coli, for example, the class Ib NrdF protein is not expressed under conditions of metal-replete aerobic growth but is induced under iron limitation and oxidative stress. In contrast to the class Ia diferric tyrosyl radical (FeIII2-Y•) cofactor, which can self-assemble and mature from FeII2-NrdF and molecular O2, generation of the MnIII2-Y• cofactor in the class Ib enzyme requires the reduced form of a flavoprotein, NrdIhq, and O2 to create superoxide anion for active site assembly and oxidative maturation. (795) Indeed, structures of the NrdI-NrdF complexes have been known for a number of years now. (795) This may not be generally true for all class Ib enzymes, since in Bacillus cereus, for example, the NrdF–NrdI complex is not strongly affected by the nature of the bound metal in the NrdF subunit. (797) Furthermore, Bacillus anthracis class Ib R2 subunit intrinsically selects manganese over iron, under both aerobic and anoxic conditions, when the metal-free protein is subjected to incubation with equimolar MnII and FeII in a crystallographic soaking experiment. (798) These findings are at face value in opposition to expectations from the Irving-Williams series (Figure 8). Interestingly, introduction of a substitution of a second coordination sphere residue (highlighted with an arrow in Figure 33B) in B. anthracis NrdF only subtly changes the site 1 metal coordination geometry and yet impacts spontaneous metalation of the protein to form a predominant heterodinuclear Mn/Fe cofactor under aerobic conditions and a Mn/Mn metal center under anoxic conditions. (798) There is clearly much more to learn about the general principles of metal selectivity in cells, particularly in class 1b RNR activating enzymes.

5. Metal Trafficking to Specific Clients and Compartments

Click to copy section linkSection link copied!
A second major regulatory outcome of metalloregulation of gene expression by metal sensor proteins is the expression of proteins that traffic or allocate transition metals to a specific recipient protein or cellular compartment (Figure 1). Seminal work on copper chaperones found in all domains of life provides the framework for much of this discussion. (799,800) These early studies showed that a metallochaperone forms a ligand-exchange labile coordination complex with a specific metal and exploits the specificity of protein–protein interactions to transfer the metal to an acceptor or client protein (Figure 34, top). Here, we discuss more recently published work on a distinct class of soluble, NTP-hydrolysis powered metallochaperones involved in metalloenzyme maturation. (688,691,801) We also discuss recent structural and biological work on metal transporters, some from the P-type ATPase superfamily, (802,803) as historical clients of the copper chaperones, (804,805) that are paired with an accessory protein(s) that may have metallochaperone-like features, or possess intrinsic metallochaperone activity itself. (792) Although chemical and structural mechanisms that underscore the processes we describe here are, in some cases, not yet known, these biological studies establish a foundation for future studies.

Figure 34

Figure 34. Conceptual framework for thermodynamically favorable metal transfer from a donor protein to an acceptor client protein. Top, metallochaperone-client protein interaction lowers the kinetic barrier to metal transfer via intermolecular metal–ligand exchange. Bottom, metallochaperone-client protein interaction reduces the thermodynamic stability of the metallochaperone-metal complex, enabling transfer by increasing the favorability of the transfer reaction.

High Resolution Image
Download MS PowerPoint Slide
One prevailing model of metallochaperone function holds that the association between a metalated donor and an apoprotein acceptor occurs such that the donor metal complex is in close physical proximity to the empty client metal binding site, allowing client metal ligands to invade the first coordination shell of the donor metal complex, and initiate a ligand exchange process. (804,806,807) This framework exploits the fact that late 3d-block divalent transition metal complexes formed with biological ligands, even complexes of considerable thermodynamic stability and metal association constants of 1010 to ≥1020 M–1, are kinetically exchange-labile, (808) particularly so if a higher coordination number complex can be transiently formed during the metal transfer process (Figure 34, top). (809,810) Alternatively, protein–protein docking disrupts or alters the first coordination shell of the donor, independent of intermolecular metal ligand exchange, lowering the metal affinity, thereby enhancing the kinetics and thermodynamics of metal transfer (Figure 34, bottom).
The rates and mechanism by which metals are exchanged between donor and acceptor remains understudied, but must be dictated in part by their position along a continuum of divalent metal stability constants with simple organic ligands, known as the Irving-Williams series (Figure 7). (157) We posit that for high affinity metals, notably CuII (and CuI), ZnII and NiII, affinities are so tight so as to render metal dissociation rates (off-rates), even if increased by the presence of metal-buffering small molecules, far too slow to be relevant during intermolecular metal transfer (Figure 34, top). A metal–ligand exchange process effectively lowers the kinetic barrier to the metal movement as long as the thermodynamic gradient from donor to acceptor is favorable. (805) This arrangement also effectively prevents deleterious “off-pathway” reactions by highly competitive metals (Figure 7) where inappropriate metal dissociation might drive noncognate mis-metalation of the proteome (Section 2.2.3), particularly problematic for metalloenzymes that require weaker binding metals for activity. (168,811) In contrast, metal dissociation rates into solvent for those weaker binding metals, including MnII and FeII on the left side of this series (Figure 7), can be sufficiently rapid to be captured by both metal-buffering small molecules and ultimately the recipient acceptor protein, particularly if an open or solvent-accessible coordination site is available on the metal bound to the donor protein.

5.1. Soluble Nucleotide-Dependent Metallochaperones

5.1.1. Overview

Enzymes associated with the superfamily of P-loop G3E GTPases and other closely related NTPases are the principal nucleotide-dependent metallochaperones (NMC) in all kingdoms of life (Figure 35A). (688,691,801) These GTPases (G-proteins) use the energy of GTP hydrolysis to facilitate metal transfer or another step in the metalloenzyme maturation cycle, shown specifically for a metal transfer reaction (Figure 35B). (56,812−815) G-proteins are characterized by several short, conserved G-loops that are responsible for GTP binding and hydrolysis and are denoted G1-G5, from N- to C-terminus (Figure 36A,B). (816) Loop G1 encompasses the canonical P-loop or Walker A motif and recognizes the α- and β-phosphates. G2 (switch 1/switch I) and G3 (Walker B or switch 2/switch II) stabilize the β- and γ-phosphates on the bound nucleotide and donate ligands to the MgII coordinated by these phosphates (Figure 36A). The MgII activates a water molecule for GTP hydrolysis. The G3E family gets its name from a Glu (E) in the G3 motif, ExxG instead of the canonical DxxG. (817) The glutamate directly participates in coordinating the MgII ion, whereas the more typical aspartate makes a water-mediated contact to the metal ion. (817) The G4 (often NKxD, where N, Asn, K, Lys; D, Asp) and G5 loops influence nucleotide specificity and release, respectively, via interactions with the nucleobase; G5, in particular, exhibits higher sequence variability than the other G-loops. (801)

Figure 35

Figure 35. Nucleotide-dependent metallochaperones. A) Known members of the small P-loop G3E superfamily of NTPases that are known or thought to be involved in metallocofactor maturation. SIMIBI, signal recognition particle, MinD, and BioD; COG0523, cluster of orthologous groups 0523. The Co MeaB square is shaded differently to emphasize that MeaB functions as an accessory protein for ATR (Section 5.1.3). B) Nucleotide-dependent metallochaperone model, which proposes that GDP hydrolysis and dissociation of Pi are just upstream or coincident with metal transfer to the client.

High Resolution Image
Download MS PowerPoint Slide

Figure 36

Figure 36. Conceptual and structural comparisons between representative SIMIBI class and COG0523 subfamily metallochaperones. A) Cartoon representation of SIMIBI G-proteins. The five G-loops are shaded as in panel B: G1, red; G2, orange; G3, yellow; G4, green; G5, blue. The GTP is purple with the three 5′ phosphate groups shaded red. The metal to be transferred is shaded green. B) Ribbon representation of Helicobacter pylori HypB (4LPS), with the five G-loops shaded as in panel A. C) Cartoon representation of a canonical COG0523 protein, colored as in panel A, with ZnII shaded black. CxCC (C, Cys; x, any amino acid) is the high affinity metal binding site of COG0523 enzymes. (820) D) AlphaFold2 model of Acinetobacter baumannii ZigA from Model Archive showing GTP and ZnII positioned as in the crystal structure of Methanococcus janaschii HypB (PDB 2HF8) (panel B). (820) Conserved motifs are indicated on the structure, shaded red, orange, yellow, green, and blue for G1-G5 respectively. MgII ion, gray sphere; ZnII ion black sphere; GTP, purple stick.

High Resolution Image
Download MS PowerPoint Slide
The G2 and G3 loops are conformationally plastic and function as molecular switches that drive a change in structure and/or dynamics within and likely beyond the G-domain upon GTP hydrolysis. The conformation of the G2 loop in particular defines the active ON (GTP-bound) and inactive or relaxed OFF (GDP-bound) states; (818,819) conformational switching by one or more G-domain “switches” upon GTP hydrolysis is likely the key structural trigger that orchestrates NMC function in a metallochaperone cycle (Figure 35B).

5.1.2. SIMIBI Proteins Involved in NiII-Enzyme Biogenesis

UreG and HypB are examples of extensively characterized G3E GTPases that utilize their G-domains to catalyze NiII insertion into urease and [NiFe]-hydrogenase, respectively (Figure 37). (819,821,822) This metalloenzyme specificity is not absolute, since UreG can substitute for HypB in the maturation of [NiFe]-hydrogenase, with the accessory protein UreE a key player that links the two metalloenzyme maturation pathways in some bacterial cells (Figure 37). (823,824) Both UreG and HypB are members of the SIMIBI (signal recognition particle, MinD, and BioD) class of ATPases and GTPases (825) that are regulated by reversible dimerization stabilized by NTP binding that is linked to formation of a metal coordination site that is destined to be transferred to a client enzyme (Figure 38A,B). (819,826) In the GTPases, the GTP is sandwiched into two symmetry-related binding sites at the protomer interface of the dimer in the ON or GTP-bound state. In the case of UreG, this creates a subunit-bridging NiII site that is coordinated by an invariant Cys-Pro-His sequence from opposite protomers and positioned between the G2 and G3 loops in the primary structure, to form an approximately square planar Cys2-His2 coordination complex (Figure 38A), (821) mutagenesis of which abolishes urease maturation. (827) ZnII can also bind to this site, but the binding is weaker. (162) These general features also characterize HypB, except that the subunit-bridging coordination chemistry is distinct, involving two invariant Cys, one of which is part of an Cys-His motif positioned between the G2 and G3 loops, (691) and another embedded in the G3/switch II motif itself (Figure 38B). (819,824,828) This suggests a conformational “gating” mechanism in HypB where GTP hydrolysis signals to the NiII binding site through G3, thus initiating metal transfer to HypA, which itself later inserts NiII into [NiFe]-hydrogenase (Figure 37, top). (829) There is evidence to suggest that the His in the Cys-His motif in HypB (Figure 38B) toggles on and off the NiII as a function of nucleotide status as an important aspect of the gating process, but more needs to be done here. (828)

Figure 37

Figure 37. Schematic illustration of the maturation of bacterial [NiFe]-hydrogenase and urease. UreE (labeled E; shaded pink) is thought to bridge the two metalation pathways, capable of accepting NiII from HypA and UreG as indicated. (824)

High Resolution Image
Download MS PowerPoint Slide

Figure 38

Figure 38. Structural features of the UreG and HypB homodimers. A) Ribbon structure of the KpUreG dimer (gray, black subunits) with G-loops shaded as in Figure 36 and the intersubunit NiII coordination sphere shown in magenta stick. NiII, cyan. GTP, purple stick. Inset, Close-up of the NiII coordination sphere with an abbreviated primary structure representation with NiII ligands (cyan circles) and the E residue as the namesake of the G3E superfamily indicated. B) Ribbon diagram of the HpHypB dimer, shaded as in panel A. Inset, close-up of the NiII coordination site, with the Cys residue drawn from the G3 loop (yellow) and CH sequence (magenta) shown.

High Resolution Image
Download MS PowerPoint Slide
5.1.2.1. Urease Maturation
There has been considerable progress in our mechanistic understanding of UreG-dependent urease maturation in both Klebsiella pneumoniae and in the gut pathogen Helicobacter pylori (Figure 37, bottom). (824,828,830) Urease hydrolyzes urea into carbon dioxide and ammonia (831) and allows H. pylori to thrive in the acidic environment of the stomach, where it portends an increased risk of peptic ulcers and gastric cancer. (832) Urease adopts a C3-symmetric trimer of trimers (αβγ) architecture (Figure 39A), harboring a buried binuclear Ni2 active site (Figure 39B) featuring a bridging carbamylated Lys (Figure 39C), with all metal ligands derived from the α-subunit. Urease maturation involves the participation of four accessory proteins that form a urease activation complex that ultimately moves NiII from UreG to UreB (H. pylori designation) via the obligatory intermediacy of UreF and UreD (UreH in other organisms) (Figure 37, bottom). Early work established that UreG, UreF and UreD form a 2-fold symmetric UreG2F2D2 complex in which the UreG dimer fits snugly between an approximately linear UreD-UreF–UreF’-UreD’ complex with UreG2 docked opposite the UreF2 core, positioning the subunit-bridging NiII ion at the UreG2F2 interface (Figure 40A). (830)

Figure 39

Figure 39. Structure of K. aerogenes urease (1FWJ). (824) A) Ribbon representation of the trimer or trimers architecture, with one αβγ heterotrimer shaded blue, red and yellow, respectively. B) Metal coordination structure of the binuclear Ni2 active site. C) chemical representation of the coordination structure, featuring a bridging carbamylated Lys.

High Resolution Image
Download MS PowerPoint Slide

Figure 40

Figure 40. Urease maturation intermediates. A) The 2-fold symmetric H. pylori UreG2F2D2 complex (4HI0), with the UreG G-loops of the homodimeric UreG shaded as in Figure 38A. G1-G5 loops are colored red, orange, yellow, green, and blue, respectively; GDP, colored purple, and the CPH Ni-binding motif colored magenta. UreF subunits are colored light and dark blue, and UreD subunits are colored light and dark purple. B) An Alphafold2 model of the UreG2•UreE2 complex (ModelArchive ma-41ctj), which functions upstream of the complex in panel A.

High Resolution Image
Download MS PowerPoint Slide
Functionally upstream of UreG is UreE, and it is the UreG2-GTP form that productively engages NiII-bound UreE2 and accepts the metal, leading to dissociation of UreE following NiII transfer (Figure 40B). The GTP-bound form UreG2 is then incorporated into the UreG2F2D2 complex (Figure 40A). Assembly of this metal transfer intermediate alters the structure of the terminal UreD subunit in such a way that this allows UreG2F2D2 complex to engage a single protomer of the apo-urease trimer, with a single UreD subunit of UreDF heterodimer necessary and sufficient to bind a urease trimer, primarily via interactions with UreB (Figure 41). (833) Remarkably, formation this UreFD-urease complex results in the formation of an extended ≈100 Å tunnel that links the UreG-NiII binding domain with the empty urease active site, thus suggesting a plausible pathway of NiII movement from UreG to urease (Figure 41). Critically, mutations that disrupt the UreG2F2D2 complex or the UreFD interface abolish urease activation in cells and In vitro, respectively. (830,833) In addition, mutations that compromise the integrity of the tunnel negatively impact urease maturation In vitro. (833)

Figure 41

Figure 41. Model of a single UreG2F2D2 complex (colored as in Figure 40A) bound to one of the three urease heterotrimers (green), based on overlaying the cryo-EM structure of H. pylori urease in complex with UreFD (8HC1) (833) and the crystal structure of H. pylori UreG2F2D2 complex (4HI0). This results in formation of an extended tunnel (dark gray), as identified by the Pymol plugin Caver 3.0.3, (834) that connects the UreG active site (magenta) with the urease active site. (833) The schematic model from Figure 37 is shown in the lower left.

High Resolution Image
Download MS PowerPoint Slide
This remarkable tunnel serves to enhance the exquisite metal specificity of urease maturation that must be dictated by the metal specificity of the UreE2G2 complex itself, while also chaperoning this highly competitive NiII (Figure 7) into the active site, so as to avoid off-pathway mis-metalation by this toxic metal ion. Mechanistically, GTP hydrolysis is known to induce a conformational switch in the G2 loop and Cys-Pro-His NiII-binding motif in UreG (Figure 38A), (821) and perhaps only as part of the activation complex, may weaken the NiII affinity and allow for dissociation of NiII into the vestibule (Figure 34, bottom). (833) How the integrity of metalation of client proteins dependent on other G-protein metallochaperones (Section 5.1.4) is not yet known, and this is an area worthy of significant future study.

5.1.3. SIMIBI Proteins: MeaB and Maturation of B12-Cofactored Enzymes

Although the work described above for HypB and UreG is consistent with the metallochaperone model (Figure 35B), both function upstream of the actual apoenzyme metalation event and neither directly interacts with the client enzyme itself. In the maturation of the adenosylcobalamin (AdoCbl; B12)-dependent enzyme, methylmalonyl-CoA mutase (MCM), the SIMIBI family G-protein MeaB (MMAA in humans) also requires the energy of GTP hydrolysis to drive metallocofactor insertion into MCM. However, MeaB (MMAA) functions not to deliver metallocofactor cargo (AdoCbl) directly, but instead uses the energy of GTP hydrolysis to close a “wedge” between the two subdomains of MCM, after another enzyme, adenosyltransferase (ATR; MMAB in humans (835,836)) inserts the cofactor into an “open” form stabilized by the MeaB-GTP dimer (Figure 42A). (816,837) ATR has two roles, one of which is to adenylate cob(II)alamin using ATP to make AdoCbl, followed by transfer of AdoCbl to the cobalamin (Cbl)-binding domain of MCM. Thus, MeaB (MMAA) leverages the energy of GTP hydrolysis to gate AdoCbl insertion into the active site of MCM, while also stimulating removal of damaged cofactor from inactive MCM (Figure 42A). (816,838)

Figure 42

Figure 42. MeaB and adenosyltransferase (ATR)-dependent maturation of MCM. MeaB is shaded gray (GDP-bound OFF state) or bronze (GTP-bound ON state), ATR, blue and MCM, green, with the Cbl-binding domain of MCM shaded dark green. (841) A) Model for a typical system in which MCM and MeaB are different polypeptide chains. GTP binding activates MeaB to recruit cofactor-loaded ATR to an “open” conformation of the mutase; GTP hydrolysis inactivates MeaB (OFF), leading to a “closed” conformation. Cofactor damage leads to the recruitment of GTP-bound MeaB and ATR, leading to release of the damaged cofactor. B) Model for maturation of a fused (IcmF) system, which requires GTP dependent dimerization of two peripheral MeaB protomers within a filamentous oligomer to create a conformation that is active for maturation. (841)

High Resolution Image
Download MS PowerPoint Slide
5.1.3.1. Standalone MeaB
Initial structures of dimeric MeaB in the apo-, product GDP- substrate GMPPNP-bound states were rather similar (Figure 43A), with the nucleotides of each protomer far apart, thus inconsistent with the dimer structures of HypB and UreG (Figure 36A,B). Further, a functionally critical switch III loop, a mobile loop that is very close to G3 and unique to MeaB, was exposed to solvent. MMAA mutant alleles in this region are associated with human disease-causing mutations in methylmalonic aciduria, an inborn error of metabolism (99) making these findings puzzling. A more recent structure of a high affinity (Kd ≤ 100 nM) 1:1 complex between MeaB and the Cbl-binding domain of an MCM, solved in the presence of GMPPCP, reveals a completely different dimer conformation, where the two nucleotides are now quite close to one another, sandwiched at the dimer interface, with the switch III loop now engaged in direct interactions with the nucleotide of the opposite protomer of MeaB, effectively burying the bound GTP (Figure 43B). The G2-loop (switch I), disordered in the “inactive” conformation, (816) becomes ordered in this complex, forming a β-strand that hydrogen-bonds with a β-strand in MCM, establishing much of the intermolecular interface. This conformation of MeaB is thought to be representative of the “active” state. (837)

Figure 43

Figure 43. Conformational change in MeaB in moving from the “inactive” GDP-bound conformation to the “active” GMPPCP-bound state enabled by the stable interacion with the Cbl-binding domain of MCM. A) GDP-bound conformation (2QM7). (842) B) The active conformation compared to a cartoon model of a dimer interface that must be formed in the filament form of IcmF. (841) This structure is that of the soluble MCMCbl-MeaB 2:2 complex (8DPB). (837) The two MeaB subunits on the left are oriented in exactly the same way in panels A and B, allowing visualization of the large movement of the MeaB’ subunit (right) in the “active” relative to “inactive” states. The G2 switch I (orange) and the switch III (magenta) regions are highlighted in each structure, with the G-loops shaded as above. The bound nucleotide is in stick (purple).

High Resolution Image
Download MS PowerPoint Slide
The binding of MCM stimulates the GTPase activity of MeaB dramatically, and thus the client functions as a GTPase activating protein (GAP), a finding nicely explained by the structure of the complex. (839) Consistent with this, switch III mutants effectively uncouple GTP hydrolysis from cofactor transfer from ATR, in that significant cofactor transfer occurs in the presence of a nonhydrolyzable GTP analog, GMPPNP; in addition, these mutants suppress the GAP function of MCM, giving rise to elevated MCM-independent GTPase activity. As expected, switch III mutants are impaired for the ability to offload damaged cofactor, (816,837) notably cobalamin(II), which results from the loss of the 5′-deoxyadenosyl moiety. (816)
5.1.3.2. Fused MeaB
Studies of another bacterial B12 enzyme maturation system provides new insights into how GTP hydrolysis gates AdoCbl transfer to the mutase. IcmF from Cupriavidas metallodurans is an AdoCbl-dependent isobutyryl-CoA mutase fused (IcmF), in which the MeaB domain exists as an independent domain of the target enzyme, which itself harbors two domains, an N-terminal TIM barrel-shaped substrate binding domain and a C-terminal Rossman fold AdoCbl-binding domain (Figure 42B). (840) The AdoCbl cofactor is situated at the interface between these two subdomains. IcmF is a homodimer, with the substrate-binding domain forming most of the dimer interface, placing the two G-domains on the periphery of the dimer (Figure 42B). (840) A key aspect of these structures was that the apo- and holo (cofactor-bound)-IcmF dimers were found to adopt “open” and “closed” states, which are representative of a structure ready to accept cofactor from ATR and the catalytically competent and cofactored state, respectively, a finding that parallels subsequent work in human MMUT-MMAA complex. (840)
These IcmF structures proved puzzling, however, in that it seemed impossible to form an obligatory G-protein dimer like that described above (Figure 43B) (816,837) furthermore, the switch III region was again fully solvent-exposed. (837) This conundrum of the “inactive” dimeric IcmF was ultimately solved when it was shown that in the presence of GMPPCP, IcmF oligomerizes into filamentous chains, thus creating the “active” MeaB-like dimer between IcmF dimers within the filament (Figure 43B). (841) This form is capable of recruiting ATR, and once one of the two immediately adjacent ICM active sites is metalated, GTP hydrolysis closes the “gate”, and metalated and catalytically active IcmF dissociates (Figure 42B). (840) Both the identity of the G-nucleotide and the chemical integrity of AdoCbl cofactor impact the oligomeric state equilibrium, albeit to distinct degrees, consistent with their functional importance. (841) The “open” state of IcmF is also readily visible in these structures and distinct from that of the “closed” state. (841)

5.1.4. COG0523 Proteins

Detailed studies like those described above of SIMIBI GTPases provide significant context for our understanding of other poorly understood NMCs. Cluster of orthologous groups (COG) 0523 proteins are a distinct subfamily of the G3E P-loop NTPases that are associated with the biosynthesis of vitamin B12, the maturation of the Fe-dependent nitrile hydratase and a large diversity of other proteins that are often, but not always, regulated by Zur and thus become cell-abundant under conditions extreme zinc starvation (Figure 35). (146,688,843−845) COG0523 enzymes are conserved in all kingdoms of life, from bacteria to plants to higher vertebrates, and many are connected to the cellular zinc limitation response. (56,691,814,846) Unlike SIMBI family proteins, COG0523 proteins may function as monomers and harbor an N-terminal G-domain of ≈240 residues and a C-terminal candidate protein–protein interaction domain connected to the G-domain by a variable length linker (Figure 36C,D). The only known COG0523 crystallographic structures, of E. coli YjiA show a C-terminal domain consisting of an antiparallel β-sheet flanked on one side by a pair of α-helices, in a β1-α1-β5-β2-β3-β4-α2 topology (Figure 44A). (818,847) Alphafold2-based modeling suggests that many families of COG0523 proteins follow the same pattern (Figure 44B), while others have additional elements of secondary structure inserted between the β4 and β5 strands (Figure 44C), and some have an additional β-strand N-terminal to β1 (Figure 44D). These features in the CTD may impact target protein specificity but this is not yet known. The extent to which the G-domain and the C-terminal domain stably interact with one another is also unknown and may well change as a function of G-nucleotide status. Although some COG0523 proteins have been reported to weakly self-associate, (125,848) the functional importance of this is not known nor are there general trends in oligomerization state that are dependent on G-nucleotide status across the COG0523 subfamily; in any case, the C-terminal domain of EcYjiA docks against the G-domain in a way that would seem to inhibit dimerization in a UreG or HypB-like fashion, at least in the conformational state. (120)

Figure 44

Figure 44. Secondary structural schematics (left) of the tertiary structure (A) or AlphaFold2 model (B-D) (right) of the C-terminal domains (CTDs) of selected COG0523 family members. A) E. coli YjiA (PDB 4IXN), (818) B) human ZNG1A (Q9BRT8) (56) C) A. baumannii ZigA (ModelArchive ma-zam5h), (820) and D) E. coli YeiR (P33030). (848) Left, α-helices, blue cylinders; β-strands, gray arrows, with the residue numbers from each primary structure that define the 2° structure boundaries indicated. NTD, N-terminal G-domain upstream. Right, the ribbon is colored from blue to red, starting with the α0 helix (front), then the most N-terminal β-strand, β1 through to the C-terminal α2 helix (back, opposite the β1 and β5 strands).

High Resolution Image
Download MS PowerPoint Slide
5.1.4.1. Coordination Chemistry
The key distinguishing feature of the G-domains of COG0523 proteins relative to SIMIBI proteins is the presence of an N(G/S)CϕCC (N, Asn; G/S, Gly/Ser; C, Cys; ϕ, hydrophobic) motif on a loop that is positioned between the G2 and G3 loop motifs that has long been thought to serve as a high affinity (KZn ≈ 1010–1011 M–1) coordination site for cognate metal (Figure 36). (688) A superposition of one subunit of the MgII•GTP-bound HypB and an Alphafold2 model of A. baumannii ZigA (Zur-induced GTPase A) G-domain reveals that the NiII transfer site in HypB is nearly precisely coincident with the CICC motif in this Zur-regulated COG0523 enzyme (Figure 45). (820) Recent X-ray absorption spectroscopy (XAS) studies of A. bauammnii ZigA reveal that all three Cys residues contribute to a S3(O/N) ZnII coordination complex in the G-nucleotide-free form, with the identity of the fourth ligand as yet unknown. (820) The structure of this chelate remains S3(N/O) in the GTPγS-bound ON and GDP-bound states OFF states, although these spectra will be somewhat insensitive to a change in the nature of the lighter donor atom in this complex. (820) This chelate structure is in contrast to an early crystallographic study of ZnII-soaked EcYjiA, which suggested a ZnII binding site composed of the second Cys in the CϕCC motif and two acidic residues derived from the G2 loop region, not conserved in AbZigA. (691) EcYjiA, of unknown biological function, and AbZigA, which clearly plays an important role in the adaptive response to transition metal starvation (125,738) are only distantly related to one another. (691) The GTPase activity both AbZigA, the same enzyme from Staphylococcus aureus (SaZigA) and another E. coli COG0523 enzyme of unknown function, EcYeiR, are both modestly (up to ≈5-fold) stimulated by ZnII binding to a high affinity site; in both ZigAs, this stimulation is ablated in a triple serine substitution mutant, CICC to SISS, which selectively eliminates high affinity ZnII coordination to S3(N/O) site. (848,849) Although both EcYeiR and EcYjiA bind ZnII with equilibrium affinity comparable to that ZigA, the ZnII affinity in neither case is appreciably increased by GTP binding, in contrast to AbZigA (see below). (146,820) Most COG0523 enzymes connected in some way to zinc homeostasis harbor one or more secondary ZnII sites of lower affinity whose functions are unknown. (56,848,849)

Figure 45

Figure 45. Superposition of one subunit of the MjHypB in the GTP-bound ON state (2HF8, cyan) (820) and the G-domain of the AlphaFold2 model of A. baumannii ZigA (ModelArchive ma-zam5h, light gray). The G-loops are colored as above, with G1-G5 colored red, orange, yellow, green, and blue, respectively; GMPNP colored purple, CxCC motif colored magenta, and the CH NiII-binding motif colored pink.

High Resolution Image
Download MS PowerPoint Slide
5.1.4.2. Structure and Dynamics of G-Nucleotide Binding
A near total lack of high-resolution structure and dynamics insights for COG0523 enzymes and how these molecular switches change from the “spring-loaded” (850) GTP-bound ON state to the “relaxed” GDP-bound OFF state, continues to severely limit our understanding of these enzymes. Insights from a family of bacterial GTPases that regulate ribosome biogenesis under conditions of nutrient limitation (850) reveal that in the MgII•GMPPCP (818)-bound ON state, the G2 helix and the G3 loop are fully folded to engage the terminal γ-phosphate; upon nucleotide hydrolysis, the G2 helix is destabilized and becomes highly dynamic (Figure 46). (850) Recent studies of AbZigA employed hydrogen–deuterium exchange mass spectrometry (HDX-MS) to investigate the ligand-free, ZnII bound, MgII•GTPγS–ZnII substrate and MgII•GDP-ZnII product states (Figure 47). (820) These and other findings reveal that ZnII coordination by the CxCC motif slightly destabilizes the G-domain, while subsequent addition of either substrate analog GTPγS and product GDP induce significant stabilization of a metastable or fluxional G-domain (Figure 47); analogous findings may well characterize the closely related FeII-dependent nitrile hydratase activator protein. (843) Quantitative metal binding studies reveal that both MgII•GDP and MgII•GTPγS enhance the ZnII binding affinity, with the effect far greater for MgII•GTPγS, by ≈10-fold; (820) these findings of significant positive linkage between substrate MgII•GTPγS and ZnII binding, also observed by CoII-bound CobW. (851) A significantly reduced ZnII affinity of the MgII•GDP-bound OFF state may well provide a (shallow) thermodynamic driving force for metalation of a client protein (Figure 48A), exactly in line with thermodynamic gradients previously observed in sequential CuI trafficking reactions, (805) and analogous to protein–protein interaction affinity gradients recently observed in [4Fe-4S] cluster enzyme maturation. (852) The HDX-MS reveal unfolding of the G2 loop specifically upon GTP hydrolysis, which impacts the stability of selected other regions of AbZigA, extending into the CTD (Figure 48B). (820) This long-range communication of nucleotide status in the G-domain to other regions of the protein is canonical long-range allostery, and a hallmark of G-proteins in general.

Figure 46

Figure 46. A depiction of a structural changes in the small GTPase involved in ribosome biogenesis, Staphylococcus aureus RsgA. A) GDP-bound OFF state. The G-loops are labeled and shaded as above; note that the G2 loop is disordered in this structure and not observed in the crystal structure (pdb 6ZHM). (850) The structure of the alarmone (ppGpp)-bound state (pdb 6ZHL) is similar, with a disordered G2 loop. (850) The immediately adjacent oligonucleotide-oligosaccharide (OB)-fold (OB) and ZF (zinc finger) domains are shown in dark gray. B) GMPPNP-bound ON state of Escherichia coli RsgA bound to the 30S ribosomal subunit (pdb 7NAR). (853)

High Resolution Image
Download MS PowerPoint Slide

Figure 47

Figure 47. Ribbon representation (Alphafold2 model) of the results of probing various allosteric states of the bacterial ZnII metallochaperone, A. baumannii ZigA, by hydrogen–deuterium exchange-mass spectrometry (HDX-MS) relative to a common apo (ligand-free) reference state. (820) ON-state, GTP bound; OFF-state, GDP bound. The ZnII-coordinating Cys Sγ atoms are.indicated (yellow spheres). The ribbon in each case is colored according to% change in deuterium uptake (see scale, right). See text for details.

High Resolution Image
Download MS PowerPoint Slide

Figure 48

Figure 48. Thermodynamic model for favorable metal transfer that is linked to GTP hydrolysis by a nucleotide-dependent metallochaperone. (820,851) Here, the ON- or substrate GTP-bound state of the metallochaperone has a higher affinity for the metal to be transferred relative to the OFF or product GDP-bound state, as experimentally verified for CobW and A. baumannii ZigA. If the recipient client protein has an affinity between these two extremes, the metal will move thermodynamically “downhill” to the client (green arrow), with concomitant dissociation of the chaperone-client protein complex. (B) Depiction of the structural and/or dynamic changes originating with the G2/switch 1 region, as experimentally determined for A. baumannii ZigA by HDX-MS, which is destabilized in the OFF state. These changes in structure or dynamics propagate to the other regions of the G-domain and the C-terminal domain (pink ribbon). The approximate positions of the G1, G2, G2 and G4 loops are indicated by the red, orange, yellow and blue Cα spheres, respectively. The Sγ atoms of the high affinity ZnII site shown as dark yellow spheres. Scale bar:% change in deuterium uptake in the GDP•ZnII state relative to the GTPγS•ZnII state, colored onto the ribbon AlphaFold2 model of ZigA.

High Resolution Image
Download MS PowerPoint Slide
5.1.4.3. Client Proteins and Mechanistic Insights
Unlike the SIMIBI GTPases, there are few insights into metalation clients of bacterial COG0523 proteins, with the exceptions of authentic CobW (146) and the Fe-dependent nitrile hydratase activator protein Nha3. (843) The acquisition of CoII by the corrin component of vitamin B12 occurs in cells via one of two distinct pathways, designated as the early or late CoII insertion pathways. In the late insertion pathway, CoII is accessed by the COG0523-family metallochaperone CobW and incorporated into hydrogenobyrinic acid a,c-diamide (HBAD), as part of a HBAD-CobNST chelatase complex, to form CoII–HBAD, in what is also known as the metallochaperone-dependent pathway. (146,851) This reaction is calculated to be unfavorable if the CoII is derived from bioavailable CoII pool based on an idealized metal pool in Salmonella (Figure 8A) but is calculated to become thermodynamically favorable only when CoII is bound to the product MgII•GDP-CobW complex (Figure 48A). (851) There is as yet no published biochemical evidence that CobW activates CobNST directly, and the involvement of intermediary protein cannot yet be ruled out.
Limited biochemical data on the Fe-dependent α2β2 heterotetrameric nitrile hydratase maturation factor Nha3 support a model in which the folded α-subunit forms a complex with Nha3; FeII binding to Nha3 followed by GTP binding then stimulates GTP hydrolysis and FeII transfer to the α-subunit. (843,844) Following transfer, FeII-α is proposed to engage the β subunit and via subunit swapping with apo-GDP Nha3, which then drives formation of the mature α2β2 heterotetramer (Figure 49A). (844) Maturation of the CoII-dependent nitrile hydratase also involves a subunit swapping step and may require ATPase activity for metallocofactor insertion, but shows no requirement for a COG0523 metallochaperone. (854−856) The GTPase activity of Nha3 is required to obtain active nitrile hydratase in cells. (857) This work on Fe-dependent nitrile hydratase raises many questions, but a molecular switch in Nha3 from the ON to the OFF state might be anticipated to modulate the stability of the FeII-α complex, so that subunit swapping can occur. The catalytic FeII site in FeII-dependent nitrile hydratase is buried at the α–β subunit interface; (192) the same is true of the CoII-dependent enzyme (Figure 49B). (858) However, it remains unclear as yet why nitrile hydratase activation requires a specific metallochaperone, and molecular details of cofactor biogenesis remain largely lacking. One possibility is that the metallochaperone coordinates maturation of the cofactor beyond insertion of the metal, including selective oxidation of the two coordinating cysteines in the first coordination shell; (859) arguing against this is the finding that in at least one case, the α-subunit alone has significant catalytic activity and harbors a fully post-translationally modified metal site. (860)

Figure 49

Figure 49. A) Proposed activation mechanism of bacterial Fe-dependent nitrile hydratase (NHase) in which FeII is transferred from the COG0523 metallochaperone, Nha3. (844) Once the α-subunit is loaded with metal, this facilitates dissociation of the OFF-state chaperone and subunit swapping into an “immature” α2β2 tetramer via an as yet unknown mechanism. B) Ribbon representation of the structure of an αβ heterodimer of mature FeIII nitrile hydratase from Rhodococcus erythropolis AJ270 (2QDY), with the alpha subunit shaded green and the beta subunit shaded blue. (861)

High Resolution Image
Download MS PowerPoint Slide
There are no known metalloenzyme clients for any bacterial Zur-regulated COG0523 protein, although the functional characterization of B. subtilis ZagA and its connection to sustaining folate biosynthesis and one-carbon (1C) metabolism under conditions of ZnII restriction has been reported. (761) FolE catalyzes the rate-determining step in bacterial folate biosynthesis and is an obligatory ZnII-dependent enzyme. (862) When ZnII levels become scarce, Zur regulates the expression of ZagA and if this fails, a ZnII-independent paralog, FolEB, which uses other divalent cations for activity, in an effort to sustain folate biosynthesis under these conditions (Section 4.3.3). (762) The simplest mechanistic model is that ZagA and apo-FolE physically interact in a way that is modulated by G-nucleotide status of ZagA, but this is not yet known. Attempts to obtain a preparation of metal-free apo-FolE from ZnII-bound FolE purified from E. coli has generally met with limited success, consistent with the idea that the ZnII bound to subunit interface in the folded decameric FolE oligomer may not be in equilibrium with the bioavailable ZnII pool. (863) Thus, other models may well be required to explain the metallochaperone activity of ZagA. It is the case that ZagA binds and hydrolyzes the alarmone, 5-amino 4-imidazole carboxamide riboside 5′-triphosphate (ZTP), which accumulates in cells under 1C folate deficiency, (198) the functional significance of which is remains unknown. (761)
The previously noted strong evolutionary similarity of bacterial ZigA/ZagA to the mycobacterial Mrf (MPY recruitment factor) deserves comment. (691) Mrf binds to the 30S subunit of the ribosome and loads mycobacterial factor Y (MPY) onto the 70S ribosome to induce ribosome inactivation under conditions of ZnII deficiency. (697) This suggests the possibility that some COG0523 proteins function by physically associating with the ribosome. Here, ZnII might be installed into an obligatory ZnII-dependent protein(s) by ZigA cotranslationally, into the nascent unfolded chain, thus avoiding kinetic or thermodynamic trapping of the metal into a holoprotein once folded. (864,865) Indeed, Mycobacterium smegmatis, the soil saprophyte of M. tuberculosis, encodes both ZigA and Mrf, and both are regulated by Zur; this may allow the organism to fine-tune control of ribosome function and metalloproteome maintenance under conditions of nutrient ZnII restriction. Ribosome profiling (866) in a ΔzigA strain grown under conditions of ZnII deficiency would provide strong support for a cotranslational metal insertion model that is stimulated by ZigA, for example.
5.1.4.4. Eukaryotic COG0523 Proteins
Landmark studies published in 2022 establish that a eukaryotic COG0523 protein, named ZNG1 for Zn-regulated GTPase metalloprotein activator 1, fulfills all the criteria that one would anticipate of a metallochaperone for methionine aminopeptidase-1 (METAP1) in vertebrates and the yeast Saccharomyces cerevisiae, a finding recapitulated recently in plants. (56,814,846) METAP1 is a cell-abundant binuclear ZnII enzyme that associates with the translating ribosome, cleaving the N-terminal methionine (Met) in over 50% of the nascent polypeptide chains. (867) This activity is critical to cell physiology since the identity of the N-terminal residue in a mature protein can dictate organellar destination in eukaryotic cells. There is as yet no evidence that bacterial or archaeal methionine aminopeptidase is dependent upon a COG0523 for metalation under conditions of metal restriction, although this is clearly worthy of investigation.
METAP1 and ZNG1 form a complex mediated in part by a conserved N-terminal CPxLxP (C, Cys; P, Pro; L, leucine; x, any amino acid) sequence motif in ZNG1 not present in bacterial homologues, and an N-terminal Cys6His2 cross-brace zinc finger (ZF) structure in METAP1 (Figure 50A). (56) The NMR solution structure of the complex reveals a core ββαβα, cross-brace fold with a 1,3,2 antiparallel β-sheet arrangement, in which the outer β2 strand hydrogen bonds with the ZNG1-derived peptide (Figure 50B). This structure fully explains the sequence conservation of the METAP1 interacting motif in ZNG1, from lower eukaryotes to plants and animals across millions of years of evolution, conforming the general sequence “[D/E]nψPxLVp (Figure 50C). (56) These studies suggest that GTP hydrolysis by ZNG1 simulates metalation of the empty METAP1 active site above that which could be obtained with GMPPNP or the OFF-state product GDP, consistent with the mechanistic model shown (Figure 51A).

Figure 50

Figure 50. Solution structure of the major site of interaction between eukaryotic ZNG1 and METAP1. (56) A) Ribbon representation of the structure of the N-terminal tail of ZNG1 (blue) bound to METAP1 ZF domain (PDB 7SEK). B) Same structure as in panel A, with the N-terminal tail of ZNG1 shown in stick (residue numbers in bold italic) and the ZF domain shown in spacefill colored as an electrostatic surface rendering. C) Conserved features of all ZNG1 N-terminal METAP1-interacting motif, with the sequence of the vertebrate ZNG1 shown below the weblogo. (56)

High Resolution Image
Download MS PowerPoint Slide

Figure 51

Figure 51. A mechanistic model of ZNG1 function. (56) A) Cartoon depiction of the catalytic cycle illustrating one round of metal insertion. B) An AlphaFold2 model of the complex between human ZNG1 (green) and human METAP1 (gray). The region of the model that corresponds to the NMR solution structure is boxed (cyan; NMR). A close-up of the interfacial region of highlighting the close juxtaposition of the native binuclear ZnII center (black) with its ligands indicated and the CXCC motif of ZNG1 (green ribbon). C) A potential metal–ligand exchange pathway that connects the ZnII bound to the CXCC motif of ZNG1 (white), through a “third” metal site (gray) and the binuclear active site of METAP1 (black). The residues highlighted (yellow stick) are conserved in METAP1 sequences and thus may function to guide the metal into the (empty) active site via a sequential metal ligand-exchange process.

High Resolution Image
Download MS PowerPoint Slide
Although structures of the complex are lacking, an Alphafold2 model (Figure 51B) places the CXCC motif of ZNG1 in close proximity to the active site of METAP1, seemingly poised to undertake a sequence of ligand exchange reactions and metal transfer down a thermodynamic gradient (see Figure 34, top). In fact, some METAP1 structures incorporate a “third” metal binding site, positioned just above the binuclear active-site cluster, coordinated by a series of conserved residues, a subset of which function in substrate binding and catalysis (Figure 51C). (868) One hypothesis is that these residues play multiple roles, one of which is to initiate ligand exchange with the metallochaperone. However, the nature of the driving force for metalation (Figure 34) awaits further study. The KZn values for isolated ZNG1 and METAP1 are very similar in the absence of G-nucleotide; furthermore, ZnII slightly inhibits the GTPase activity of ZNG1, like in EcYjiA, its closest bacterial homologue; (691) addition of Zn2 METAP (with the active sites empty) to ZnII-ZNG1 restores the GTPase activity to that of apoprotein, as expected if metal moves from donor ZNG1 to client.
It is interesting to note that the N-terminal ZF domain of METAP1 is also required for association with the translating ribosome, via the C-terminus of the β-subunit of the nascent polypeptide-associated complex (NAC). NAC is analogous to bacterial trigger factor, which is anchored near ribosome exit tunnel on what is considered a “universal” ribosome docking site. (865) This suggests a model, not yet tested, in which the β-NAC peptide and the N-terminal tail of ZNG1 are in competition with one another and given increased cellular expression of ZNG1 under conditions of ZnII limitation, (56) may help to ensure that inactive metal-free METAP1 does not readily associate with the ribosome. Alternatively, the metalation status of METAP1 may regulate ribosome binding in some way.
Finally, of the hundreds of zinc metalloenzymes present in even a simple eukaryotic cell, (869) why does METAP1 require what may be a dedicated metallochaperone under conditions of nutritional zinc deficiency? One possibility is that cotranslational processing of the nascent chain by a highly cell-abundant enzyme is simply “too big to fail”. Indeed, cell abundant proteins in specialized cells often have their own specialized protein chaperone as a means to maintain proteostasis, exactly parallel to metallostasis (Figure 1). These ideas are in line with metalloproteomics experiments in Saccharomyces cerevisiae, which reveal that 20 cell-abundant ZnII metalloenzymes, if fully metalated, account for ≥90% of the zinc quota in the cell, one of which is methionine aminopeptidase. (869) It will be interesting to discover the extent to which cells use this or other approaches to prioritize utilization of ZnII under conditions of metal nutrient limitation.

5.2. Copper Storage Proteins and Metal Allocation

5.2.1. Overview of Copper Toxicity

Cu is an essential micronutrient required as a cofactor for various cuproenzymes in nearly all forms of aerobic life. Copper homeostasis, like the homeostasis of other transition metals, (120) is tightly controlled by the combined action of transcriptional regulators that sense copper levels (Section 3), efflux pumps, storage proteins and metallochaperones, the latter of which allocate Cu to partner proteins without dissociation into solvent. (807,870−872) It is worth noting that the number of cuproenzymes that are resident in cytoplasm tends to be quite small (or zero) in nearly all organisms, with cell-abundant cuproproteins sequestered in intracellular membrane vesicles, e.g., thylakoids, intracytoplasmic membrane structures, (873) e.g., for particulate methane monooxygenase (pMMO), (112) or simply embedded in the plasma membrane with active sites facing out. As such, Cu present in the cytoplasm may be recognized as toxic. A major mechanism of bacterial resistance to Cu toxicity is cytoplasmic efflux of CuI via one or more P1B- type ATPase exporters, e.g., CopA or CopB (Section 5.3.2). These transporters function alongside a cytoplasmic CuI netallochaperone that plays a role in buffering cytoplasmic CuI to very low concentrations (Figure 8) and is capable of delivering CuI directly to an exporter for efflux (874,875) and in some cases, to the CuI-sensing transcriptional repressor that orchestrates the adaptive response Cu toxicity (Figure 34). (876−878)
In Gram-negative bacteria, the fate of this periplasmic CuI differs depending on the organism, but includes oxidation by a multicopper oxidase (MCO) to the less toxic CuII, (879−882) sequestration by periplasmic CuI/CuII binding proteins, (883−885) and capture by CusF and efflux via the periplasm-spanning CusABC system. (886,887) More recently, a novel membrane-anchored and surface-exposed Cu binding lipoprotein, CopL, from methicillin-resistant Staphylococcus aureus (MRSA), was shown to provide hyper-resistance against Cu toxicity in clinically relevant MRSA strains; this activity is independent of the virulence-associated CopB efflux pump, but whose function is dependent on CuI binding to a novel CuI-binding architecture. (170,888,889) This suggests that extracytoplasmic Cu sequestration may be important mode of resistance to copper toxicity. Within the cytoplasm, CuI coordination and sequestration represents another mode of bacterial resistance to copper toxicity, and in some cases, but not all, the genes that encode these proteins are regulated by a Cu-sensing repressor. These include a mycobacterial metallothionein, MymT, (359,890) CutC found in some Firmicutes, (891) a CopZ, (876) and more recently, proteins termed copper storage proteins (Csp) discussed here. (892,893)

5.2.2. Copper Storage Protein Structure

Csps are a family of closely related cysteine thiolate-rich four-helix bundle proteins that form homotetramers that are capable of binding up to 80 CuI per tetramer. (894) The inner surface of this tube-like structure harbors ≈13–20 cysteines, and often a cluster of histidine residues at the open end or “mouth” of each four-helix bundle (Figure 52). Though first discovered as a soluble, periplasmic binding protein (Csp1/Csp2) in methanotrophs and proposed to function in the metalation of pMMO, the cytoplasmic Csp3 is now considered the far more prevalent Csp, and has now been identified in at least ten different bacterial phyla. (892) Csp3-like proteins differ slightly in CuI-binding stoichiometries, but all contain a common core of three tetrameric Cu-binding clusters (which collectively bind 12 Cu), with the additional coordination chemistry variable above or below this cluster (Figure 52). Csp3s bind CuI with high equilibrium affinities (KCu ≈ 1016 M–1) to multiple classes of sites that vary slightly in affinity. Although the details differ for different Csp3s, the kinetics of CuI binding are generally quite fast but only for an initial phase, with the kinetics of CuI removal by a high affinity Cu(I) chelator very slow, beyond the first couple mol equivalents of CuI. (894,895)

Figure 52

Figure 52. Crystallographic structures of one protomer of tetrameric copper storage protein (Csp) from the organisms indicated. Left, Streptomyces lividans Csp3 (6EK9); (896) center, Methylocystis sp. strain Rockwell Csp3 (6ZIF); and right, Methylosinus trichosporium OB3b Csp3 (5ARN). CuI ions (copper spheres) and CuI-ligating residues are shown in stick. The tri-tetrad inner core of 12 CuI ions is common to all Csps, with the base and mouth regions variable.

High Resolution Image
Download MS PowerPoint Slide

5.2.3. Copper Storage Protein Function

Despite their discovery some 10 years ago, the functional role(s) of Csp3 in bacterial cells remain largely enigmatic. These CuI binding properties are potentially consistent with a role in Cu storage under Cu-replete or Cu-toxic conditions, (897) or as part of an “anticipatory” Cu-toxicity response, when cells are chronically metal-starved, and thus particularly vulnerable to mismetalation mediated by even modest Cu toxicity (see Section 5.3.2). (174) There is also some evidence to suggest that Csp3 can function as a cushion for very high levels of CuI in some specialized cases, e.g., where the ability of cells to efflux CuI is compromised (in a ΔcopA strain). (898) However, csp3 expression, unlike the primary CuI toxicity regulon discussed above, is rarely regulated by Cu stress.
There is recent evidence in support of the original proposal (893) that Csp3 functions as a copper chaperone to mediate CuI delivery to a client enzyme, in Bacillus subtilis. (899) Here, it was demonstrated the CuI-loaded Csp3 is capable of metalating a Cu-requiring endospore-resident multicopper oxidase CotA, a laccase, which is one just two known Cu-requiring enzymes in B. subtilis. Importantly, B. subtilis Csp3 does not protect cells from Cu intoxication and thus must play some other role. Both CotA and Csp3 are upregulated during late stages of spore formation (900) and appear to protect spores against hydrogen peroxide and UV light via production of a melanin-like molecule. Csp3 stimulates the extent of CotA metalation, albeit over very long time points (hours), as evidenced by activity assays, and a soluble CuI chaperone CopZ does not function in this assay, consistent but not confirmatory of a specific protein–protein interaction. (899) The extent to which Csp3s broadly function in metalloenzyme maturation is unknown.

5.3. Metal Ion Transport and Supramolecular Assemblies

We next describe recent biologically focused discoveries in transition metal transport in bacterial cells that highlight the involvement of accessory proteins that function as part of supramolecular efflux assemblies to facilitate metal transport across a membrane. These transporters and cognate accessory proteins may function as membrane-associated metallochaperones, in the regulation of transport, or in the metalation of metalloenzymes. The major metal transporter discussed here are a poorly understood subclass of the P-type (phosphorylation-type) ATPase efflux pumps, (901) notably those derived from the P1B-6 subclass of transporters. We place these discoveries in the context of a discussion of new structural advances in our understanding of the metal selectivity and transport mechanisms of P-type ATPases.

5.3.1. P-Type ATPases

P-type ATPases are conserved across all kingdoms of life and collectively mediate the movement of metals, protons and lipids across biological membranes from the cytoplasm to the extracellular space or an extracytoplasmic compartment or organelle. (901) All P-type ATPases possess three domains, a transmembrane domain (TM) composed of six to eight membrane-spanning α-helices that comprise the transport channel, an ATP-binding domain composed of nucleotide (N) binding and phosphorylation (P) subdomains, and an actuator domain (A) that allosterically signals a change in the phosphorylation status in the P domain to the TM domain to allow for transport. Metal transport is well-modeled by the alternating states Post-Albers cycle (902) which involves two distinct conformational states, an inward-facing E1 state and an outward-facing E2 state. The E1 state has high affinity for the cognate metal(s) and binds and hydrolyzes ATP; formation of the phosphoenzyme intermediate on an Asp residue in the P subdomain is coincident with formation of the low-metal binding affinity E2 (E2P) state, which then releases Pi (E2) and ultimately relaxes to the E1 state (Figure 53). Transient phosphorylation drives significant conformational changes in the membrane-spanning domain of transporter required for transport against a concentration gradient. (903,904) These changes in transporter domain ultimately expose the metal egress pathway on the opposite side of the membrane, thus allowing for vectorial or unidirectional transfer across the membrane. Some P-type ATPases including the P1B-type subfamily, contain ancillary domains, often metal binding domains (MBDs), or other regulatory domains that can modulate or gate transport activity (Figure 54, top).

Figure 53

Figure 53. Metal transport is well-modeled by the alternating states Post-Albers cycle (902) which involves two distinct conformational states, an inward-facing E1 state and an outward-facing E2 state. The E1 state has high affinity for the cognate metal(s) and binds and hydrolyzes ATP; formation of the phosphoenzyme intermediate on an Asp residue in the phosphorylation subdomain is coincident with formation of the low-metal binding affinity E2 (E2P) state, which then releases Pi (E2) and ultimately transitions to the E1 state.

High Resolution Image
Download MS PowerPoint Slide

Figure 54

Figure 54. P-type ATPase transporter structures derived from the P2 and P1 subfamilies. Schematic representations (top) are color-coded as shown in the ribbon diagram of each strucure (middle). ADP-AlF4–-bound human (h) SERCA1a (PDB 5XA8), AMPPCP-bound hSPCA1 (8IWR), ligand-free AfCopA (7R0H), AlF4–-bound SsZntA (4UMW), BeF3–-bound sCoaT (7QC0). The Post-Albers conformational state is indicated in each case, as well as ligands resolved in each structure. Bottom, coordination chemistry of each transporter as determined by X-ray crystallography (rSERCA; 1SU4) or cryo-EM (hSPCA1; 8IWR, CaII bound; 7YAJ, MnII bound) or via X-ray absorption spectroscopy and mutagenesis studies (others). The minimal sequence motif characteristic of each subfamily is shown below the metal coordination domain model for the three P1B transporters, positioned on TM4 or TM6 helices, as indicated. (909) The structures of P1B-3, P1B-5, P1B-6 and P1B-7 transporters are not yet known.

High Resolution Image
Download MS PowerPoint Slide
5.3.1.1. P2A-Type CaII and CaII/MnII Transporters
Metal-transporting P-type ATPases typically form intramembrane metal sites derived from the TM4 and TM6 helices in the core TM domain, with the identity of these metal-binding residues strongly impacting the metal specificity (or promiscuity) of the transporter. (905) This is true for both the paradigm sarco(endo)plasmic reticulum calcium ATPase (SERCA) pumps and related transporters from the P2A subclass that function in the ER and cis-Golgi in vertebrates, as well as the “heavy metal” transporting P1B-type transporters described below (vide infra) (Figure 54). Of particular interest regarding metal specificity of P2A ATPases is a suite of cryo-EM structures of the secretory-pathway CaII-ATPases (SPCA1), long known to transport MnII as well, (906) which function both to metalate key metalloenzymes in this compartment and beyond, including those associated with O-linked glycosylation, while also mediating MnII detoxification in neuronal tissues by actively pumping cytosolic MnII into the secretory pathway (Figure 55). (907)

Figure 55

Figure 55. A suite of cryo-EM structures of the secretory-pathway CaII-ATPases (hSPCA1) which pump CaII and MnII into the trans-Golgi lumen. (907) The domains of the transporter are colored as in Figure 55. CaII ions, green; ATP, ADP and Pi, purple stick. New insights into the mechanism of CaII release obtained from inspection of the CaII-bound E2P conformation, prior to hydrolysis of the covalent phosphate ester (E2.Pi) and Pi dissociation (E2), were derived from this work.

High Resolution Image
Download MS PowerPoint Slide
These cryo-EM structures provide high resolution snapshots of all intermediates in a complete transport cycle and nicely illustrate the large amplitude conformational transitions of one domain relative to another that occur as a function of metal binding and phosphorylation status of the P-domain (Figure 55). (907) Pertinent to the discussion here, these structures provide unprecedented insights into the relaxed metal specificity of SPCA1 relative to SERCA enzymes, which are strictly CaII transporters (Figure 54, bottom) (see Section 3.3.3). (908) In SPCA1, MnII was found to bind to the same metal binding site II as CaII, but more tightly in forming a lower coordination number square pyramidal complex that results in residues on TM6, in particular D742, moving closer to those ligands derived from TM4, which donates three main-chain carbonyl oxygens to the metal. (908) In the E2-Pi modeled complex that contains no metal, those residues that coordinate the CaII or MnII move far apart, allowing for metal release into the trans-Golgi lumen in eukaryotic cells. In addition, SPCA1 lacks the second metal site found in SERCA (metal site I), which may help enforce MnII transport activity by SPCA1 relative to SERCA (Figure 54, bottom).
5.3.1.2. P1B-Type Transporters
It is generally accepted that P1B-ATPases conform to seven distinct structural subclasses, denoted P1B-1-P1B-7, all of which transport “heavy” metals (Figure 56). (901,910) Many P1B transporters differ significantly from the P2A transporters discussed above in that they have evolved MBDs appended to either the N-terminus or C-terminus on the core transporter structure (Figure 54, right). These MBDs often, but not always, provide hints as to the metal selectivity of the pump itself, and in the case of the P1B-1-P1B-3 transporters that transport CuI and ZnII, are in direct communication with the cytoplasmic environment. (911) Their metal affinities are tuned to the bioavailable “pool” of each metal. CuI-transporting P1B-1 ATPases nearly always harbor an N-terminal domain (or multiple repeated isostructural domains) that adopt a ferredoxin-like or cupredoxin fold, that accepts metal from an isostructural copper metallochaperone (Figure 34), likely via a metal ligand exchange intermediate. (804,807,875,912) These domains are not required for transport activity, and may well be regulatory; (911) indeed, recent structural studies suggests that that the CuI chaperone has direct access to the high-affinity metal binding site in the membrane, where an intermediate Met residue mediates metal dissociation. (913)

Figure 56

Figure 56. Primary structure determinants of P1B-type ATPase transporters that impact their transition metal specificity profiles. A) Primary structure motifs in the TM4, TM5 and TM6 that impact metal specificity of P1B subclass ATPases. B) Ribbon structure of AfCopA in the E1•CuI state (7R0H). Inset, metal binding site region with the metal ligands and the Tyr-Asn sequence from TM5 shown in stick. (913) C) Ribbon structure of S. sonnei ZntA in the E2P state (4UMV), with the structure of the EcZntA metal binding domain also shown (1MWZ). Inset, metal binding site region with the four known metal ligands shown in stick. (918)

High Resolution Image
Download MS PowerPoint Slide
Like the P2A transporters discussed above, the metal specificity of a P1B-ATPase is also impacted by conserved signature sequences present in transmembrane segments 4, 5, and 6 (TM4, TM5, and TM6; numbered such that the first two helices in the P1B transporters not found in P2A transporters are labeled TMA and TMB (901)) (Figure 54, top; Figure 56). The short TM4 sequence CPC is characteristic of both P1B-1 and P1B-2 transporters, widely accepted to transport CuI and ZnII/CdII/PbII, respectively, but each differs on the basis of residues in TM6 (Figure 56). In P1B-1 transporters, a Met in the TM6 is part of a MxxSS sequence coordination bond to the TM CuI site in the E1 state to complete a trigonal planar CPC. . .M chelate (Figure 54, bottom). (914) In P1B-2 transporters, a bidentate Asp completes a distorted tetrahedral complex in the E1 state, except for the PbII enzyme, which excludes one the O ligands of Asp replacing it with its “stereochemically active” lone pair (Figure 54, bottom). (915,916)
A number of high-resolution P1B-type ATPase structures have now been solved to date in various conformations within the Post-Albers catalytic cycle (Figure 54, middle). These include the P1B-1 CuI-specific CopA from Archaeoglobus fulgidus in the CuI-free E1 state (Figure 54, middle), CuI-specific CopA from Legionella pneumophilia in the CuI-free E2•Pi and E2 states, ZntA, a P1B-2-subclass ZnII transporter from Shigella sonnei, in the metal-free E2P and E2-Pi states (Figure 54, middle). (913,917,918) Each structure was used to inform the coordination chemistry confirmed by X-ray absorption spectroscopy and other methods (Figure 54, bottom). More recently, the structure of a canonical P1B-4-type enzyme has been solved, Sulfitobacter sp. NAS14–1 sCoaT, in the metal-free E1 state, as representative of the simplest member of the entire P-type ATPase superfamily (259,910) and which lacks ancillary MBDs altogether (Figure 54, top). (909) P1B-4-subclass ATPases were originally proposed to be CoII-specific transporters, (919) but may well be characterized by a relaxed metal specificity profile; (920,921) in at least three cases, they have been shown to be FeII-efflux transporters in bacterial cells. (920,922−924) Here, FeII transport is coupled to the H2O2 stress response and the expression is regulated by PerR or Fur (Section 3). These are extremely interesting transporters from a therapeutic perspective, since although they were functionally characterized first in the model organisms B. subtilis (PfeT), P1B-4 ATPases are present in many bacterial pathogens, including Mycobacterium tuberculosis (CtpD), (921) Group A Streptococcus (PmtA), (923−925) and Listeria monocytogenes (FrvA). (920)

5.3.2. Ferrous Iron Transport into Bacterial Ferrosomes

Given its abundance in the metalloproteome and the extremely low solubility of FeIII in aerobic environments, iron is considered a precious metal nutrient for nearly all cells. As a result, most cells go to extraordinary lengths to scavenge and if possible, store, this metal nutrient, to protect from subsequent Fe starvation. Ferrous iron, FeII, is the oxidation state that is bioavailable in the cytoplasm and dominates extracellularly in oxygen-deplete anaerobic and acidic environments of the stomach, for example. Unlike FeIII, FeII is highly soluble and exchange labile, and recent work reveals that FeII is bioavailable in these specific niches and recognized to play an important role in nutritional immunity against invading pathogens. (926−928)
5.3.2.1. Overview
It has long been known that one way to obtain a “pool” of bioavailable FeII required for protein metalation in bacterial cells is to store it in intracellular compartments, notably heme-based bacterioferritins, ferritins, or in mini-ferritins, derived from the superfamily of DNA-binding proteins from starved cells (Dps), all of which are protein-coated biomineralized compartments of FeII and FeIII. However, iron storage in these compartments is dependent upon the availability of oxygen, and therefore cannot function efficiently in an anoxic environment. The only other known strategy available to protect bacteria from iron intoxication is to efflux the metal from the cytoplasm, via ferrous ion transporters. Their structures and transport mechanisms have been reviewed recently, (929,930) and we focus on one here.
5.3.2.2. Ferrosomes and P1B-6 ATPase Transporters
There is considerable excitement around the P1B-6 subclass of transporters, given the fact that up until recently they were poorly characterized as to both metal transport specificity and function in cells (Figure 56A). A preliminary genomics analysis suggested that these transporters might transport FeII, given an occasional genomic colocalization with feoB, known to encode the primary bacterial ferrous ion uptake transporter FeoAB. (910,931) Remarkably, this prediction turns out to be true, although only in part. Seminal work from the Komeili group published in 2010 revealed that lipid-encased electron dense particles exist in the anaerobic bacterium Desulfovibrio magneticus, long studied as a model organism for the formation of “bullet-shaped” magnetosomes, which biomineralize Fe as magnetite (Fe3O4) in a membrane-encased “organelle” organized into long chains that use geomagnetic fields for direction sensing. (932,933) In that work, the presence of another electron-dense particle rich in Fe and phosphate and surrounded by a lipid membrane was noted, and whose biogenesis was clearly distinct from that of magnetosomes. (933)
In more recent studies, these authors further characterized these amorphous granules as specifically enriched in Fe and phosphate in what are now known as ferrosome organelles (ferrosomes), and identified a three-gene cluster arranged into a putative operon responsible for their formation as encoding ferrosome-associated (Fez) proteins (Figure 57A,B). (674) Core components of this operon are FezA, a predicted α-helical transmembrane membrane protein (DUF6110) and FezB, a P1B-6-ATPase transporter, as well as a third small protein denoted FezC or FezX, depending on the organism, of unknown function. P1B-6-ATPases harbor a characteristic SPC motif in TM4 and a TxxNHN motif in TM6, which may well impact metal specificity (Figure 56A). Purification of ferrosomes from bacterial cells reveal that both FezA and FezB are greatly enriched in these electron-dense granules, with no detectable FezX. (673) Thus, although biochemical experiments are not yet available, FezB is clearly an FeII transporter, but is tasked with moving metal across an intracellular membrane, a remarkable finding given that bacteria are not broadly assumed to harbor intracellular membrane-encased organelles.

Figure 57

Figure 57. P-type ATPase transporter genomic localization and transporter function. A) Genomic organization of the fez operon regions in selected organisms. Fur boxes and two transcription start sites (TSS) are indicated on the C. difficile genome. (673,674,934) B) Ferrosome function in C. difficile. Ferrosomes assemble at or near the plasma membrane. FeII is driven into the interior of the ferrosome by the P1B-6-class transporter, FezB. The function of FezA (DUF1610) is unknown but may assist in deforming the plasma membrane to facilitate ferrosome biogenesis, or play some other role in activating FeII transport. C) Three distinct P1B-type transporters in M. tuberculosis that function as part of metal efflux platforms and incorporate a PacLn (DUF1490) homologue. (935)

High Resolution Image
Download MS PowerPoint Slide
Ferrosomes have now been imaged by energy-dispersive X-ray spectroscopy (EDS) and characterized in the Gram-positive, spore-forming, intestinal anaerobe and human pathogen, Clostridioides difficile. They have been shown to play an integral role in iron homeostasis in this organism and defense against nutritional immunity mediated by CP stress during an infection. (673) In C. difficile, ferrosomes function in iron storage, and the expression of fezX and fezAB is regulated by Fur and inducible by the metal chelator dipyridyl, often used to starve cells for Fe. These genes are constitutively expressed in a Δfur strain and appear to protect C. difficile from Fe toxicity during an acute switch from a Fe-deplete to an Fe-replete growth conditions (Figure 57B). (673) High-resolution tomographic imaging of C. difficile engineered to constitutively express fezAB suggests that although these compartments are clearly cytoplasmic, ferrosome clusters are frequently found adjacent to the cell membrane, for reasons that are not yet clear (Figure 57B). Heterologous expression of fezXAB, but not fezAB alone, in B. subtilis, reveals that despite the fact that FezX cannot be detected in these organelles isolated from C. difficile, it is required for their assembly, at least in this heterologous host organism. (673) CP induces the transcription of fezXAB in cell culture and in infected animals and it appears that FezB and the inhibitory impact of CP in nutritional immunity and function most acutely during colonization of the gastrointestinal tract during infections.
These comprehensive studies establish a new paradigm for membrane-encased organelles or compartments in bacteria, but also raise many important questions that bear on P1B-type ATPase function, assembly and mechanism like other work described below (Section 5.3.4), (935) The biological function of the DUF6110 protein FezA is completely unknown. Does it function as a novel ferrosome-membrane-associated metallochaperone or scavenger of cytoplasmic FeII funneled to FezB, as a regulator of FezB transport activity, a membrane bilayer remodeling protein, or a canonical protein chaperone that enables assembly of active and folded FezB in the membrane? FezX (29 amino acids), on the other hand, is perhaps reminiscent of small proteins (<50 residues) that regulate bacterial MnII homeostasis, (936) whose functions are only beginning to gain some clarity. (937) How does FezX, not found in all fez operons, impact this process in C. difficile? Detailed biochemical reconstitution and cell biological studies will be required to address these exciting questions, answers to which to greatly enhance our understanding of P1B-6-ATPase transporters.

5.3.3. Membrane Efflux Platforms

A prevailing view in the P-type ATPase field is that these transporters tend to function independently, with the exception of soluble cytoplasmic metallochaperones most strongly associated with the P1B-1-type CuI chaperones, which deliver metal to the transport site. (799,913,938) The work on bacterial FezA-FezB described above suggests that this need not be in the case. In fact, it is interesting to consider the singular example of a membrane-anchored metallochaperone, CupA, and its cognate P1B-1-ATPase CopA in the Gram-positive respiratory pathogen Streptococccus pneumoniae in this context. (939) CupA was shown to be required for the cellular copper resistance, and a heterologous CopZ-like chaperone which adopts a distinct fold could not substitute for CupA in cells; further, membrane anchoring was required for resistance against copper toxicity. (912) It is not yet known if and how S. pneumoniae CupA and CopA physically interact, or form a stable complex in the membrane, but this is of course the expectation. (940)
5.3.3.1. DUF1490 Proteins Impact P1B ATPase Function
Mycobacterium tuberculosis encodes seven P1B-type ATPases encoded by ctpA-D, ctpG, ctpJ and ctpV that mediate comprehensive resistance to transition metal toxicity in the hostile intracellular environment of the phagolysosome. (941) These include at least three CuI-specific transporters (CtpA, CtpB and CtpV), with CtpV, regulated by the CuI-sensing repressor CsoR (Section 3), required for full virulence of the organism. (203,374,942) Interestingly, the genes encoding CtpC, CtpG and CtpV are each immediately downstream of a gene encoding a DUF1490 protein (Figure 57C), found in a genomic arrangement that is reminiscent of fezAB in D. magneticus and C. difficile; in fact, DUF1490 and DUF6110 (fezA) proteins are distantly related to one another in that both are predicted to be α-helical membrane proteins with a C-terminal tail. (935) Boudehen et al. have extensively functionally characterized the DUF1490 immediately in front of CtpC, a P1B-6-like ATPase transporter that harbors a canonical P1B-1- and P1B-2 CPC motif in the TM4 associated with CuI an ZnII transporters (Figure 54, bottom), but contains a canonical P1B-6-class TxxNHN motif in TM6 (Figure 56A), consistent with a global SSN analysis that places these CPC enzymes in their own cluster of closely related sequences. (910) The expression of pacL1-ctpC is induced specifically by zinc toxicity in both liquid cultures and infected macrophages, with the identity of the zinc-sensing efflux repressor not yet known. (935,943) These authors establish that this DUF1490, renamed PacL1, for P-ATPase-associated chaperone-like protein 1, binds ZnII via a C-terminal “DLHDHDH” metal binding motif, and may function as a metallochaperone for CtpC at high cellular ZnII toxicity. (935) Furthermore, PacL1 is also required under conditions of modest ZnII toxicity in a way that is independent of its metal binding function where it plays a role as a scaffold to stabilize CtpC in the plasma membrane. A strain harboring a deletion of pacL1-ctpC is sensitive to zinc toxicity while single pacL1- or ctpC-complemented strains fail to rescue this growth phenotype; this is consistent with a model in which PacL1 and CtpC are both required to efficiently efflux ZnII from the cytoplasm. (935)
Light microscopy-based imaging reveals that CtpC and PacL1 physically interact and form high molecular weight complexes in the mycobacterial membrane that the authors dub “metal efflux platforms” (935) that enhance resistance to metal poisoning. Such membrane microdomains like these, which impact enzyme activity, may be more common in bacterial membranes than previously believed. (937,944) Although more work needs to be done, orthologs PacL2, encoded in front of CtpG and containing a single D-E (Asp-Glu) sequence, and PacL3, in front of CtpV and harboring a “DDGHDH” sequence in its C-terminal metal-binding motif (Figure 57C) are partially redundant when expressed at high intracellular levels. This is a function of their conserved motifs in the protein chaperone-specific regions of these proteins rich in Glu/Ala repeats are reminiscent of the canonical protein chaperone GroEL1. (935) These Glu/Ala motifs mediate a physical interaction with CtpC in these membrane microdomains. (935) Biochemical reconstitution experiments and structural studies of a PacL1-CtpC complex will be required to significantly extend these studies.

5.3.4. Extracellular Metalation of Manganese Metalloenzymes

A final frontier in metalloprotein metalation is the metalation of exoenzymes, which include those enzymes secreted into the extracellular milieu as well as those that are integral membrane proteins whose active sites face outward, relative to the cytoplasm. There are a number of reports of P1B-type ATPases that exhibit metal efflux rates that are too slow to be effective in resisting the effects of cellular metal toxicity, and thus must play some other role. One potential role is in metalation of extracytoplasmic enzymes. (945,946) This becomes particularly critical for weakly binding metals, including MnII and FeII, which may be outcompeted by extracellular ZnII or CuII, which generally bind much more tightly to proteins (Section 2.2.2). (157) Seminal work by Robinson and co-workers showed that the twin-arginine translocation (Tat)-dependent secretion system, which exports folded metalloproteins, could be used to ensure cognate metalation by a MnII-metalloenzyme; (157) however, the Tat system only functions for soluble proteins and not integral membrane proteins or those anchored in the plasma membrane with their active sites facing out. This suggests the strong possibility that metalation with MnII might be dependent on a specific efflux transporter.
5.3.4.1. TerC Proteins Function in Exoenzyme Metalation
TerC proteins are a subgroup of the lysine transporter superfamily, and were originally described as mediating cellular resistance to tellurite oxide anion (TeO32–) in extremophiles, (947) via a mechanism connected to cell envelope stress, (948) but more recently linked to MnII homeostasis in both B. subtilis and E. coli. (949−951) These apparently conflicting accounts have now been reconciled by recent findings that reveal that TerC proteins function during protein export via the generalized secretory pathway that is dependent on the SecYEG holotranslocon, activated by the SecA ATPase, that collectively define a core component of the secretosome (Figure 58). (952) B. subtilis encodes three TerC homologues, two of which were renamed MeeF (metalation of exoenzymes with Mn) (formerly YceF) and MeeY (formerly YkoY), the latter of which is regulated by the yybP-ykoY Mn-sensing riboswitch in both B. subtilis and E. coli (named Alx) (Section 3.3.3). meeF appears to constitutively expressed in B. subtilis but can be upregulated further under conditions of acute cellular stress.

Figure 58

Figure 58. B. subtilis TerC proteins MeeF/MeeY physically associate with the SecYEG translocon-secretosome where the mediate the extracellular metalation of MnII-requiring soluble exoenzymes (slate square) and integral membrane proteins (represented by LtaS/LtaSa). Strains that lack both meeY/meeF genes result in general reduction in cellular protein secretion. Pac-man symbol (brown), Sec-associated extracellular processing enzyme. LtaS, major lipoteichoic acid (LTA) synthase; LtaSa, stress-induced LTA synthase. (952)

High Resolution Image
Download MS PowerPoint Slide
Remarkably, coimmunoprecipation experiments reveal a direct interaction of both TerC proteins with common components of the secretosome. This protein assembly must mediate cotranslational folding of secretosome clients since B. subtilis strains deficient in both MeeF and MeeY are strongly defective for protein secretion, independent of the metalation of a client (Figure 58). The ΔmeeF-ΔmeeY (FY) mutant strain can be rescued by induction of the expression of FtsH, an ATP-dependent metalloprotease that is known to clear “jammed” SecYEG translocons (Figure 58, left). The FY strains also secretes fewer extracellular proteases as well. Thus, TerC proteins appear to be integral components of the bacterial secretory apparatus, but also endow the secretosome with the ability to metalate exoenzymes, notably, the MnII-dependent lipoteichoic acid synthase (LtaS), the major lipotechoic acid synthase in B. subtilis (Figure 58, right). LtaS is an integral membrane protein that features an outward facing MnII active site that functions in cell wall homeostasis. These findings in B. subtilis appear general, since heterologous expression of TerCs from L. monocytogenes and B. anthracis in B. subtilis complements the FY mutant, restoring fitness and normal levels of protease and MnII secretion into the spent growth media. (952)
The mechanism of cotranslocational metalation by TerC proteins of LtaS and perhaps other clients during other stages of growth is not known, but several possibilities exist. (952) One is that of a classical metallochaperone role that requires a functional interaction that drives ligand exchange and metal transfer between a conserved metal site on the MeeF/MeeY donor (951) and the client LtaS (Figure 34). Another is that TerC proteins increase the local concentration of MnII so that it better competes with more tightly binding metals, e.g., ZnII, in a way that does not require a direct physical interaction. Indeed, ZnII inhibits MnII-dependent LtaS metalation in a way that is exacerbated in the FY mutant, a finding consistent with other reports of the mismetalation of MnII enzymes by ZnII (Section 2.2.3). How LtaS holds on to its metal is not known, but kinetic trapping may well be involved. As this cotranslocational metal insertion system appears broadly conserved in all kingdoms of life, (952) there exists a strong motivation to identify of other TerC clients and fundamental bioinorganic chemistry features of this process.

6. Conclusions

Click to copy section linkSection link copied!
Metalloproteins, (953) ranging from proteins that use a metal as an enzymatic cofactor to those with structural metals that stabilize the protein fold, play myriad distinct and often evolutionarily conserved roles in all kingdoms of life. (1) Since approximately one-third of a typical bacterial proteome consists of metalloproteins, cells must ensure that the metalloproteome is metalated with integrity (3) and is resistant to mismetalation or under-metalation that might occur under conditions of environmental metal toxicity or metal restriction, respectively. Metallostasis (Section 2) encompasses all those processes that maintain the integrity of the metalloproteome and its associated metabolic and regulatory functions, the complexity of which is orchestrated by the upstream “sentinels” of this process, the metallosensors (Section 3). These model regulatory proteins and RNAs exploit fundamental coordination chemistry to drive allosteric switching (18) in a wide range of molecular scaffolds (575) to induce a specific biological output of physiological or biotechnological value. (19)
Historically, our understanding of the metal (ligand) specificity and regulatory mechanisms of these sensory biomolecules has come from the study of “one molecule at a time” through application of classical approaches in bioinorganic chemistry, biophysical chemistry and structural biology (Section 3). While powerful, this approach becomes limiting in the search for general features that impact metal selectivity in a superfamily of proteins that do not necessarily share all specific features of the model protein. (241,253) The gap between detailed chemical and physical studies and bacterial phylogeny continues to close with the emergence and broad application of sophisticated comparative genomics tools, e.g., sequence similarity network (SSN) and related tools, that allows our community to exploit continuous advances in genome sequencing. (239,245,246) Machine learning approaches now make it possible to leverage reasonable quality structures of a significant number of evolutionarily related, yet biologically uncharacterized proteins, to determine the degree to which findings from a model system, on metal selectivity, for example, can be extended to the group. (954,955) Emerging advances in machine learning will continue to move the field of structural biology from a labor-intensive to high-throughput structure and dynamics prediction pipeline that will ultimately inform comparative genomics efforts to understand molecular evolution of specificity. (956) This paradigm shift, well underway, will permit new applications of metalloregulatory proteins as biotechnological “parts”, while also informing new experimental strategies to develop new functionalities on these molecular scaffolds. (957) For example, high throughput selection and counter-selection strategies have led to the development of metal ion-specific DNAzymes with unprecedented metal specificity that continue to teach us how nucleic acids distinguish among various transition metal ions, while also creating reagents for applications as biosensing tools. (958)
Myriad cell-adaptive mechanisms, some newly discovered, (673,935,952,959) allow bacterial cells to maintain a metal quota that is capable of reacting to both chronic and acute metal starvation or toxicity, whatever the source. The expression of metalloprotein “paralogs”, which occurs often under conditions of restriction of a specific metal, (6,680) or alternatively in distinct organisms with distinct lifestyles and access to distinct profiles of metal nutrients, is a form of metalloproteome remodeling, the footprint of which is likely far more widespread than discussed here (Section 4). Indeed, one enzyme superfamily only briefly touched upon here encompasses the ribonucleotide reductases (Section 4.3.6) required for the synthesis of DNA in all life forms. (793,960) A number of clearly distinct metallo-forms exist and have been characterized in detail in microbes, which remarkably, includes a metal-free variant of the enzyme, (8,794,961,962) which in human pathogens could retain function under conditions of host-mediated metal restriction. (963)
We also discuss new mechanistic insights into “old friends”, the nucleotide-dependent metallochaperones (Section 5) (56,801,814,824) leading to a far more sophisticated understanding of enzyme metalation and transport, with many unanswered questions left for future work. (964,965) These adaptive responses, which occur in the context of an extraordinary ability of cells to “buffer” transition metal ions, enhance the resiliency of the metalloproteome while avoiding off-pathway metal toxicity. (966) They are remarkable examples of the extent to which evolution has gone to maintain the catalytic power of cellular metabolism dependent on metalloenzymes, particularly those enzymes involved in central carbon metabolism, one-carbon shuttling, scavenging of reactive oxygen species, and diverse aspects of the central dogma, when critical nutrient metals are in short supply. The implementation and development of global computational and experimental methods to map and explore the metalation status of the proteome, (62,160,184,967,968) i.e, metalloproteomics, (68,969) as well as workflows to better define the metal buffering components in cells, (119) promises novel insights into metalloregulation of cell biology and mechanisms of metallostasis from single cells to whole organisms in all kingdoms of life.

Author Information

Click to copy section linkSection link copied!
  • Corresponding Authors
  • Authors
    • Johnma J. Rondón - Fundación Instituto Leloir, Instituto de Investigaciones Bioquímicas de Buenos Aires (IIBBA-CONICET), C1405 BWE Buenos Aires, Argentina
    • Katherine A. Edmonds - Department of Chemistry, Indiana University, Bloomington, Indiana 47405-7102, United States
    • Joseph S. Rocchio - Department of Chemistry, Indiana University, Bloomington, Indiana 47405-7102, United States
    • Matias Villarruel Dujovne - Fundación Instituto Leloir, Instituto de Investigaciones Bioquímicas de Buenos Aires (IIBBA-CONICET), C1405 BWE Buenos Aires, Argentina
  • Author Contributions

    CRediT: Daiana A. Capdevila conceptualization, funding acquisition, visualization, writing - original draft, writing - review & editing;Johnma J. Rondón investigation, visualization, writing - original draft, writing - review & editing; Katherine A. Edmonds visualization, writing - original draft, writing - review & editing; Joseph S. Rocchio writing - review & editing; Matias Villarruel Dujovne visualization, writing - original draft, writing - review & editing; David P. Giedroc conceptualization, funding acquisition, visualization, writing - original draft, writing - review & editing.

  • Notes
    The authors declare no competing financial interest.

Biographies

Click to copy section linkSection link copied!
Download MS PowerPoint Slide

Daiana A. Capdevila earned her bachelor’s degree in Chemistry at the University of Buenos Aires, Argentina. In 2015, she obtained her Ph.D. in Chemistry in the same institution under the supervision of Prof. Daniel Murgida. She later earned a Latin American PEW fellowship for a postdoctoral position in the Giedroc group where she studied the role of conformational dynamics in metalloregulatory proteins. In 2019, she returned to Argentina to start her own research group at the Leloir Institute Foundation in Buenos Aires. She was appointed Professor of Chemistry at the University of San Martin in 2021.

Download MS PowerPoint Slide

Johnma J. Rondón Guerrero earned his bachelor’s degree in Biology at the Universidad de los Andes, Venezuela. He later obtained his Ph.D. in Cellular Biology at the same institution. He is currently working as a postdoctoral fellow at the Capdevila group studying the molecular evolution of the ArsR family of transcription factors in Mycobacterium tuberculosis.

Download MS PowerPoint Slide

Joseph S. Rocchio earned his undergraduate degree in Chemistry from Rensselaer Polytechnic Institute. He is currently a Ph.D. student in Chemistry at Indiana University in the Giedroc group, where he studies the biochemistry of the metal starvation response in Acinetobacter baumannii.

Download MS PowerPoint Slide

Katherine A. Edmonds earned her B.S. in Chemistry at the University of Chicago. She obtained her Ph.D. in Biophysics at Harvard University under the supervision of Prof. Gerhard Wagner, where she studied protein NMR and eukaryotic translation initiation. Dr. Edmonds is a research scientist in the Giedroc group interested in transition metal homeostasis.

Download MS PowerPoint Slide

Matias Villarruel Dujovne earned his bachelor’s degree in Chemistry at the University of Buenos Aires, Argentina. He is currently a Ph.D. student in Chemistry at the same institution while working in the Capdevila group, where he is combining the biophysical characterization of bacterial transcription factors with the development of cell-free biosensors.

Download MS PowerPoint Slide

David P. Giedroc earned a B.S. in Biochemistry from the Pennsylvania State University and Ph.D. in Biochemistry from the Vanderbilt University School of Medicine in 1984. Following a postdoctoral stint as an NIH postdoctoral fellow in the laboratory of the late Joseph Coleman at Yale University, he joined the faculty in the Department of Biochemistry and Biophysics at Texas A&M University in 1988. He was appointed Professor of Chemistry at Indiana University in 2007, served as Chair of Department from 2010-2015 and is currently holder of the Lilly Chemistry Alumni Chair. He was appointed Distinguished Professor of Chemistry in 2020. Prof. Giedroc is a Fellow of the AAAS, the American Academy of Microbiology, and the Royal Society of Chemistry.

Acknowledgments

Click to copy section linkSection link copied!

This work was supported by grants from the National Institutes of Health (R35 GM118157, R01 AI101171, and R01 AI178929) to D.P.G. and by Bunge & Born, Argentina, the Williams Foundations and MinCyT Argentina PICT-2021-GRF-TI-0415 and B81-CYTCH (to D.A.C.). D.A.C. is a staff member from CONICET; J.J.R. and M.V.D. are supported by a postdoctoral and a doctoral fellowship provided by CONICET, Argentina.

Abbreviations

Click to copy section linkSection link copied!
Mn

manganese

Zn

zinc

Fe

iron

Co

cobalt

Ni

nickel

Cu

copper

Se

selenium

Mo

molybdenum

Mg

magnesium

K

potassium

Pho

phosphate

Hg

mercury

Au

gold

Cd

cadmium

As

arsenic

Pb

lead

Fe–S

iron–sulfur

Cys

cysteine

His

histidine

Asp

aspartate

Glu

glutamate

Met

methionine

Pro

proline

Trp

tryptophan

Asn

asparagine

Gly

glycine

Ser

serine

Lys

lysine

RSSH

organic persulfide

CSSH

cysteine persulfide

GSSH

glutathione persulfide

H2S

hydrogen sulfide

ROS

reactive oxygen species

RNS

reactive nitrogen species

GSNO

S-Nitrosoglutathione

NOx

nitric oxide (NO) and nitrogen dioxide (NO2)

RES

reactive electrophile species

RSS

reactive sulfur species

Moco

molybdenum cofactor

GMPPNP

5′-Guanylyl imidodiphosphate

GMPPCP

Guanylyl 5′- (β,γ–methylenediphosphonate)

ppGpp

guanosine tetraphosphate

pppGpp

guanosine pentaphosphate

GTP

guanosine triphosphate

GDP

guanosine diphosphate

ATP

adenosine triphosphate

H2NTP

7,8-dihydroneopterin triphosphate

THF

tetrahydrofolate

Q

queuosine

CAA

carbamoyl aspartate

DHO

dihydroorotate

EDTA

Ethylenediaminetetraacetic acid

PPRPP

Phosphoribosyl diphosphate

AMA

aspergillomarasmine A

ZTP

5-amino 4-imidazole carboxamide riboside 5′-triphosphate

GTPγS

guanosine 5′-O-[γ-thio]triphosphate

HDX-MS

hydrogen–deuterium exchange mass spectrometry

AdoCbl

adenosylcobalamin

HMW

high molecular weight

ZF

zinc finger

NTD

N-terminal domain

CC

coiled-coil

HTH

helix-turn-helix DNA binding motif

TM

transmembrane domain

MBD

metal-binding domain

DBD

DNA-binding domain

MBL

metal binding loop

FCD

FadR C-terminal domain

DNA

DNA

RNA

ribonucleic acid

mRNA

mRNA

tRNA

tRNA

5′ UTR

5′ Untranslated Region

aTFs

allosteric transcription factors

MoTTs

modification-tunable transcripts

WCB

whole cell biosensor

LOD

limit of detection

SDA

strand displacement amplification

NAST

nicked DNA template-assisted signal transduction

Tx-Tl

transcription-translation

SELIS

Seamless Enrichment of Ligand-Inducible Sensors

EXAFS

Extended X-ray Absorption Fine Structure

XAS

X-ray absorption spectroscopy

EDS

energy-dispersive X-ray spectroscopy

Cryo-EM

cryogenic electron microscopy

SSN

sequence similarity network

NMR

nuclear magnetic resonance spectroscopy

SAXS

small-angle X-ray scattering

FRET

Förster resonance energy transfer

ICP-MS

inductively coupled plasma-mass spectrometry

2D-LC-ICP-MS

Two-dimensional liquid chromatography inductively coupled plasma mass spectrometry

GE-LS

polyacrylamide gel electrophoresis laser ablation

isoTOP-ABPP

isotopic tandem orthogonal proteolysis-activity-based protein profiling

Kd

dissociation equilibrium constant

pM

picomolar

nM

nanomolar

EC

Enzyme Commission

PDB

Protein Databank

CP

calprotectin

NMC

nucleotide-dependent metallochaperones

GAP

GTPase activating protein

ArsR

arsenic repressor

IdeR

Iron-dependent repressor

MerR

mercuric ion resistance regulator

CsoR

copper sensitive operon repressor

CstR

CsoR-like sulfurtransferase repressor

CopY

copper-responsive repressor

Fur

ferrous iron uptake repressor

Zur

Fur family zinc uptake regulator

DtxR

Diphtheria toxin repressor

NikR

Nickel responsive regulator of the nik operon

GntR

gluconate repressor

Mar

multiple antibiotic resistance repressor

TetR

Tetracycline Repressor

LysR

lysine repressor

DHOase

dihydroorotase

PyrC

pyrimidine biosynthetic enzyme dihydroorotase

QueD: 6-carboxy-5

6,7,8-tetrahydropterin synthase

DksA

RNA polymerase-binding transcription factor DksA/DnaK suppressor protein

FolE

GTP cyclohydrolase-IA (GCYH-1A)

CA

carbonic anhydrase

SODs

superoxide dismutases

camSOD

cambialistic superoxide dismutase

nsp

nonstructural protein

PTP1B

protein tyrosine phosphatase 1B

PBP2

penicillin-binding protein

MBLs

metallo-β-lactamases

SIMIBI

Signal recognition particle, MinD and BioD

ZigA

Zur-induced GTPase A

MCM

methylmalonyl-CoA mutase

MeaB

MCM accessory GTPase

UreG

urease accessory GTPase

NHase

nitrile hydratase

MPY

mycobacterial factor Y

Mrf

MPY recruitment factor

ZNG1

Zn-regulated GTPase metalloprotein activator

METAP1

methionine aminopeptidase-1

NAC

nascent polypeptide-associated complex

Csp

copper storage protein

MCO

multicopper oxidase

SERCA

sarco(endo)plasmic reticulum calcium ATPase

SPCA1

secretory-pathway CaII-ATPases

References

Click to copy section linkSection link copied!

This article references 969 other publications.

  1. 1
    Andreini, C.; Cavallaro, G.; Lorenzini, S.; Rosato, A. MetalPDB: A Database of Metal Sites in Biological Macromolecular Structures. Nucleic Acids Res. 2012, 41, D312– 9,  DOI: 10.1093/nar/gks1063
    Google Scholar
    There is no corresponding record for this reference.
  2. 2
    Andreini, C.; Bertini, I.; Cavallaro, G.; Holliday, G. L.; Thornton, J. M. Metal Ions in Biological Catalysis: From Enzyme Databases to General Principles. J. Biol. Inorg. Chem. 2008, 13, 1205– 1218,  DOI: 10.1007/s00775-008-0404-5
    Google Scholar
    2
    Metal ions in biological catalysis: from enzyme databases to general principles
    Andreini, Claudia; Bertini, Ivano; Cavallaro, Gabriele; Holliday, Gemma L.; Thornton, Janet M.
    JBIC, Journal of Biological Inorganic Chemistry (2008), 13 (8), 1205-1218CODEN: JJBCFA; ISSN:0949-8257. (Springer GmbH)
    The roles and distribution of metal ions in enzymic catalysis were analyzed using available public databases and the authors' new resource, Metal-MACiE (http://www.ebi.ac.uk/thornton-srv/databases/MetalMACiE/home.html). In Metal-MACiE, a database of metal-based reaction mechanisms, 116 entries covering 21% of the metal-dependent enzymes and 70% of the types of enzyme-catalyzed chem. transformations are annotated according to metal function. The authors used Metal-MACiE to assess the functions performed by metals in biol. catalysis and the relative frequencies of different metals in different roles, which could be related to their individual chem. properties and availability in the environment. The overall picture that emerged from the overview of Metal-MACiE was that redox-inert metal ions are used in enzymes to stabilize neg. charges and to activate substrates by virtue of their Lewis acid properties, whereas redox-active metal ions can be used both as Lewis acids and as redox centers. Mg2+ and Zn2+ were by far the most common ions of the 1st type, while Ca2+ was relatively less frequently used. Mg2+, however, was most often bound to phosphate groups of substrates and interacted with the enzyme only transiently, whereas the other metals were stably bound to the enzyme. The most common metal of the 2nd type was Fe2+/Fe3+, which was prevalent in the catalysis of redox reactions, followed by Mn2+, Co2+, Mo4+/Mo6+, Cu+/Cu2+, and Ni2+. The control of the reactivity of redox-active metal ions may involve their assocn. with org. cofactors to form stable units. This occurs sometimes for Fe2+/Fe3+ and Ni2+, and quite often for Co2+ and Mo4+/Mo6+.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXhtlGgtr%252FP&md5=8ce874b5110831f43821edb2f3e9d92c
  3. 3
    Waldron, K. J.; Rutherford, J. C.; Ford, D.; Robinson, N. J. Metalloproteins and Metal Sensing. Nature 2009, 460, 823– 30,  DOI: 10.1038/nature08300
    Google Scholar
    3
    Metalloproteins and metal sensing
    Waldron, Kevin J.; Rutherford, Julian C.; Ford, Dianne; Robinson, Nigel J.
    Nature (London, United Kingdom) (2009), 460 (7257), 823-830CODEN: NATUAS; ISSN:0028-0836. (Nature Publishing Group)
    A review. Almost half of all enzymes must assoc. with a particular metal to function. An aim of researchers is to understand why each metal-protein partnership arose and how it is maintained. Metal availability provides part of the explanation, and has changed over geol. time and varies between habitats but is held within vital limits in cells. Such homeostasis needs metal sensors, and there is an ongoing search to discover the metal-sensing mechanisms. For metalloproteins to acquire the right metals, metal sensors must correctly distinguish between the inorg. elements.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1MXpvFCruro%253D&md5=2680f7e2d354c7ce94f0f3b26a919c21
  4. 4
    Jordan, M. R.; Dujovne, M. V.; Capdevila, D. A.; Giedroc, D. P. Metal Ion Homeostasis: Metalloenzyme Paralogs in the Bacterial Adaptative Response to Zinc Restriction. In Comprehensive Inorganic Chemistry III; Elsevier, 2023; pp 30– 52.
    Google Scholar
    There is no corresponding record for this reference.
  5. 5
    Radin, J. N.; Kelliher, J. L.; Solórzano, P. K. P.; Grim, K. P.; Ramezanifard, R.; Slauch, J. M.; Kehl-Fie, T. E. Metal-Independent Variants of Phosphoglycerate Mutase Promote Resistance to Nutritional Immunity and Retention of Glycolysis during Infection. PLoS Pathog 2019, 15, e1007971  DOI: 10.1371/journal.ppat.1007971
    Google Scholar
    5
    Metal-independent variants of phosphoglycerate mutase promote resistance to nutritional immunity and retention of glycolysis during infection
    Radin, Jana N.; Kelliher, Jessica L.; Solorzano, Paola K. Parraga; Grim, Kyle P.; Ramezanifard, Rouhallah; Slauch, James M.; Kehl-Fie, Thomas E.
    PLoS Pathogens (2019), 15 (7), e1007971CODEN: PPLACN; ISSN:1553-7374. (Public Library of Science)
    The ability of Staphylococcus aureus and other pathogens to consume glucose is crit. during infection. However, glucose consumption increases the cellular demand for manganese sensitizing S. aureus to host-imposed manganese starvation. The current investigations were undertaken to elucidate how S. aureus copes with the need to consume glucose when metal-limited by the host. A crit. component of host defense is prodn. of the manganese binding protein calprotectin. S. aureus has two variants of phosphoglycerate mutase, one of which is manganese-dependent, GpmI, and another that is manganeseindependent, GpmA. Leveraging the ability to impose metal starvation in culture utilizing calprotectin revealed that the loss of GpmA, but not GpmI, sensitized S. aureus to manganese starvation. Metabolite feeding expts. revealed that the growth defect of GpmA when manganese-starved was due to a defect in glycolysis and not gluconeogenesis. Loss of GpmA reduces the ability of S. aureus to cause invasive disease in wild type mice. However, GpmA was dispensable in calprotectin-deficient mice, which have defects in manganese sequestration, indicating that this isoenzyme contributes to the ability of S. aureus to overcome manganese limitation during infection. Cumulatively, these observations suggest that expressing a metal-independent variant enables S. aureus to consume glucose while mitigating the neg. impact that glycolysis has on the cellular demand for manganese. S. aureus is not the only bacterium that expresses manganese-dependent and -independent variants of phosphoglycerate mutase. Similar results were also obsd. in culture with Salmonella enterica serovar Typhimurium mutants lacking the metal-independent isoenzyme. These similar observations in both Gram-pos. and Gram-neg. pathogens suggest that expression of metal-independent glycolytic isoenzymes is a common strategy employed by bacteria to survive in metal-limited environments, such as the host.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXitVCiur%252FM&md5=93cac0d5436518d70fadf274ce3035ca
  6. 6
    Parraga Solorzano, P. K.; Bastille, T. S.; Radin, J. N.; Kehl-Fie, T. E. A Manganese-Independent Aldolase Enables Staphylococcus aureus To Resist Host-Imposed Metal Starvation. mBio 2023, 14, e0322322  DOI: 10.1128/mbio.03223-22
    Google Scholar
    There is no corresponding record for this reference.
  7. 7
    Cotruvo, J. A., Jr.; Stubbe, J. Metallation and Mismetallation of Iron and Manganese Proteins in vitro and in vivo: The Class I Ribonucleotide Reductases as a Case Study. Metallomics 2012, 4, 1020– 1036,  DOI: 10.1039/c2mt20142a
    Google Scholar
    7
    Metallation and mismetallation of iron and manganese proteins in vitro and in vivo: the class I ribonucleotide reductases as a case study
    Cotruvo, Joseph A., Jr.; Stubbe, JoAnne
    Metallomics (2012), 4 (10), 1020-1036CODEN: METAJS; ISSN:1756-591X. (Royal Society of Chemistry)
    A review. How cells ensure correct metalation of a given protein and whether a degree of promiscuity in metal binding has evolved are largely unanswered questions. In a classic case, iron- and manganese-dependent superoxide dismutases (SODs) catalyze the disproportionation of superoxide using highly similar protein scaffolds and nearly identical active sites. However, most of these enzymes are active with only one metal, although both metals can bind in vitro and in vivo. Iron(ii) and manganese(ii) bind weakly to most proteins and possess similar coordination preferences. Their distinct redox properties suggest that they are unlikely to be interchangeable in biol. systems except when they function in Lewis acid catalytic roles, yet recent work suggests this is not always the case. This review summarizes the diversity of ways in which iron and manganese are substituted in similar or identical protein frameworks. As models, we discuss (1) enzymes, such as epimerases, thought to use FeII as a Lewis acid under normal growth conditions but which switch to MnII under oxidative stress; (2) extradiol dioxygenases, which have been found to use both FeII and MnII, the redox role of which in catalysis remains to be elucidated; (3) SODs, which use redox chem. and are generally metal-specific; and (4) the class I ribonucleotide reductases (RNRs), which have evolved unique biosynthetic pathways to control metalation. The primary focus is the class Ib RNRs, which can catalyze formation of a stable radical on a tyrosine residue in their β2 subunits using either a di-iron or a recently characterized dimanganese cofactor. The physiol. roles of enzymes that can switch between iron and manganese cofactors are discussed, as are insights obtained from the studies of many groups regarding iron and manganese homeostasis and the divergent and convergent strategies organisms use for control of protein metalation. We propose that, in many of the systems discussed, "discrimination" between metals is not performed by the protein itself, but it is instead detd. by the environment in which the protein is expressed.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38XhsVWgtrrJ&md5=8f407c40ec2e066607971c2965802902
  8. 8
    Ruskoski, T. B.; Boal, A. K. The Periodic Table of Ribonucleotide Reductases. J. Biol. Chem. 2021, 297, 101137,  DOI: 10.1016/j.jbc.2021.101137
    Google Scholar
    There is no corresponding record for this reference.
  9. 9
    Jordan, M. R.; Gonzalez-Gutierrez, G.; Trinidad, J. C.; Giedroc, D. P. Metal Retention and Replacement in QueD2 Protect Queuosine-tRNA Biosynthesis in Metal-Starved Acinetobacter baumannii. Proc. Natl. Acad. Sci. U. S. A. 2022, 119, e2213630119  DOI: 10.1073/pnas.2213630119
    Google Scholar
    There is no corresponding record for this reference.
  10. 10
    Antelo, G. T.; Vila, A. J.; Giedroc, D. P.; Capdevila, D. A. Molecular Evolution of Transition Metal Bioavailability at the Host-Pathogen Interface. Trends Microbiol 2021, 29, 441– 457,  DOI: 10.1016/j.tim.2020.08.001
    Google Scholar
    10
    Molecular Evolution of Transition Metal Bioavailability at the Host-Pathogen Interface
    Antelo, Giuliano T.; Vila, Alejandro J.; Giedroc, David P.; Capdevila, Daiana A.
    Trends in Microbiology (2021), 29 (5), 441-457CODEN: TRMIEA; ISSN:0966-842X. (Elsevier Ltd.)
    A review mol. evolution of the adaptive response at the host-pathogen interface has been frequently referred to as an 'arms race' between the host and bacterial pathogens. The innate immune system employs multiple strategies to starve microbes of metals. Pathogens, in turn, develop successful strategies to maintain access to bioavailable metal ions under conditions of extreme restriction of transition metals, or nutritional immunity. However, the processes by which evolution repurposes or re-engineers host and pathogen proteins to perform or refine new functions have been explored only recently. Here we review the mol. evolution of several human metalloproteins charged with restricting bacterial access to transition metals. These include the transition metal-chelating S100 proteins, natural resistance-assocd. macrophage protein-1 (NRAMP-1), transferrin, lactoferrin, and heme-binding proteins. We examine their coevolution with bacterial transition metal acquisition systems, involving siderophores and membrane-spanning metal importers, and the biol. specificity of allosteric transcriptional regulatory proteins tasked with maintaining bacterial metallostasis. We also discuss the evolution of metallo-β-lactamases; this illustrates how rapid antibiotic-mediated evolution of a zinc metalloenzyme obligatorily occurs in the context of host-imposed nutritional immunity.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXhslejtr3P&md5=47198f603b4e3d2bbe42d6fdc9da56a3
  11. 11
    Zygiel, E. M.; Nolan, E. M. Transition Metal Sequestration by the Host-Defense Protein Calprotectin. Annu. Rev. Biochem. 2018, 87, 621– 643,  DOI: 10.1146/annurev-biochem-062917-012312
    Google Scholar
    11
    Transition Metal Sequestration by the Host-Defense Protein Calprotectin
    Zygiel, Emily M.; Nolan, Elizabeth M.
    Annual Review of Biochemistry (2018), 87 (), 621-643CODEN: ARBOAW; ISSN:0066-4154. (Annual Reviews)
    A review. In response to microbial infection, the human host deploys metal-sequestering host-defense proteins, which reduce nutrient availability and thereby inhibit microbial growth and virulence. Calprotectin (CP) is an abundant antimicrobial protein released from neutrophils and epithelial cells at sites of infection. CP sequesters divalent first-row transition metal ions to limit the availability of essential metal nutrients in the extracellular space. While functional and clin. studies of CP have been pursued for decades, advances in our understanding of its biol. coordination chem., which is central to its role in the host-microbe interaction, have been made in more recent years. In this review, we focus on the coordination chem. of CP and highlight studies of its metal-binding properties and contributions to the metal-withholding innate immune response. Taken together, these recent studies inform our current model of how CP participates in metal homeostasis and immunity, and they provide a foundation for further investigations of a remarkable metal-chelating protein at the host-microbe interface and beyond.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXhtFOrsrvN&md5=7782548a1ccbde6ebabddb3e231a7d52
  12. 12
    Sheldon, J. R.; Skaar, E. P. Metals as Phagocyte Antimicrobial Effectors. Curr. Opin. Immunol 2019, 60, 1– 9,  DOI: 10.1016/j.coi.2019.04.002
    Google Scholar
    12
    Metals as phagocyte antimicrobial effectors
    Sheldon, Jessica R.; Skaar, Eric P.
    Current Opinion in Immunology (2019), 60 (), 1-9CODEN: COPIEL; ISSN:0952-7915. (Elsevier Ltd.)
    A review. Transition metal ions are essential to bacterial pathogens and their hosts alike but are harmful in excess. In an effort to curtail the replication of intracellular bacteria, host phagocytes exploit both the essentiality and toxicity of transition metals. In the paradigmatic description of nutritional immunity, iron and manganese are withheld from phagosomes to starve microbial invaders of these nutrients. Conversely, the destructive properties of copper and zinc appear to be harnessed by phagocytes, where these metals are delivered in excess to phagosomes to intoxicate internalized bacteria. Here, we briefly summarize key players in metal withholding from intracellular pathogens, before focusing on recent findings supporting the function of copper and zinc as phagocyte antimicrobial effectors. The mechanisms of copper and zinc toxicity are explored, along with strategies employed by intracellular bacterial pathogens to avoid killing by these metals.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXos12ksbg%253D&md5=3c27bb23bc8ee289ebf152f0abd7eff8
  13. 13
    Liu, C.; Jursa, T.; Aschner, M.; Smith, D. R.; Mukhopadhyay, S. Up-Regulation of the Manganese Transporter SLC30A10 by Hypoxia-Inducible Factors Defines a Homeostatic Response to Manganese Toxicity. Proc. Natl. Acad. Sci. U. S. A. 2021, 118, 1– 12,  DOI: 10.1073/pnas.2107673118
    Google Scholar
    There is no corresponding record for this reference.
  14. 14
    Banerjee, R.; Gouda, H.; Pillay, S. Redox-Linked Coordination Chemistry Directs Vitamin B12 Trafficking. Acc. Chem. Res. 2021, 54, 2003– 2013,  DOI: 10.1021/acs.accounts.1c00083
    Google Scholar
    There is no corresponding record for this reference.
  15. 15
    Bagheri, S.; Squitti, R.; Haertlé, T.; Siotto, M.; Saboury, A. A. Role of Copper in the Onset of Alzheimer’s Disease Compared to Other Metals. Front Aging Neurosci 2018, 9, 1– 15,  DOI: 10.3389/fnagi.2017.00446
    Google Scholar
    There is no corresponding record for this reference.
  16. 16
    de Bie, P.; Muller, P.; Wijmenga, C.; Klomp, L. W. J. Molecular Pathogenesis of Wilson and Menkes Disease: Correlation of Mutations with Molecular Defects and Disease Phenotypes. J. Med. Genet 2007, 44, 673– 688,  DOI: 10.1136/jmg.2007.052746
    Google Scholar
    16
    Molecular pathogenesis of Wilson and Menkes diseases: correlation of mutations with molecular defects and disease phenotypes
    de Bie, P.; Muller, P.; Wijmenga, C.; Klomp, L. W. J.
    Journal of Medical Genetics (2007), 44 (11), 673-688CODEN: JMDGAE; ISSN:0022-2593. (BMJ Publishing Group)
    A review. The trace metal copper is essential for a variety of biol. processes, but extremely toxic when present in excessive amts. Therefore, concns. of this metal in the body are kept under tight control. Central regulators of cellular copper metab. are the copper-transporting P-type ATPases ATP7A and ATP7B. Mutations in ATP7A or ATP7B disrupt the homeostatic copper balance, resulting in copper deficiency (Menkes disease) or copper overload (Wilson disease), resp. ATP7A and ATP7B exert their functions in copper transport through a variety of interdependent mechanisms and regulatory events, including their catalytic ATPase activity, copper-induced trafficking, post-translational modifications and protein-protein interactions. This paper reviews the extensive efforts that have been undertaken over the past few years to dissect and characterize these mechanisms, and how these are affected in Menkes and Wilson disease. As both disorders are characterized by an extensive clin. heterogeneity, we will discus how the underlying genetic defects correlate with the mol. functions of ATP7A and ATP7B and with the clin. expression of these disorders.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXhsVGqurzM&md5=19583f444207eceee75afb77899180f6
  17. 17
    Rodriguez-Granillo, A.; Crespo, A.; Wittung-Stafshede, P. Conformational Dynamics of Metal-Binding Domains in Wilson Disease Protein: Molecular Insights into Selective Copper Transfer. Biochemistry 2009, 48, 5849– 5863,  DOI: 10.1021/bi900235g
    Google Scholar
    17
    Conformational Dynamics of Metal-Binding Domains in Wilson Disease Protein: Molecular Insights into Selective Copper Transfer
    Rodriguez-Granillo, Agustina; Crespo, Alejandro; Wittung-Stafshede, Pernilla
    Biochemistry (2009), 48 (25), 5849-5863CODEN: BICHAW; ISSN:0006-2960. (American Chemical Society)
    ATP7A/B are human P1B-type ATPases involved in cellular Cu homeostasis. The N-terminal parts of these multidomain proteins contain six metal-binding domains (MBDs) connected by linkers. The MBDs are similar in structure to each other and to the human copper chaperone Atox1, although their distinct roles in Cu transfer appear to vary. All domains have the ferredoxin-like fold and a solvent-exposed loop with a MXCXXC motif that can bind CuI. Here, we investigated the dynamic behavior of the individual MBDs (WD1-WD6) in ATP7B in apo forms using mol. dynamic simulations. We also performed simulations of three Cu-bound forms (WD2c, WD4c, and WD6c). Our results reveal mol. features that vary distinctly among the MBDs. Whereas WD1, WD2, and WD6 have well-defined Cu loop conformations stabilized by a network of interactions, WD4 and WD5 exhibit greater loop flexibility and, in WD4, helix α1 unwinds and rewinds. WD3, which has the lowest sequence identity, behaves differently and its Cu loop is rigid with respect to the rest of the domain. Cu coordination reduces structural dynamics in all domains but WD4c. In agreement with predictions on individual domains, simulations of the six possible Atox1-WD heterocomplexes show that Atox1 interactions with WD4 are the strongest. This study provides mol. explanations for reported Cu transfer and protein-protein interaction specificity.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1MXmvVelurk%253D&md5=c0996d0f18e119940ba716bbcbf05992
  18. 18
    Giedroc, D. P.; Arunkumar, A. I. Metal Sensor Proteins: Nature’s Metalloregulated Allosteric Switches. Dalton Trans 2007, 3107– 3120,  DOI: 10.1039/b706769k
    Google Scholar
    18
    Metal sensor proteins: Nature's metalloregulated allosteric switches
    Giedroc, David P.; Arunkumar, Alphonse I.
    Dalton Transactions (2007), (29), 3107-3120CODEN: DTARAF; ISSN:1477-9226. (Royal Society of Chemistry)
    A review. Metalloregulatory proteins control the expression of genes that allow organisms to quickly adapt to chronic toxicity or deprivation of both biol. essential metal ions and heavy metal pollutants found in their microenvironment. Emerging evidence suggests that metal ion homeostasis and resistance defines an important tug-of-war in human host-bacterial pathogen interactions. This adaptive response originates with the formation of "metal receptor" complexes of exquisite selectivity. In this perspective, we summarize consensus structural features of metal sensing coordination complexes and the evolution of distinct metal selectivities within seven characterized metal sensor protein families. In addn., we place recent efforts to understand the structural basis of metal-induced allosteric switching of these metalloregulatory proteins in a thermodn. framework, and review the degree to which coordination chem. drives changes in protein structure and dynamics in selected metal sensor systems. New insights into how metal sensor proteins function in the complex intracellular milieu of the cytoplasm of cells will require a more sophisticated understanding of the "metallome" and will benefit greatly from ongoing collaborative efforts in bioinorg., biophys. and anal. chem., structural biol. and microbiol.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXnslKktr4%253D&md5=30fe0abf8606fa713be41dfeaa7a38bd
  19. 19
    Jung, J. K.; Alam, K. K.; Verosloff, M. S.; Capdevila, D. A.; Desmau, M.; Clauer, P. R.; Lee, J. W.; Nguyen, P. Q.; Pastén, P. A.; Matiasek, S. J. Cell-Free Biosensors for Rapid Detection of Water Contaminants. Nat. Biotechnol. 2020, 38, 1451– 1459,  DOI: 10.1038/s41587-020-0571-7
    Google Scholar
    19
    Cell-free biosensors for rapid detection of water contaminants
    Jung, Jaeyoung K.; Alam, Khalid K.; Verosloff, Matthew S.; Capdevila, Daiana A.; Desmau, Morgane; Clauer, Phillip R.; Lee, Jeong Wook; Nguyen, Peter Q.; Pasten, Pablo A.; Matiasek, Sandrine J.; Gaillard, Jean-Francois; Giedroc, David P.; Collins, James J.; Lucks, Julius B.
    Nature Biotechnology (2020), 38 (12), 1451-1459CODEN: NABIF9; ISSN:1087-0156. (Nature Research)
    Abstr.: Lack of access to safe drinking water is a global problem, and methods to reliably and easily detect contaminants could be transformative. We report the development of a cell-free in vitro transcription system that uses RNA Output Sensors Activated by Ligand Induction (ROSALIND) to detect contaminants in water. A combination of highly processive RNA polymerases, allosteric protein transcription factors and synthetic DNA transcription templates regulates the synthesis of a fluorescence-activating RNA aptamer. The presence of a target contaminant induces the transcription of the aptamer, and a fluorescent signal is produced. We apply ROSALIND to detect a range of water contaminants, including antibiotics, small mols. and metals. We also show that adding RNA circuitry can invert responses, reduce crosstalk and improve sensitivity without protein engineering. The ROSALIND system can be freeze-dried for easy storage and distribution, and we apply it in the field to test municipal water supplies, demonstrating its potential use for monitoring water quality.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXhtlagtLrK&md5=1db92d8b43f6619693da1cdddde3738a
  20. 20
    Zhang, Y.; Zheng, J. Bioinformatics of Metalloproteins and Metalloproteomes. Molecules 2020, 25, 3366,  DOI: 10.3390/molecules25153366
    Google Scholar
    20
    Bioinformatics of Metalloproteins and Metalloproteomes
    Zhang, Yan; Zheng, Junge
    Molecules (2020), 25 (15), 3366CODEN: MOLEFW; ISSN:1420-3049. (MDPI AG)
    A review. Trace metals are inorg. elements that are required for all organisms in very low quantities. They serve as cofactors and activators of metalloproteins involved in a variety of key cellular processes. While substantial effort has been made in exptl. characterization of metalloproteins and their functions, the application of bioinformatics in the research of metalloproteins and metalloproteomes is still limited. In the last few years, computational prediction and comparative genomics of metalloprotein genes have arisen, which provide significant insights into their distribution, function, and evolution in nature. This review aims to offer an overview of recent advances in bioinformatic anal. of metalloproteins, mainly focusing on metalloprotein prediction and the use of different metals across the tree of life. We describe current computational approaches for the identification of metalloprotein genes and metal-binding sites/patterns in proteins, and then introduce a set of related databases. Furthermore, we discuss the latest research progress in comparative genomics of several important metals in both prokaryotes and eukaryotes, which demonstrates divergent and dynamic evolutionary patterns of different metalloprotein families and metalloproteomes. Overall, bioinformatic studies of metalloproteins provide a foundation for systematic understanding of trace metal utilization in all three domains of life.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3cXhsFKjtrnP&md5=ec070065b71b42093dcf48ec9690208b
  21. 21
    Sigel, H.; Sigel, A. The Bio-Relevant Metals of the Periodic Table of the Elements. Zeitschrift fur Naturforschung - Sec B J. Chem. Sci. 2019, 74, 461– 471,  DOI: 10.1515/znb-2019-0056
    Google Scholar
    There is no corresponding record for this reference.
  22. 22
    Freisinger, E.; Sigel, R. K. O. The Bioinorganic Periodic Table. Chimia (Aarau) 2019, 73, 185– 193,  DOI: 10.2533/chimia.2019.185
    Google Scholar
    There is no corresponding record for this reference.
  23. 23
    Probst, C.; Yang, J.; Krausze, J.; Hercher, T. W.; Richers, C. P.; Spatzal, T.; Kc, K.; Giles, L. J.; Rees, D. C. Mechanism of Molybdate Insertion into Pterin-Based Molybdenum Cofactors. Nat. Chem. 2021, 13, 758– 765,  DOI: 10.1038/s41557-021-00714-1
    Google Scholar
    23
    Mechanism of molybdate insertion into pterin-based molybdenum cofactors
    Probst, Corinna; Yang, Jing; Krausze, Joern; Hercher, Thomas W.; Richers, Casseday P.; Spatzal, Thomas; Kc, Khadanand; Giles, Logan J.; Rees, Douglas C.; Mendel, Ralf R.; Kirk, Martin L.; Kruse, Tobias
    Nature Chemistry (2021), 13 (8), 758-765CODEN: NCAHBB; ISSN:1755-4330. (Nature Portfolio)
    The molybdenum cofactor (Moco) is found in the active site of numerous important enzymes that are crit. to biol. processes. The bidentate ligand that chelates molybdenum in Moco is the pyranopterin dithiolene (molybdopterin, MPT). However, neither the mechanism of molybdate insertion into MPT nor the structure of Moco prior to its insertion into pyranopterin molybdenum enzymes is known. Here, the authors report this final maturation step, where adenylated MPT (MPT-AMP) and molybdate are the substrates. X-ray crystallog. of the Arabidopsis thaliana Mo-insertase variant Cnx1E S269D D274S identified adenylated Moco (Moco-AMP) as an unexpected intermediate in this reaction sequence. X-ray absorption spectroscopy revealed the first coordination sphere geometry of Moco trapped in the Cnx1E active site. The authors used this structural information to deduce a mechanism for molybdate insertion into MPT-AMP. Given their high degree of structural and sequence similarity, the authors suggest that this mechanism was employed by all eukaryotic Mo-insertases.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3MXhsVCltrzO&md5=d636c78885689b5aa05320ad6b220b0a
  24. 24
    Zhang, Y.; Gladyshev, V. N. Molybdoproteomes and Evolution of Molybdenum Utilization. J. Mol. Biol. 2008, 379, 881– 899,  DOI: 10.1016/j.jmb.2008.03.051
    Google Scholar
    24
    Molybdoproteomes and Evolution of Molybdenum Utilization
    Zhang, Yan; Gladyshev, Vadim N.
    Journal of Molecular Biology (2008), 379 (4), 881-899CODEN: JMOBAK; ISSN:0022-2836. (Elsevier Ltd.)
    The trace element molybdenum (Mo) is utilized in many life forms, and it is a key component of several enzymes involved in nitrogen, sulfur, and carbon metab. With the exception of nitrogenase, Mo is bound in proteins to a pterin, thus forming the molybdenum cofactor (Moco) at the catalytic sites of molybdoenzymes. Although a no. of molybdoenzymes are well characterized structurally and functionally, evolutionary analyses of Mo utilization are limited. Here, we carried out comparative genomic and phylogenetic analyses to examine the occurrence and evolution of Mo utilization in bacteria, archaea and eukaryotes at the level of (i) Mo transport and Moco utilization trait, and (ii) Mo-dependent enzymes. Our results revealed that most prokaryotes and all higher eukaryotes utilize Mo, whereas many unicellular eukaryotes including parasites and most yeasts lost the ability to use this metal. In addn., eukaryotes have fewer molybdoenzyme families than prokaryotes. Dimethylsulfoxide reductase (DMSOR) and sulfite oxidase (SO) families were the most widespread molybdoenzymes in prokaryotes and eukaryotes, resp. A distant group of the ModABC transport system, was predicted in the hyperthermophilic archaeon Pyrobaculum. ModE-type regulation of Mo uptake occurred in less than 30% of Moco-utilizing organisms. A link between Mo and selenocysteine utilization in prokaryotes was also identified wherein the selenocysteine trait was largely a subset of the Mo trait, presumably due to formate dehydrogenase, a Mo- and selenium-contg. protein. Finally, anal. of environmental conditions and organisms that do or do not depend on Mo revealed that host-assocd. organisms and organisms with low G + C content tend to reduce their Mo utilization. Overall, our data provide new insights into Mo utilization and show its wide occurrence, yet limited use of this metal in individual organisms in all three domains of life.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD1cXmsFKqsLk%253D&md5=e074cdff825ebc86754907146f706df1
  25. 25
    Zupok, A.; Iobbi-Nivol, C.; Mejean, V.; Leimkühler, S. The Regulation of Moco Biosynthesis and Molybdoenzyme Gene Expression by Molybdenum and Iron in Bacteria. Metallomics 2019, 11, 1602– 1624,  DOI: 10.1039/c9mt00186g
    Google Scholar
    There is no corresponding record for this reference.
  26. 26
    Kang, W. Structural Insights and Mechanistic Understanding of Iron-Molybdenum Cofactor Biosynthesis by NifB in Nitrogenase Assembly Process. Mol. Cells 2023, 46, 736– 742,  DOI: 10.14348/molcells.2023.0140
    Google Scholar
    There is no corresponding record for this reference.
  27. 27
    Weber, J. N.; Minner-Meinen, R.; Behnecke, M.; Biedendieck, R.; Hansch, V. G.; Hercher, T. W.; Hertweck, C.; van den Hout, L.; Knuppel, L.; Sivov, S. Moonlighting Arabidopsis Molybdate Transporter 2 Family and GSH-Complex Formation Facilitate Molybdenum Homeostasis. Commun. Biol. 2023, 6, 801,  DOI: 10.1038/s42003-023-05161-x
    Google Scholar
    There is no corresponding record for this reference.
  28. 28
    Tejada-Jimenez, M.; Leon-Miranda, E.; Llamas, A. Chlamydomonas Reinhardtii-A Reference Microorganism for Eukaryotic Molybdenum Metabolism. Microorganisms 2023, 11, 1671,  DOI: 10.3390/microorganisms11071671
    Google Scholar
    There is no corresponding record for this reference.
  29. 29
    Johannes, L.; Fu, C. Y.; Schwarz, G. Molybdenum Cofactor Deficiency in Humans. Molecules 2022, 27, 6896,  DOI: 10.3390/molecules27206896
    Google Scholar
    There is no corresponding record for this reference.
  30. 30
    Hausinger, R. P. New Metal Cofactors and Recent Metallocofactor Insights. Curr. Opin Struct Biol. 2019, 59, 1– 8,  DOI: 10.1016/j.sbi.2018.12.008
    Google Scholar
    30
    New metal cofactors and recent metallocofactor insights
    Hausinger, Robert P.
    Current Opinion in Structural Biology (2019), 59 (), 1-8CODEN: COSBEF; ISSN:0959-440X. (Elsevier Ltd.)
    A review. A vast array of metal cofactors are assocd. with the active sites of metalloenzymes. This Opinion describes the most recently discovered metal cofactor, a nickel-pincer nucleotide (NPN) coenzyme that is covalently tethered to lactate racemase from Lactobacillus plantarum. The enzymic function of the NPN cofactor and its pathway for biosynthesis are reviewed. Furthermore, insights are summarized from recent advances involving other selected organometallic and inorg.-cluster cofactors including the lanthanide-pyrroloquinoline quinone found in certain alc. dehydrogenases, tungsten-pyranopterins or molybdenum-pyranopterins in chosen enzymes, the iron-guanylylpyridinol cofactor of [Fe] hydrogenase, the nickel-tetrapyrrole coenzyme F430 of Me coenzyme M reductase, the vanadium-iron cofactor of nitrogenase, redox-dependent rearrangements of the nickel-iron-sulfur C-cluster in carbon monoxide dehydrogenase, and light-dependent changes in the multi-manganese cluster of the oxygen-evolving complex.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXhsVOqtL4%253D&md5=22a40e4f304eec5a4172b92fffb4c552
  31. 31
    Daumann, L. J.; Pol, A.; Op den Camp, H. J. M.; Martinez-Gomez, N. C. A Perspective on the Role of Lanthanides in Biology: Discovery, Open Questions and Possible Applications. Adv. Microb Physiol 2022, 81, 1– 24,  DOI: 10.1016/bs.ampbs.2022.06.001
    Google Scholar
    There is no corresponding record for this reference.
  32. 32
    Featherston, E. R.; Cotruvo, J. A., Jr The Biochemistry of Lanthanide Acquisition, Trafficking, and Utilization. Biochim Biophys Acta Mol. Cell Res. 2021, 1868, 118864,  DOI: 10.1016/j.bbamcr.2020.118864
    Google Scholar
    There is no corresponding record for this reference.
  33. 33
    Mattocks, J. A.; Jung, J. J.; Lin, C. Y.; Dong, Z.; Yennawar, N. H.; Featherston, E. R.; Kang-Yun, C. S.; Hamilton, T. A.; Park, D. M.; Boal, A. K. Enhanced Rare-Earth Separation with a Metal-Sensitive Lanmodulin Dimer. Nature 2023, 618, 87– 93,  DOI: 10.1038/s41586-023-05945-5
    Google Scholar
    33
    Enhanced rare-earth separation with a metal-sensitive lanmodulin dimer
    Mattocks, Joseph A.; Jung, Jonathan J.; Lin, Chi-Yun; Dong, Ziye; Yennawar, Neela H.; Featherston, Emily R.; Kang-Yun, Christina S.; Hamilton, Timothy A.; Park, Dan M.; Boal, Amie K.; Cotruvo Jr, Joseph A.
    Nature (London, United Kingdom) (2023), 618 (7963), 87-93CODEN: NATUAS; ISSN:1476-4687. (Nature Portfolio)
    Technol. crit. rare-earth elements are notoriously difficult to sep., owing to their subtle differences in ionic radius and coordination no.1-3. The natural lanthanide-binding protein lanmodulin (LanM)4,5 is a sustainable alternative to conventional solvent-extn.-based sepn.6. Here we characterize a new LanM, from Hansschlegelia quercus (Hans-LanM), with an oligomeric state sensitive to rare-earth ionic radius, the lanthanum(III)-induced dimer being >100-fold tighter than the dysprosium(III)-induced dimer. X-ray crystal structures illustrate how picometre-scale differences in radius between lanthanum(III) and dysprosium(III) are propagated to Hans-LanMs quaternary structure through a carboxylate shift that rearranges a second-sphere hydrogen-bonding network. Comparison to the prototypal LanM from Methylorubrum extorquens reveals distinct metal coordination strategies, rationalizing Hans-LanMs greater selectivity within the rare-earth elements. Finally, structure-guided mutagenesis of a key residue at the Hans-LanM dimer interface modulates dimerization in soln. and enables single-stage, column-based sepn. of a neodymium(III)/dysprosium(III) mixt. to >98% individual element purities. This work showcases the natural diversity of selective lanthanide recognition motifs, and it reveals rare-earth-sensitive dimerization as a biol. principle by which to tune the performance of biomol.-based sepn. processes.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3sXhtFaqtrfF&md5=b957ebd84259b6e125217debdb42a537
  34. 34
    Zoidis, E.; Seremelis, I.; Kontopoulos, N.; Danezis, G. P. Selenium-Dependent Antioxidant Enzymes: Actions and Properties of Selenoproteins. Antioxidants 2018, 7, e00624-19  DOI: 10.3390/antiox7050066
    Google Scholar
    There is no corresponding record for this reference.
  35. 35
    Santesmasses, D.; Mariotti, M.; Gladyshev, V. N. Bioinformatics of Selenoproteins. Antioxid Redox Signal 2020, 33, 525– 536,  DOI: 10.1089/ars.2020.8044
    Google Scholar
    There is no corresponding record for this reference.
  36. 36
    Andreini, C.; Bertini, I. A Bioinformatics View of Zinc Enzymes. J. Inorg. Biochem 2012, 111, 150– 156,  DOI: 10.1016/j.jinorgbio.2011.11.020
    Google Scholar
    36
    A bioinformatics view of zinc enzymes
    Andreini, Claudia; Bertini, Ivano
    Journal of Inorganic Biochemistry (2012), 111 (), 150-156CODEN: JIBIDJ; ISSN:0162-0134. (Elsevier)
    Thanks to the contributions of scientists like Bert L. Vallee, Zn enzymol. is an area of research with a rich history and a strong basis of biochem. and biophys. knowledge. In recent years, the dramatic development of genomic and post-genomic research has provided this as well as all other fields of life sciences with a massive body of new data, including, but not limited to, protein sequence and structural data. By integrating these new data with the wealth of information available in the literature, it is possible to achieve an unprecedented overview of the properties and functions of Zn-contg. enzymes in the context of biol. systems. To this aim, the role of bioinformatics is essential. Here, the authors used bioinformatics tools and databases that they developed for the study of metalloproteins to gain insights into the functions of Zn in Zn-contg. enzymes, its coordination properties, and the usage of Zn-contg. enzymes in living organisms.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38XosVWksrw%253D&md5=e3b25937a50ef6e1bb3297fff55e5c83
  37. 37
    Choi, S.; Bird, A. J. Zinc’ing Sensibly: Controlling Zinc Homeostasis at the Transcriptional Level. Metallomics 2014, 6, 1198– 1215,  DOI: 10.1039/C4MT00064A
    Google Scholar
    37
    Zinc'ing sensibly: controlling zinc homeostasis at the transcriptional level
    Choi, Sangyong; Bird, Amanda J.
    Metallomics (2014), 6 (7), 1198-1215CODEN: METAJS; ISSN:1756-591X. (Royal Society of Chemistry)
    A review. Zinc-responsive transcription factors are found in all kingdoms of life and include the transcriptional activators ZntR, SczA, Zap1, bZip19, bZip23, and MTF-1, and transcriptional repressors Zur, AdcR, Loz1, and SmtB. These factors have two defining features; their activity is regulated by zinc and they all play a central role in zinc homeostasis by controlling the expression of genes that directly affect zinc levels or its availability. This review summarizes what is known about the mechanisms by which each of these factors sense changes in intracellular zinc levels and how they control zinc homeostasis through target gene regulation. Other factors that influence zinc ion sensing are also discussed.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhtVKjsLzM&md5=da417ae101c50d7ec062f87e75318455
  38. 38
    Damon, L. J.; Ocampo, D.; Sanford, L.; Taylor, J.; Allen, M. A.; Dowell, R. D.; Palmer, A. E. Cellular Zinc Status Alters Chromatin Accessibility and Binding of Transcription Factor p53 to Genomic Sites. bioRxiv 2023. DOI: 10.1101/2023.11.20.567954 .
    Google Scholar
    There is no corresponding record for this reference.
  39. 39
    Laity, J. H.; Andrews, G. K. Understanding the Mechanisms of Zinc-Sensing by Metal-Response Element Binding Transcription Factor-1 (MTF-1). Arch. Biochem. Biophys. 2007, 463, 201– 210,  DOI: 10.1016/j.abb.2007.03.019
    Google Scholar
    39
    Understanding the mechanisms of zinc-sensing by metal-response element binding transcription factor-1 (MTF-1)
    Laity, John H.; Andrews, Glen K.
    Archives of Biochemistry and Biophysics (2007), 463 (2), 201-210CODEN: ABBIA4; ISSN:0003-9861. (Elsevier)
    A review. The regulation of Zn2+ has been obsd. in a wide range of organisms. Since this metal is an essential nutrient, but also toxic in excess, Zn2+ homeostasis is crucial for normal cellular functioning. Metal-responsive-element-binding transcription factor-1 (MTF-1) is a key regulator of Zn2+ in higher eukaryotes ranging from insects to mammals. MTF-1 controls the expression of metallothioneins (MTs) and a no. of other genes directly involved in the intracellular sequestration and transport of Zn2+. Although the diverse functions of MTF-1 extend well beyond Zn2+ homeostasis to include stress-responses to heavy metal toxicity, oxidative stress, and selected chem. agents, in this review the authors focus on recent advances in understanding the mechanisms whereby MTF-1 regulates MT gene expression to protect the cell from fluctuations in environmental Zn2+. Particular emphasis is devoted to recent studies involving the Cys2His2 zinc finger DNA-binding domain of MTF-1, which is an important contributor to the zinc-sensing and metal-dependent transcriptional activation functions of this protein.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BD2sXnslWhurs%253D&md5=b3b0e308a3d164916addf0f05379640e
  40. 40
    Wang, Z.; Feng, L. S.; Matskevich, V.; Venkataraman, K.; Parasuram, P.; Laity, J. H. Solution Structure of a Zap1 Zinc-Responsive Domain Provides Insights into Metalloregulatory Transcriptional Repression in Saccharomyces cerevisiae. J. Mol. Biol. 2006, 357, 1167– 1183,  DOI: 10.1016/j.jmb.2006.01.010
    Google Scholar
    There is no corresponding record for this reference.
  41. 41
    Lange, M.; Ok, K.; Shimberg, G. D.; Bursac, B.; Marko, L.; Ivanovic-Burmazovic, I.; Michel, S. L. J.; Filipovic, M. R. Direct Zinc Finger Protein Persulfidation by H2S Is Facilitated by Zn2+. Angew. Chem., Int. Ed. Engl. 2019, 58, 7997– 8001,  DOI: 10.1002/anie.201900823
    Google Scholar
    There is no corresponding record for this reference.
  42. 42
    Cui, B.; Pan, Q.; Clarke, D.; Villarreal, M. O.; Umbreen, S.; Yuan, B.; Shan, W.; Jiang, J.; Loake, G. J. S-Nitrosylation of the Zinc Finger Protein SRG1 Regulates Plant Immunity. Nat. Commun. 2018, 9, 4226,  DOI: 10.1038/s41467-018-06578-3
    Google Scholar
    There is no corresponding record for this reference.
  43. 43
    Mely, Y.; Rocquigny, H.; Shvadchak, V.; Avilov, S.; Dong, C.; Dietrich, U.; Darlix, J.-L. Targeting the Viral Nucleocapsid Protein in Anti-HIV-1 Therapy. Mini Rev. Med. Chem. 2008, 8, 24– 35,  DOI: 10.2174/138955708783331603
    Google Scholar
    There is no corresponding record for this reference.
  44. 44
    Miller Jenkins, L. M.; Ott, D. E.; Hayashi, R.; Coren, L. V.; Wang, D.; Xu, Q.; Schito, M. L.; Inman, J. K.; Appella, D. H.; Appella, E. Small-Molecule Inactivation of HIV-1 NCp7 by Repetitive Intracellular Acyl Transfer. Nat. Chem. Biol. 2010, 6, 887– 889,  DOI: 10.1038/nchembio.456
    Google Scholar
    44
    Small-molecule inactivation of HIV-1 NCp7 by repetitive intracellular acyl transfer
    Miller Jenkins, Lisa M.; Ott, David E.; Hayashi, Ryo; Coren, Lori V.; Wang, De-Yun; Xu, Qun; Schito, Marco L.; Inman, John K.; Appella, Daniel H.; Appella, Ettore
    Nature Chemical Biology (2010), 6 (12), 887-889CODEN: NCBABT; ISSN:1552-4450. (Nature Publishing Group)
    The zinc fingers of the HIV-1 nucleocapsid protein, NCp7, are prime targets for antiretroviral therapeutics. Here we show that S-acyl-2-mercaptobenzamide thioester (SAMT) chemotypes inhibit HIV by modifying the NCp7 region of Gag in infected cells, thereby blocking Gag processing and reducing infectivity. The thiol produced by SAMT reaction with NCp7 is acetylated by cellular enzymes to regenerate active SAMTs via a recycling mechanism unique among small-mol. inhibitors of HIV.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3cXhtlSgurzE&md5=4519c9f254c0bb714ae2fd9750b4e727
  45. 45
    Tedbury, P. R.; Freed, E. O. HIV-1 Gag: An Emerging Target for Antiretroviral Therapy. Curr. Top Microbiol Immunol 2015, 389, 171– 201,  DOI: 10.1007/82_2015_436
    Google Scholar
    There is no corresponding record for this reference.
  46. 46
    Deshmukh, L.; Tugarinov, V.; Appella, D. H.; Clore, G. M. Targeting a Dark Excited State of HIV-1 Nucleocapsid by Antiretroviral Thioesters Revealed by NMR Spectroscopy. Angew. Chem., Int. Ed. Engl. 2018, 57, 2687– 2691,  DOI: 10.1002/anie.201713172
    Google Scholar
    There is no corresponding record for this reference.
  47. 47
    Shimberg, G. D.; Pritts, J. D.; Michel, S. L. J. Iron-Sulfur Clusters in Zinc Finger Proteins. Methods Enzymol 2018, 599, 101– 137,  DOI: 10.1016/bs.mie.2017.09.005
    Google Scholar
    47
    Iron-sulfur clusters in zinc finger proteins
    Shimberg, Geoffrey D.; Pritts, Jordan D.; Michel, Sarah L. J.
    Methods in Enzymology (2018), 599 (Fe-S Cluster Enzymes, Part B), 101-137CODEN: MENZAU; ISSN:0076-6879. (Elsevier Inc.)
    A review. Zinc finger (ZF) proteins are proteins that use zinc as a structural cofactor. The common feature among all ZFs is that they contain repeats of four cysteine and/or histidine residues within their primary amino acid sequence. With the explosion of genome sequencing in the early 2000s, a large no. of proteins were annotated as ZFs based solely upon amino acid sequence. As these proteins began to be characterized exptl., it was discovered that some of these proteins contain iron-sulfur sites either in place of or in addn. to zinc. Here, we describe methods to isolate and characterize one such ZF protein, cleavage and polyadenylation specificity factor 30 (CPSF30) with respect to its metal-loading and RNA-binding activity.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXmvF2qsrw%253D&md5=19a750e4caba13bb2298d48e467730a6
  48. 48
    O’Brien, E.; Holt, M. E.; Thompson, M. K.; Salay, L. E.; Ehlinger, A. C.; Chazin, W. J.; Barton, J. K. The [4Fe4S] Cluster of Human DNA Primase Functions as a Redox Switch Using DNA Charge Transport. Science 2017, 355. DOI: 10.1126/science.aag1789 .
    Google Scholar
    There is no corresponding record for this reference.
  49. 49
    O’Brien, E.; Salay, L. E.; Epum, E. A.; Friedman, K. L.; Chazin, W. J.; Barton, J. K. Yeast Require Redox Switching in DNA Primase. Proc. Natl. Acad. Sci. U. S. A. 2018, 115, 13186– 13191,  DOI: 10.1073/pnas.1810715115
    Google Scholar
    49
    Yeast require redox switching in DNA primase
    O'Brien Elizabeth; Barton Jacqueline K; Salay Lauren E; Chazin Walter J; Salay Lauren E; Chazin Walter J; Salay Lauren E; Chazin Walter J; Epum Esther A; Friedman Katherine L
    Proceedings of the National Academy of Sciences of the United States of America (2018), 115 (52), 13186-13191 ISSN:.
    Eukaryotic DNA primases contain a [4Fe4S] cluster in the C-terminal domain of the p58 subunit (p58C) that affects substrate affinity but is not required for catalysis. We show that, in yeast primase, the cluster serves as a DNA-mediated redox switch governing DNA binding, just as in human primase. Despite a different structural arrangement of tyrosines to facilitate electron transfer between the DNA substrate and [4Fe4S] cluster, in yeast, mutation of tyrosines Y395 and Y397 alters the same electron transfer chemistry and redox switch. Mutation of conserved tyrosine 395 diminishes the extent of p58C participation in normal redox-switching reactions, whereas mutation of conserved tyrosine 397 causes oxidative cluster degradation to the [3Fe4S](+) species during p58C redox signaling. Switching between oxidized and reduced states in the presence of the Y397 mutations thus puts primase [4Fe4S] cluster integrity and function at risk. Consistent with these observations, we find that yeast tolerate mutations to Y395 in p58C, but the single-residue mutation Y397L in p58C is lethal. Our data thus show that a constellation of tyrosines for protein-DNA electron transfer mediates the redox switch in eukaryotic primases and is required for primase function in vivo.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB3cnht1eksw%253D%253D&md5=cc956735ff88e27f3609ae4d81e034f7
  50. 50
    Pritts, J. D.; Michel, S. L. J. Fe-S Clusters Masquerading as Zinc Finger Proteins. J. Inorg. Biochem 2022, 230, 111756,  DOI: 10.1016/j.jinorgbio.2022.111756
    Google Scholar
    50
    Fe-S clusters masquerading as zinc finger proteins
    Pritts, Jordan D.; Michel, Sarah L. J.
    Journal of Inorganic Biochemistry (2022), 230 (), 111756CODEN: JIBIDJ; ISSN:0162-0134. (Elsevier Inc.)
    A review. Metal ions are commonly found as protein co-factors in biol., and it is estd. that over a quarter of all proteins require a metal cofactor. The distribution and utilization of metals in biol. has changed over time. As the earth evolved, the atm. became increasingly oxygen rich which affected the bioavailability of certain metals such as iron, which in the oxidized ferric form is significantly less sol. than its reduced ferrous counterpart. Addnl., proteins that utilize metal cofactors for structural purposes grew in abundance, necessitating the use of metal co-factors that are not redox active, such as zinc. One common class of Zn co-factored proteins are zinc finger proteins (ZFs). ZFs bind zinc utilizing cysteine and histidine ligands to promote structure and function. Bioinformatics has annotated 5% of the human genome as ZFs; however, many of these proteins have not been studied empirically. In recent years, examples of annotated ZFs that instead harbor Fe-S clusters have been reported. In this review we highlight four examples of mis-annotated ZFs: mitoNEET, CPSF30, nsp12, and Fep1 and describe methods that can be utilized to differentiate the metal-cofactor.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB38Xlslehsbo%253D&md5=bc38950239c6930a1d2d0dd5157f8810
  51. 51
    Shimberg, G. D.; Michalek, J. L.; Oluyadi, A. A.; Rodrigues, A. V.; Zucconi, B. E.; Neu, H. M.; Ghosh, S.; Sureschandra, K.; Wilson, G. M.; Stemmler, T. L.; Michel, S. L. J. Cleavage and Polyadenylation Specificity Factor 30: An RNA-Binding Zinc-Finger Protein with an Unexpected 2Fe-2S Cluster. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 4700– 4705,  DOI: 10.1073/pnas.1517620113
    Google Scholar
    There is no corresponding record for this reference.
  52. 52
    White, M. F.; Dillingham, M. S. Iron-Sulphur Clusters in Nucleic Acid Processing Enzymes. Curr. Opin Struct. Biol. 2012, 22, 94– 100,  DOI: 10.1016/j.sbi.2011.11.004
    Google Scholar
    52
    Iron-sulphur clusters in nucleic acid processing enzymes
    White, Malcolm F.; Dillingham, Mark S.
    Current Opinion in Structural Biology (2012), 22 (1), 94-100CODEN: COSBEF; ISSN:0959-440X. (Elsevier Ltd.)
    A review. Several unexpected reports of iron-sulfur clusters in nucleic acid binding proteins have recently appeared in the literature. Once thought to be relatively rare in these systems, iron-sulfur clusters are now known to be essential components of diverse nucleic acid processing machinery including glycosylases, primases, helicases, nucleases, transcription factors, RNA polymerases and RNA methyltransferases. In many cases, the function of the cluster is poorly understood and crystal structures of these iron-sulfur enzymes reveal little in common between them. In this article, we review the recent developments in the field and discuss to what extent there might exist common mechanistic roles for iron-sulfur clusters in nucleic acid enzymes.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38XjtFSjs7g%253D&md5=cee08ff6385b7c60ce6ad49616d14b72
  53. 53
    Khodour, Y.; Kaguni, L. S.; Stiban, J. Iron-Sulfur Clusters in Nucleic Acid Metabolism: Varying Roles of Ancient Cofactors. Enzymes (Essen) 2019, 45, 225– 256,  DOI: 10.1016/bs.enz.2019.08.003
    Google Scholar
    There is no corresponding record for this reference.
  54. 54
    Maio, N.; Raza, M. K.; Li, Y.; Zhang, D. L.; Bollinger, J. M., Jr; Krebs, C.; Rouault, T. A. An Iron-Sulfur Cluster in the Zinc-Binding Domain of the SARS-CoV-2 Helicase Modulates Its RNA-Binding and -Unwinding Activities. Proc. Natl. Acad. Sci. U. S. A. 2023, 120, e2303860120  DOI: 10.1073/pnas.2303860120
    Google Scholar
    There is no corresponding record for this reference.
  55. 55
    Maio, N.; Lafont, B. A. P.; Sil, D.; Li, Y.; Bollinger, J. M., Jr; Krebs, C.; Pierson, T. C.; Linehan, W. M.; Rouault, T. A. Fe-S Cofactors in the SARS-CoV-2 RNA-Dependent RNA Polymerase Are Potential Antiviral Targets. Science 2021, 373, 236– 241,  DOI: 10.1126/science.abi5224
    Google Scholar
    55
    Fe-S cofactors in the SARS-CoV-2 RNA-dependent RNA polymerase are potential antiviral targets
    Maio, Nunziata; Lafont, Bernard A. P.; Sil, Debangsu; Li, Yan; Bollinger, J. Martin, Jr.; Krebs, Carsten; Pierson, Theodore C.; Linehan, W. Marston; Rouault, Tracey A.
    Science (Washington, DC, United States) (2021), 373 (6551), 236-241CODEN: SCIEAS; ISSN:1095-9203. (American Association for the Advancement of Science)
    Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), the causal agent of COVID-19, uses an RNA-dependent RNA polymerase (RdRp) for the replication of its genome and the transcription of its genes. The catalytic subunit of the RdRp, nsp12, ligates 2 iron-sulfur metal cofactors in sites that were modeled as zinc centers in the available cryoelectron microscopy structures of the RdRp complex. These metal binding sites are essential for replication and for interaction with the viral helicase. Oxidn. of the clusters by the stable nitroxide TEMPOL caused their disassembly, potently inhibited the RdRp, and blocked SARS-CoV-2 replication in cell culture. These iron-sulfur clusters thus serve as cofactors for the SARS-CoV-2 RdRp and are targets for therapy of COVID-19.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3MXhsFaqsLjF&md5=6cc5d644ea4ee3a5a8818e284873f4fc
  56. 56
    Weiss, A.; Murdoch, C. C.; Edmonds, K. A.; Jordan, M. R.; Monteith, A. J.; Perera, Y. R.; Rodríguez Nassif, A. M.; Petoletti, A. M.; Beavers, W. N.; Munneke, M. J. Zn-Regulated GTPase Metalloprotein Activator 1 Modulates Vertebrate Zinc Homeostasis. Cell 2022, 185, 2148– 2163,  DOI: 10.1016/j.cell.2022.04.011
    Google Scholar
    56
    Zn-regulated GTPase metalloprotein activator 1 modulates vertebrate zinc homeostasis
    Weiss, Andy; Murdoch, Caitlin C.; Edmonds, Katherine A.; Jordan, Matthew R.; Monteith, Andrew J.; Perera, Yasiru R.; Rodriguez Nassif, Aslin M.; Petoletti, Amber M.; Beavers, William N.; Munneke, Matthew J.; Drury, Sydney L.; Krystofiak, Evan S.; Thalluri, Kishore; Wu, Hongwei; Kruse, Angela R. S.; DiMarchi, Richard D.; Caprioli, Richard M.; Spraggins, Jeffrey M.; Chazin, Walter J.; Giedroc, David P.; Skaar, Eric P.
    Cell (Cambridge, MA, United States) (2022), 185 (12), 2148-2163.e27CODEN: CELLB5; ISSN:0092-8674. (Cell Press)
    Zinc (Zn) is an essential micronutrient and cofactor for up to 10% of proteins in living organisms. During Zn limitation, specialized enzymes called metallochaperones are predicted to allocate Zn to specific metalloproteins. This function was putatively assigned to G3E GTPase COG0523 proteins, yet no Zn metallochaperone was exptl. identified in any organism. Here, the authors functionally characterize a family of COG0523 proteins that is conserved across vertebrates. The authors identify Zn metalloprotease methionine aminopeptidase 1 (METAP1) as a COG0523 client, leading to the redesignation of this group of COG0523 proteins as the Zn-regulated GTPase metalloprotein activator (ZNG1) family. Using biochem., structural, genetic, and pharmacol. approaches across evolutionarily divergent models, including zebrafish and mice, the authors demonstrate a crit. role for ZNG1 proteins in regulating cellular Zn homeostasis. Collectively, these data reveal the existence of a family of Zn metallochaperones and assign ZNG1 an important role for intracellular Zn trafficking.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB38XhtlSltrvL&md5=a9b120f2fc0dcbbb8e99a2f1e5d69c9a
  57. 57
    Summers, M. F.; Henderson, L. E.; Chance, M. R.; South, T. L.; Blake, P. R.; Perez-Alvarado, G.; Bess, J. W.; Sowder, R. C.; Arthur, L. O.; Sagi, I. Nucleocapsid Zinc Fingers Detected in Retroviruses: EXAFS Studies of Intact Viruses and the Solution-state Structure of the Nucleocapsid Protein from HIV-1. Protein Sci. 1992, 1, 563– 574,  DOI: 10.1002/pro.5560010502
    Google Scholar
    57
    Nucleocapsid zinc fingers detected in retroviruses: EXAFS studies of intact viruses and the solution-state structure of the nucleocapsid protein from HIV-1
    Summers, Michael F.; Henderson, Louis E.; Chance, Mark R.; Bess, Julian W., Jr.; South, Terri L.; Blake, Paul R.; Sagi, Irit; Perez-Alvarado, Gabriela; Sowder, Raymond C., III; et al.
    Protein Science (1992), 1 (5), 563-74CODEN: PRCIEI; ISSN:0961-8368.
    All retroviral nucleocapsid (NC) proteins contain one or two copies of an invariant 3Cys-1His array (CCHC = C-X2-C-X4-H-X4-C; C = Cys, H = His, X = variable amino acid) that are essential for RNA genome packaging and infectivity and have been proposed to function as zinc-binding domains. Although the arrays are capable of binding zinc in vitro, the physiol. relevance of zinc coordination has not been firmly established. Here, zinc-edge extended X-ray absorption fine structure (EXAFS) spectra was obtained for intact retroviruses in order to det. if virus-bound zinc, which is present in quantities nearly stoichiometric with the CCHC arrays (Bess, J.W., et al., 1992) exists in a unique coordination environment. The viral EXAFS spectra obtained are remarkably similar to the spectrum of a model CCHC zinc finger peptide with known 3Cys-1His zinc coordination structure. This finding, combined with other biochem. results, indicates that the majority of the viral zinc is coordinated to the NC CCHC arrays in mature retroviruses. Based on the these findings, NMR studies of the HIV-1 NC protein were extended and the three-dimensional soln.-state structure of the protein detd . The CCHC arrays of HIV-1 NC exist as independently folded, noninteracting domains on a flexible polypeptide chain, with conservatively substituted arom. residues forming hydrophobic patches on the zinc finger surfaces. These residues are essential for RNA genome recognition, and fluorescence measurements indicate that at least one residue (Trp37) participates directly in binding to nucleic acids in vitro. The NC is only the third HIV-1 protein to be structurally characterized, and the combined EXAFS, structural, and nucleic acid-binding results provide a basis for the rational design of new NC-targeted antiviral agents and vaccines for the control of AIDS.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADyaK38XkvVGnu7s%253D&md5=590108d91cc63cad9a2a179ed8cbc88f
  58. 58
    Newman, J. A.; Douangamath, A.; Yadzani, S.; Yosaatmadja, Y.; Aimon, A.; Brandão-Neto, J.; Dunnett, L.; Gorrie-stone, T.; Skyner, R.; Fearon, D. Structure, Mechanism and Crystallographic Fragment Screening of the SARS-CoV-2 NSP13 Helicase. Nat. Commun. 2021, 12, 4848,  DOI: 10.1038/s41467-021-25166-6
    Google Scholar
    58
    Structure, mechanism and crystallographic fragment screening of the SARS-CoV-2 NSP13 helicase
    Newman, Joseph A.; Douangamath, Alice; Yadzani, Setayesh; Yosaatmadja, Yuliana; Aimon, Antony; Brandao-Neto, Jose; Dunnett, Louise; Gorrie-stone, Tyler; Skyner, Rachael; Fearon, Daren; Schapira, Matthieu; von Delft, Frank; Gileadi, Opher
    Nature Communications (2021), 12 (1), 4848CODEN: NCAOBW; ISSN:2041-1723. (Nature Research)
    Abstr.: There is currently a lack of effective drugs to treat people infected with SARS-CoV-2, the cause of the global COVID-19 pandemic. The SARS-CoV-2 Non-structural protein 13 (NSP13) has been identified as a target for anti-virals due to its high sequence conservation and essential role in viral replication. Structural anal. reveals two "druggable" pockets on NSP13 that are among the most conserved sites in the entire SARS-CoV-2 proteome. Here we present crystal structures of SARS-CoV-2 NSP13 solved in the APO form and in the presence of both phosphate and a non-hydrolysable ATP analog. Comparisons of these structures reveal details of conformational changes that provide insights into the helicase mechanism and possible modes of inhibition. To identify starting points for drug development we have performed a crystallog. fragment screen against NSP13. The screen reveals 65 fragment hits across 52 datasets opening the way to structure guided development of novel antiviral agents.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3MXhslOns7%252FI&md5=3e643800631b8bb56328c4ddd97fb7a7
  59. 59
    Bellomo, E.; Birla Singh, K.; Massarotti, A.; Hogstrand, C.; Maret, W. The Metal Face of Protein Tyrosine Phosphatase 1B. Coord. Chem. Rev. 2016, 327–328, 70– 83,  DOI: 10.1016/j.ccr.2016.07.002
    Google Scholar
    59
    The metal face of protein tyrosine phosphatase 1B
    Bellomo, Elisa; Birla Singh, Kshetrimayum; Massarotti, Alberto; Hogstrand, Christer; Maret, Wolfgang
    Coordination Chemistry Reviews (2016), 327-328 (), 70-83CODEN: CCHRAM; ISSN:0010-8545. (Elsevier B.V.)
    A review. A new paradigm in metallobiochem. describes the activation of inactive metalloenzymes by metal ion removal. Protein tyrosine phosphatases (PTPs) do not seem to require a metal ion for enzymic activity. However, both metal cations and metal anions modulate their enzymic activity. One binding site is the phosphate binding site at the catalytic cysteine residue. Oxyanions with structural similarity to phosphate, such as vanadate, inhibit the enzyme with nanomolar to micromolar affinities. In addn., zinc ions (Zn2+) inhibit with picomolar to nanomolar affinities. We mapped the cation binding site close to the anion binding site and established a specific mechanism of inhibition occurring only in the closed conformation of the enzyme when the catalytic cysteine is phosphorylated and the catalytic aspartate moves into the active site. We discuss this dual inhibition by anions and cations here for PTP1B, the most thoroughly investigated protein tyrosine phosphatase. The significance of the inhibition in phosphorylation signaling is becoming apparent only from the functions of PTP1B in the biol. context of metal cations as cellular signaling ions. Zinc ion signals complement redox signals but provide a different type of control and longer lasting inhibition on a biol. time scale owing to the specificity and affinity of zinc ions for coordination environments. Inhibitor design for PTP1B and other PTPs is a major area of research activity and interest owing to their prominent roles in metabolic regulation in health and disease, in particular cancer and diabetes. Our results explain the apparent dichotomy of both cations (Zn2+) and oxyanions such as vanadate inhibiting PTP1B and having insulin-enhancing ("anti-diabetic") effects and suggest different approaches, namely targeting PTPs in the cell by affecting their physiol. modulators and considering a metallodrug approach that builds on the knowledge of the insulin-enhancing effects of both zinc and vanadium compds.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XhsFSjsLrI&md5=b82d368ac3fc1a494d0c7683b28ab00a
  60. 60
    Bellomo, E.; Massarotti, A.; Hogstrand, C. Zinc Ions Modulate Protein Tyrosine Phosphatase 1B Activity. Metallomics 2014, 6, 1229– 1239,  DOI: 10.1039/C4MT00086B
    Google Scholar
    60
    Zinc ions modulate protein tyrosine phosphatase 1B activity
    Bellomo, Elisa; Massarotti, Alberto; Hogstrand, Christer; Maret, Wolfgang
    Metallomics (2014), 6 (7), 1229-1239CODEN: METAJS; ISSN:1756-591X. (Royal Society of Chemistry)
    Protein tyrosine phosphatases (PTPs) are key enzymes in cellular regulation. The 107 human PTPs are regulated by redox signaling, phosphorylation, dimerization, and proteolysis. Recent findings of very strong inhibition of some PTPs by Zn2+ ions at concns. relevant in a cellular environment suggest yet another mechanism of regulation. One of the most extensively investigated PTPs is PTP1B (PTPN1). It regulates the insulin and leptin signaling pathway and is implicated in cancer and obesity/diabetes. The development of novel assay conditions to investigate Zn2+ inhibition of PTP1B provided ests. of ∼5.6 nM affinity for inhibitory Zn2+ ions. Anal. of 3 PTP1B 3-dimensional structures (PDB id: 2CM2, 3I80, and 1A5Y) identified putative Zn2+-binding sites and supported the kinetic studies in suggesting an inhibitory Zn2+ only in the closed and cysteinyl-phosphate intermediate forms of the enzyme. These observations gained significance with regard to recent findings of regulatory roles of Zn2+ ions released from the endoplasmic reticulum.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhtVKjsLzN&md5=91376647da286fc94298ffcbf60c3b14
  61. 61
    Bellomo, E.; Abro, A.; Hogstrand, C.; Maret, W.; Domene, C. Role of Zinc and Magnesium Ions in the Modulation of Phosphoryl Transfer in Protein Tyrosine Phosphatase 1B. J. Am. Chem. Soc. 2018, 140, 4446– 4454,  DOI: 10.1021/jacs.8b01534
    Google Scholar
    61
    Role of Zinc and Magnesium Ions in the Modulation of Phosphoryl Transfer in Protein Tyrosine Phosphatase 1B
    Bellomo, Elisa; Abro, Asma; Hogstrand, Christer; Maret, Wolfgang; Domene, Carmen
    Journal of the American Chemical Society (2018), 140 (12), 4446-4454CODEN: JACSAT; ISSN:0002-7863. (American Chemical Society)
    While the majority of phosphatases are metalloenzymes, the prevailing model for the reactions catalyzed by protein tyrosine phosphatases does not involve any metal ion. Yet, both metal cations and oxoanions affect their enzymic activity. Mg2+ and Zn2+ activate and inhibit, resp., protein tyrosine phosphatase 1B (PTP1B). Mol. dynamics simulations, metadynamics and quantum chem. calcns. in combination with exptl. investigations demonstrate that Mg2+ and Zn2+ compete for the same binding site in the active site only in the closed conformation of the enzyme in its phosphorylated state. The two cations have different effects on the arrangements and activities of water mols. that are necessary for the hydrolysis of the phosphocysteine intermediate in the second catalytic step of the reaction. Remarkable differences between the established structural enzymol. of PTP1B investigated ex vivo and the function of PTP1B in vivo become evident. Different reaction pathways are viable when the presence of metal ions and their cellular concns. are considered. The findings suggest that the substrate delivers the inhibitory Zn2+ ion to the active site. The inhibition and activation can be ascribed to the different coordination chem. of Zn2+ and Mg2+ ions and the orientation of the metal-coordinated water mols. Metallochem. adds an addnl. dimension to the regulation of PTP1B, and presumably other members of this enzyme family.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXjvFWnsr8%253D&md5=1b44f32bfbc431c825e282ede4448c6c
  62. 62
    Cheng, Y.; Wang, H.; Xu, H.; Liu, Y.; Ma, B.; Chen, X.; Zeng, X.; Wang, X.; Wang, B.; Shiau, C. Co-Evolution-Based Prediction of Metal-Binding Sites in Proteomes by Machine Learning. Nat. Chem. Biol. 2023, 19, 548– 555,  DOI: 10.1038/s41589-022-01223-z
    Google Scholar
    62
    Co-evolution-based prediction of metal-binding sites in proteomes by machine learning
    Cheng, Yao; Wang, Haobo; Xu, Hua; Liu, Yuan; Ma, Bin; Chen, Xuemin; Zeng, Xin; Wang, Xianghe; Wang, Bo; Shiau, Carina; Ovchinnikov, Sergey; Su, Xiao-Dong; Wang, Chu
    Nature Chemical Biology (2023), 19 (5), 548-555CODEN: NCBABT; ISSN:1552-4450. (Nature Portfolio)
    Metal ions have various important biol. roles in proteins, including structural maintenance, mol. recognition and catalysis. Previous methods of predicting metal-binding sites in proteomes were based on either sequence or structural motifs. Here we developed a co-evolution-based pipeline named 'MetalNet' to systematically predict metal-binding sites in proteomes. We applied MetalNet to proteomes of four representative prokaryotic species and predicted 4,849 potential metalloproteins, which substantially expands the currently annotated metalloproteomes. We biochem. and structurally validated previously unannotated metal-binding sites in several proteins, including apo-citrate lyase phosphoribosyl-dephospho-CoA transferase citX, an Escherichia coli enzyme lacking structural or sequence homol. to any known metalloprotein (Protein Data Bank (PDB) codes: 7DCM and 7DCN). MetalNet also successfully recapitulated all known zinc-binding sites from the human spliceosome complex. The pipeline of MetalNet provides a unique and enabling tool for interrogating the hidden metalloproteome and studying metal biol. [graphic not available: see fulltext].
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3sXitlClsA%253D%253D&md5=141b5e802d6126c1108157954a1ad713
  63. 63
    Durr, S. L.; Levy, A.; Rothlisberger, U. Metal3D: A General Deep Learning Framework for Accurate Metal Ion Location Prediction in Proteins. Nat. Commun. 2023, 14, 2713,  DOI: 10.1038/s41467-023-37870-6
    Google Scholar
    There is no corresponding record for this reference.
  64. 64
    Putignano, V.; Rosato, A.; Banci, L.; Andreini, C. MetalPDB in 2018: A Database of Metal Sites in Biological Macromolecular Structures. Nucleic Acids Res. 2018, 46, D459– D464,  DOI: 10.1093/nar/gkx989
    Google Scholar
    64
    MetalPDB in 2018: a database of metal sites in biological macromolecular structures
    Putignano, Valeria; Rosato, Antonio; Banci, Lucia; Andreini, Claudia
    Nucleic Acids Research (2018), 46 (D1), D459-D464CODEN: NARHAD; ISSN:1362-4962. (Oxford University Press)
    A review. MetalPDB is a database providing information on metal-binding sites detected in the three-dimensional (3D) structures of biol. macromols. MetalPDB represents such sites as 3D templates, called Minimal Functional Sites (MFSs), which describe the local environment around the metal(s) independently of the larger context of the macromol. structure. The 2018 update of MetalPDB includes new contents and tools. A major extension is the inclusion of proteins whose structures do not contain metal ions although their sequences potentially contain a known MFS. In addn., MetalPDB now provides extensive statistical analyses addressing several aspects of general metal usage within the PDB, across protein families and in catalysis. Users can also query MetalPDB to ext. statistical information on structural aspects assocd. with individual metals, such as preferred coordination geometries or amino-acidic environment. A further major improvement is the functional annotation of MFSs; the annotation is manually performed via a password-protected annotator interface. At present, ∼50% of all MFSs have such a functional annotation. Other noteworthy improvements are bulk query functionality, through the upload of a list of PDB identifiers, and ftp access to MetalPDB contents, allowing users to carry out in-depth analyses on their own computational infrastructure.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXitlGisL3N&md5=a3c7d882fd20192f2668fa0c9e9f1805
  65. 65
    Laveglia, V.; Giachetti, A.; Sala, D.; Andreini, C.; Rosato, A. Learning to Identify Physiological and Adventitious Metal-Binding Sites in the Three-Dimensional Structures of Proteins by Following the Hints of a Deep Neural Network. J. Chem. Inf Model 2022, 62, 2951– 2960,  DOI: 10.1021/acs.jcim.2c00522
    Google Scholar
    There is no corresponding record for this reference.
  66. 66
    Feehan, R.; Franklin, M. W.; Slusky, J. S. G. Machine Learning Differentiates Enzymatic and Non-Enzymatic Metals in Proteins. Nat. Commun. 2021, 12, 3712,  DOI: 10.1038/s41467-021-24070-3
    Google Scholar
    66
    Machine learning differentiates enzymatic and non-enzymatic metals in proteins
    Feehan, Ryan; Franklin, Meghan W.; Slusky, Joanna S. G.
    Nature Communications (2021), 12 (1), 3712CODEN: NCAOBW; ISSN:2041-1723. (Nature Research)
    Abstr.: Metalloenzymes are 40% of all enzymes and can perform all seven classes of enzyme reactions. Because of the physicochem. similarities between the active sites of metalloenzymes and inactive metal binding sites, it is challenging to differentiate between them. Yet distinguishing these two classes is crit. for the identification of both native and designed enzymes. Because of similarities between catalytic and non-catalytic metal binding sites, finding physicochem. features that distinguish these two types of metal sites can indicate aspects that are crit. to enzyme function. In this work, we develop the largest structural dataset of enzymic and non-enzymic metalloprotein sites to date. We then use a decision-tree ensemble machine learning model to classify metals bound to proteins as enzymic or non-enzymic with 92.2% precision and 90.1% recall. Our model scores electrostatic and pocket lining features as more important than pocket vol., despite the fact that vol. is the most quant. different feature between enzyme and non-enzymic sites. Finally, we find our model has overall better performance in a side-to-side comparison against other methods that differentiate enzymic from non-enzymic sequences. We anticipate that our model's ability to correctly identify which metal sites are responsible for enzymic activity could enable identification of new enzymic mechanisms and de novo enzyme design.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3MXhsFKnsr%252FN&md5=d90cfaf3fcde254c98b6439f9c12705f
  67. 67
    Laveglia, V.; Bazayeva, M.; Andreini, C.; Rosato, A. Hunting down Zinc(II)-Binding Sites in Proteins with Distance Matrices. Bioinformatics 2023, 39, btad653,  DOI: 10.1093/bioinformatics/btad653
    Google Scholar
    There is no corresponding record for this reference.
  68. 68
    Zhou, Y.; Li, H.; Sun, H. Metalloproteomics for Biomedical Research: Methodology and Applications. Annu. Rev. Biochem. 2022, 91, 449– 473,  DOI: 10.1146/annurev-biochem-040320-104628
    Google Scholar
    68
    Metalloproteomics for Biomedical Research: Methodology and Applications
    Zhou Ying; Li Hongyan; Sun Hongzhe
    Annual review of biochemistry (2022), 91 (), 449-473 ISSN:.
    Metals are essential components in life processes and participate in many important biological processes. Dysregulation of metal homeostasis is correlated with many diseases. Metals are also frequently incorporated into diagnosis and therapeutics. Understanding of metal homeostasis under (patho)physiological conditions and the molecular mechanisms of action of metallodrugs in biological systems has positive impacts on human health. As an emerging interdisciplinary area of research, metalloproteomics involves investigating metal-protein interactions in biological systems at a proteome-wide scale, has received growing attention, and has been implemented into metal-related research. In this review, we summarize the recent advances in metalloproteomics methodologies and applications. We also highlight emerging single-cell metalloproteomics, including time-resolved inductively coupled plasma mass spectrometry, mass cytometry, and secondary ion mass spectrometry. Finally, we discuss future perspectives in metalloproteomics, aiming to attract more original research to develop more advanced methodologies, which could be utilized rapidly by biochemists or biologists to expand our knowledge of how metal functions in biology and medicine.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A280%3ADC%252BB2MzlsFansQ%253D%253D&md5=311079a04a9d40feee923025da4a465c
  69. 69
    Micelli, C.; Dai, Y.; Raustad, N.; Isberg, R. R.; Dowson, C. G.; Lloyd, A. J.; Geisinger, E.; Crow, A.; Roper, D. I. A Conserved Zinc-Binding Site in Acinetobacter baumannii PBP2 Required for Elongasome-Directed Bacterial Cell Shape. Proc. Natl. Acad. Sci. U. S. A. 2023, 120, e2215237120  DOI: 10.1073/pnas.2215237120
    Google Scholar
    There is no corresponding record for this reference.
  70. 70
    Murdoch, C. C.; Skaar, E. P. Nutritional Immunity: The Battle for Nutrient Metals at the Host-Pathogen Interface. Nat. Rev. Microbiol 2022, 20, 657– 670,  DOI: 10.1038/s41579-022-00745-6
    Google Scholar
    70
    Nutritional immunity: the battle for nutrient metals at the host-pathogen interface
    Murdoch, Caitlin C.; Skaar, Eric P.
    Nature Reviews Microbiology (2022), 20 (11), 657-670CODEN: NRMACK; ISSN:1740-1526. (Nature Portfolio)
    A review. Trace metals are essential micronutrients required for survival across all kingdoms of life. From bacteria to animals, metals have crit. roles as both structural and catalytic cofactors for an estd. third of the proteome, representing a major contributor to the maintenance of cellular homeostasis. The reactivity of metal ions engenders them with the ability to promote enzyme catalysis and stabilize reaction intermediates. However, these properties render metals toxic at high concns. and, therefore, metal levels must be tightly regulated. Having evolved in close assocn. with bacteria, vertebrate hosts have developed numerous strategies of metal limitation and intoxication that prevent bacterial proliferation, a process termed nutritional immunity. In turn, bacterial pathogens have evolved adaptive mechanisms to survive in conditions of metal depletion or excess. In this review, we discuss mechanisms by which nutrient metals shape the interactions between bacterial pathogens and animal hosts. We explore the cell-specific and tissue-specific roles of distinct trace metals in shaping bacterial infections, as well as implications for future research and new therapeutic development.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB38XhsVahsLbP&md5=fe3589d7b9cdd4c249eef63151b80755
  71. 71
    Weinberg, E. D. Nutritional Immunity. Host’s Attempt to Withhold Iron from Microbial Invaders. JAMA 1975, 231, 39– 41,  DOI: 10.1001/jama.1975.03240130021018
    Google Scholar
    There is no corresponding record for this reference.
  72. 72
    Kehl-Fie, T. E.; Skaar, E. P. Nutritional Immunity beyond Iron: A Role for Manganese and Zinc. Curr. Opin Chem. Biol. 2010, 14, 218– 224,  DOI: 10.1016/j.cbpa.2009.11.008
    Google Scholar
    72
    Nutritional immunity beyond iron: A role for manganese and zinc
    Kehl-Fie, Thomas E.; Skaar, Eric P.
    Current Opinion in Chemical Biology (2010), 14 (2), 218-224CODEN: COCBF4; ISSN:1367-5931. (Elsevier B.V.)
    A review. Vertebrates sequester iron from invading pathogens, and conversely, pathogens express a variety of factors to steal iron from the host. Recent work has demonstrated that in addn. to iron, vertebrates sequester zinc and manganese both intracellularly and extracellularly to protect against infection. Intracellularly, vertebrates utilize the ZIP/ZnT families of transporters to manipulate zinc levels, as well as Nramp1 to manipulate manganese levels. Extracellularly, the S100 protein calprotectin sequesters manganese and potentially zinc to inhibit microbial growth. To circumvent these defenses, bacteria possess high affinity transporters to import specific nutrient metals. Limiting the availability of zinc and manganese as a mechanism to defend against infection expands the spectrum of nutritional immunity and further establishes metal sequestration as a key defense against microbial invaders.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC3cXjvFKnt7k%253D&md5=8428db56891c63c53ffc76284287450e
  73. 73
    Lonergan, Z. R.; Nairn, B. L.; Wang, J.; Hsu, Y.-P.; Hesse, L. E.; Beavers, W. N.; Chazin, W. J.; Trinidad, J. C.; VanNieuwenhze, M. S.; Giedroc, D. P. An Acinetobacter baumannii, Zinc-Regulated Peptidase Maintains Cell Wall Integrity during Immune-Mediated Nutrient Sequestration. Cell Rep 2019, 26, 2009– 2018,  DOI: 10.1016/j.celrep.2019.01.089
    Google Scholar
    73
    An Acinetobacter baumannii, Zinc-Regulated Peptidase Maintains Cell Wall Integrity during Immune-Mediated Nutrient Sequestration
    Lonergan, Zachery R.; Nairn, Brittany L.; Wang, Jiefei; Hsu, Yen-Pang; Hesse, Laura E.; Beavers, William N.; Chazin, Walter J.; Trinidad, Jonathan C.; VanNieuwenhze, Michael S.; Giedroc, David P.; Skaar, Eric P.
    Cell Reports (2019), 26 (8), 2009-2018.e6CODEN: CREED8; ISSN:2211-1247. (Cell Press)
    Acinetobacter baumannii is an important nosocomial pathogen capable of causing wound infections, pneumonia, and bacteremia. During infection, A. baumannii must acquire Zn to survive and colonize the host. Vertebrates have evolved mechanisms to sequester Zn from invading pathogens by a process termed nutritional immunity. One of the most upregulated genes during Zn starvation encodes a putative cell wall-modifying enzyme which we named ZrlA. We found that inactivation of zrlA diminished growth of A. baumannii during Zn starvation. Addnl., this mutant strain displays increased cell envelope permeability, decreased membrane barrier function, and aberrant peptidoglycan muropeptide abundances. This altered envelope increases antibiotic efficacy both in vitro and in an animal model of A. baumannii pneumonia. These results establish ZrlA as a crucial link between nutrient metal uptake and cell envelope homeostasis during A. baumannii pathogenesis, which could be targeted for therapeutic development.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1MXjsVegtLg%253D&md5=21f01ebeef4071e76a0b99a6add51de2
  74. 74
    Sychantha, D.; Rotondo, C. M.; Tehrani, K.; Martin, N. I.; Wright, G. D. Aspergillomarasmine A Inhibits Metallo-Beta-Lactamases by Selectively Sequestering Zn2+. J. Biol. Chem. 2021, 297, 100918,  DOI: 10.1016/j.jbc.2021.100918
    Google Scholar
    There is no corresponding record for this reference.
  75. 75
    King, A. M.; Reid-Yu, S. A.; Wang, W.; King, D. T.; De Pascale, G.; Strynadka, N. C.; Walsh, T. R.; Coombes, B. K.; Wright, G. D. Aspergillomarasmine A Overcomes Metallo-Beta-Lactamase Antibiotic Resistance. Nature 2014, 510, 503– 506,  DOI: 10.1038/nature13445
    Google Scholar
    75
    Aspergillomarasmine A overcomes metallo-β-lactamase antibiotic resistance
    King, Andrew M.; Reid-Yu, Sarah A.; Wang, Wenliang; King, Dustin T.; De Pascale, Gianfranco; Strynadka, Natalie C.; Walsh, Timothy R.; Coombes, Brian K.; Wright, Gerard D.
    Nature (London, United Kingdom) (2014), 510 (7506), 503-506CODEN: NATUAS; ISSN:0028-0836. (Nature Publishing Group)
    The emergence and spread of carbapenem-resistant Gram-neg. pathogens is a global public health problem. The acquisition of metallo-β-lactamases (MBLs) such as NDM-1 is a principle contributor to the emergence of carbapenem-resistant Gram-neg. pathogens that threatens the use of penicillin, cephalosporin and carbapenem antibiotics to treat infections. To date, a clin. inhibitor of MBLs that could reverse resistance and re-sensitize resistant Gram-neg. pathogens to carbapenems has not been found. Here we have identified a fungal natural product, aspergillomarasmine A (AMA), that is a rapid and potent inhibitor of the NDM-1 enzyme and another clin. relevant MBL, VIM-2. AMA also fully restored the activity of meropenem against Enterobacteriaceae, Acinetobacter spp. and Pseudomonas spp. possessing either VIM or NDM-type alleles. In mice infected with NDM-1-expressing Klebsiella pneumoniae, AMA efficiently restored meropenem activity, demonstrating that a combination of AMA and a carbapenem antibiotic has therapeutic potential to address the clin. challenge of MBL-pos. carbapenem-resistant Gram-neg. pathogens.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2cXhtVCqu7fN&md5=d60cc033c506e48e3c4d7b71f003181a
  76. 76
    Bahr, G.; Gonzalez, L. J.; Vila, A. J. Metallo-Beta-Lactamases in the Age of Multidrug Resistance: From Structure and Mechanism to Evolution, Dissemination, and Inhibitor Design. Chem. Rev. 2021, 121, 7957– 8094,  DOI: 10.1021/acs.chemrev.1c00138
    Google Scholar
    76
    Metallo-β-lactamases in the Age of Multidrug Resistance: From Structure and Mechanism to Evolution, Dissemination, and Inhibitor Design
    Bahr, Guillermo; Gonzalez, Lisandro J.; Vila, Alejandro J.
    Chemical Reviews (Washington, DC, United States) (2021), 121 (13), 7957-8094CODEN: CHREAY; ISSN:0009-2665. (American Chemical Society)
    A review. Antimicrobial resistance is one of the major problems in current practical medicine. The spread of genes coding for resistance determinants among bacteria challenges the use of approved antibiotics, narrowing the options for treatment. Resistance to carbapenems, last resort antibiotics, is a major concern. Metallo-β-lactamases (MBLs) hydrolyze carbapenems, penicillins, and cephalosporins, becoming central to this problem. These enzymes diverge with respect to serine-β-lactamases by exhibiting a different fold, active site, and catalytic features. Elucidating their catalytic mechanism has been a big challenge in the field that has limited the development of useful inhibitors. This review covers exhaustively the details of the active-site chemistries, the diversity of MBL alleles, the catalytic mechanism against different substrates, and how this information has helped developing inhibitors. We also discuss here different aspects crit. to understand the success of MBLs in conferring resistance: the mol. determinants of their dissemination, their cell physiol., from the biogenesis to the processing involved in the transit to the periplasm, and the uptake of the Zn(II) ions upon metal starvation conditions, such as those encountered during an infection. In this regard, the chem., biochem. and microbiol. aspects provide an integrative view of the current knowledge of MBLs.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BB3MXhtlSmu7vL&md5=84c7a6e182fe24e73b4715920306e0bc
  77. 77
    Bahr, G.; Vitor-Horen, L.; Bethel, C. R.; Bonomo, R. A.; González, L. J.; Vila, A. J. Clinical Evolution of New Delhi Metallo-Beta-Lactamase (NDM) Optimizes Resistance under Zn(II) Deprivation. Antimicrob. Agents Chemother. 2018, 62, e01849-17  DOI: 10.1128/AAC.01849-17
    Google Scholar
    There is no corresponding record for this reference.
  78. 78
    Gonzalez, L. J.; Bahr, G.; Gonzalez, M. M.; Bonomo, R. A.; Vila, A. J. In-Cell Kinetic Stability Is an Essential Trait in Metallo-Beta-Lactamase Evolution. Nat. Chem. Biol. 2023, 19, 1116– 1126,  DOI: 10.1038/s41589-023-01319-0
    Google Scholar
    There is no corresponding record for this reference.
  79. 79
    Heffern, M. C.; Park, H. M.; Au-Yeung, H. Y.; Van De Bittner, G. C.; Ackerman, C. M.; Stahl, A.; Chang, C. J. In Vivo Bioluminescence Imaging Reveals Copper Deficiency in a Murine Model of Nonalcoholic Fatty Liver Disease. Proc. Natl. Acad. Sci. U. S. A. 2016, 113, 14219– 14224,  DOI: 10.1073/pnas.1613628113
    Google Scholar
    79
    In vivo bioluminescence imaging reveals copper deficiency in a murine model of nonalcoholic fatty liver disease
    Heffern, Marie C.; Park, Hyo Min; Au-Yeung, Ho Yu; Van de Bittner, Genevieve C.; Ackerman, Cheri M.; Stahl, Andreas; Chang, Christopher J.
    Proceedings of the National Academy of Sciences of the United States of America (2016), 113 (50), 14219-14224CODEN: PNASA6; ISSN:0027-8424. (National Academy of Sciences)
    Copper is a required metal nutrient for life, but global or local alterations in its homeostasis are linked to diseases spanning genetic and metabolic disorders to cancer and neurodegeneration. Technologies that enable longitudinal in vivo monitoring of dynamic copper pools can help meet the need to study the complex interplay between copper status, health, and disease in the same living organism over time. Here, the authors present the synthesis, characterization, and in vivo imaging applications of Copper-Caged Luciferin-1 (CCL-1), a bioluminescent reporter for tissue-specific copper visualization in living animals. CCL-1 uses a selective copper(I)-dependent oxidative cleavage reaction to release d-luciferin for subsequent bioluminescent reaction with firefly luciferase. The probe can detect physiol. changes in labile Cu+ levels in live cells and mice under situations of copper deficiency or overload. Application of CCL-1 to mice with liver-specific luciferase expression in a diet-induced model of nonalcoholic fatty liver disease reveals onset of hepatic copper deficiency and altered expression levels of central copper trafficking proteins that accompany symptoms of glucose intolerance and wt. gain. The data connect copper dysregulation to metabolic liver disease and provide a starting point for expanding the toolbox of reactivity-based chem. reporters for cell- and tissue-specific in vivo imaging.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XhvFGgu7zM&md5=d86c2b8129e1c52a93a1e3f72ebd0a18
  80. 80
    Ackerman, C. M.; Chang, C. J. Copper Signaling in the Brain and Beyond. J. Biol. Chem. 2018, 293, 4628– 4635,  DOI: 10.1074/jbc.R117.000176
    Google Scholar
    80
    Copper signaling in the brain and beyond
    Ackerman, Cheri M.; Chang, Christopher J.
    Journal of Biological Chemistry (2018), 293 (13), 4628-4635CODEN: JBCHA3; ISSN:0021-9258. (American Society for Biochemistry and Molecular Biology)
    A review. Transition metals have been recognized and studied primarily in the context of their essential roles as structural and metabolic cofactors for biomols. that compose living systems. More recently, an emerging paradigm of transition-metal signaling, where dynamic changes in transitional metal pools can modulate protein function, cell fate, and organism health and disease, has broadened our view of the potential contributions of these essential nutrients in biol. Using Cu as a canonical example of transition-metal signaling, we highlight key expts. where direct measurement and/or visualization of dynamic Cu pools, in combination with biochem., physiol., and behavioral studies, have deciphered sources, targets, and physiol. effects of Cu signals.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC1cXms1Gmsrw%253D&md5=b573deb8d779360483e4956d836a178a
  81. 81
    Krishnamoorthy, L.; Cotruvo, J. A.; Chan, J.; Kaluarachchi, H.; Muchenditsi, A.; Pendyala, V. S.; Jia, S.; Aron, A. T.; Ackerman, C. M.; Wal, M. N. V. Copper Regulates Cyclic-AMP-Dependent Lipolysis. Nat. Chem. Biol. 2016, 12, 586– 592,  DOI: 10.1038/nchembio.2098
    Google Scholar
    81
    Copper regulates cyclic-AMP-dependent lipolysis
    Krishnamoorthy, Lakshmi; Cotruvo, Joseph A., Jr.; Chan, Jefferson; Kaluarachchi, Harini; Muchenditsi, Abigael; Pendyala, Venkata S.; Jia, Shang; Aron, Allegra T.; Ackerman, Cheri M.; Vander Wal, Mark N.; Guan, Timothy; Smaga, Lukas P.; Farhi, Samouil L.; New, Elizabeth J.; Lutsenko, Svetlana; Chang, Christopher J.
    Nature Chemical Biology (2016), 12 (8), 586-592CODEN: NCBABT; ISSN:1552-4450. (Nature Publishing Group)
    Cell signaling relies extensively on dynamic pools of redox-inactive metal ions such as sodium, potassium, calcium and zinc, but their redox-active transition metal counterparts such as copper and iron have been studied primarily as static enzyme cofactors. Here we report that copper is an endogenous regulator of lipolysis, the breakdown of fat, which is an essential process in maintaining body wt. and energy stores. Using a mouse model of genetic copper misregulation, in combination with pharmacol. alterations in copper status and imaging studies in a 3T3-L1 white adipocyte model, we found that copper regulates lipolysis at the level of the second messenger, cAMP, by altering the activity of the cAMP-degrading phosphodiesterase PDE3B. Biochem. studies of the copper-PDE3B interaction establish copper-dependent inhibition of enzyme activity and identify a key conserved cysteine residue in a PDE3-specific loop that is essential for the obsd. copper-dependent lipolytic phenotype.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC28XpsVCrs7k%253D&md5=8e8bad22b14a32decedc017dddc52da5
  82. 82
    Chang, C. J. Searching for Harmony in Transition-Metal Signaling. Nat. Chem. Biol. 2015, 11, 744– 7,  DOI: 10.1038/nchembio.1913
    Google Scholar
    82
    Searching for harmony in transition-metal signaling
    Chang, Christopher J.
    Nature Chemical Biology (2015), 11 (10), 744-747CODEN: NCBABT; ISSN:1552-4450. (Nature Publishing Group)
    A review. The recent emergence of signaling roles for transition metals presages a broader contribution of these elements beyond their traditional functions as metabolic cofactors. New chem. approaches to identify the sources, targets and physiologies of transition-metal signaling can help expand understanding of the periodic table in a biol. context.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC2MXhsFeisb3O&md5=0b0dc1699e35a2ec1a2a3750a3731a8d
  83. 83
    Lutsenko, S. Dynamic and Cell-Specific Transport Networks for Intracellular Copper Ions. J. Cell Sci. 2021, 134, jcs240523,  DOI: 10.1242/jcs.240523
    Google Scholar
    There is no corresponding record for this reference.
  84. 84
    Vitaliti, A.; De Luca, A.; Rossi, L. Copper-Dependent Kinases and Their Role in Cancer Inception, Progression and Metastasis. Biomolecules 2022, 12, 1520,  DOI: 10.3390/biom12101520
    Google Scholar
    There is no corresponding record for this reference.
  85. 85
    Lu, J.; Liu, X.; Li, X.; Li, H.; Shi, L.; Xia, X.; He, B.; Meyer, T. F.; Li, X.; Sun, H.; Yang, X. Copper Regulates the Host Innate Immune Response against Bacterial Infection via Activation of ALPK1 Kinase. Proc. Natl. Acad. Sci. U. S. A. 2024, 121, 2017,  DOI: 10.1073/pnas.2311630121
    Google Scholar
    There is no corresponding record for this reference.
  86. 86
    Shanbhag, V. C.; Gudekar, N.; Jasmer, K. Copper Metabolism as a Unique Vulnerability in Cancer. Biochim Biophys Acta - Mol. Cell Res. 2021, 1868, 118893,  DOI: 10.1016/j.bbamcr.2020.118893
    Google Scholar
    There is no corresponding record for this reference.
  87. 87
    Turski, M. L.; Brady, D. C.; Kim, H. J.; Kim, B.-E.; Nose, Y.; Counter, C. M.; Winge, D. R.; Thiele, D. J. A Novel Role for Copper in Ras/Mitogen-Activated Protein Kinase Signaling. Mol. Cell. Biol. 2012, 32, 1284– 1295,  DOI: 10.1128/MCB.05722-11
    Google Scholar
    87
    A novel role for copper in Ras/mitogen-activated protein kinase signaling
    Turski, Michelle L.; Brady, Donita C.; Kim, Hyung J.; Kim, Byung-Eun; Nose, Yasuhiro; Counter, Christopher M.; Winge, Dennis R.; Thiele, Dennis J.
    Molecular and Cellular Biology (2012), 32 (7), 1284-1295CODEN: MCEBD4; ISSN:0270-7306. (American Society for Microbiology)
    Copper (Cu) is essential for development and proliferation, yet the cellular requirements for Cu in these processes are not well defined. We report that Cu plays an unanticipated role in the mitogen-activated protein (MAP) kinase pathway. Ablation of the Ctr1 high-affinity Cu transporter in flies and mouse cells, mutation of Ctr1, and Cu chelators all reduce the ability of the MAP kinase kinase Mek1 to phosphorylate the MAP kinase Erk. Moreover, mice bearing a cardiac-tissue-specific knockout of Ctr1 are deficient in Erk phosphorylation in cardiac tissue. In vitro investigations reveal that recombinant Mek1 binds two Cu atoms with high affinity and that Cu enhances Mek1 phosphorylation of Erk in a dose-dependent fashion. Coimmunopptn. expts. suggest that Cu is important for promoting the Mek1-Erk phys. interaction that precedes the phosphorylation of Erk by Mek1. These results demonstrate a role for Ctr1 and Cu in activating a pathway well known to play a key role in normal physiol. and in cancer.
    >> More from SciFinder ®
    https://chemport.cas.org/services/resolver?origin=ACS&resolution=options&coi=1%3ACAS%3A528%3ADC%252BC38XktlOlsrw%253D&md5=9f96dc863d3727d3b1f6d6bf8cc66530
  88. 88
    Brady, D. C.; Crowe, M. S.; Turski, M. L.; Hobbs, G. A.; Yao, X.; Chaikuad, A.; Knapp, S.; Xiao, K.; Campbell, S. L.; Thiele, D. J.; Counter, C. M. Copper Is Required for Oncogenic BRAF Signalling and Tumorigenesis. Nature 2014, 509, 492– 496,  DOI: 10.1038/nature13180