Systems Approaches for Unveiling the Mechanism of Action of

Dec 31, 2018 - Hongyan Li received her Ph.D. from the University of Edinburgh. She was a postdoctoral researcher at Hong Kong Polytechnic University p...
37 downloads 0 Views 7MB Size
Download PDF
Article Cite This: Acc. Chem. Res. 2019, 52, 216−227

pubs.acs.org/accounts

Systems Approaches for Unveiling the Mechanism of Action of Bismuth Drugs: New Medicinal Applications beyond Helicobacter Pylori Infection Hongyan Li, Runming Wang, and Hongzhe Sun*

Downloaded via IOWA STATE UNIV on January 16, 2019 at 10:20:06 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

Department of Chemistry, The University of Hong Kong, Pokfulam Road, Hong Kong, China CONSPECTUS: Metallodrugs have been widely used as diagnostic and therapeutic agents. Understanding their mechanisms of action may lead to advances in rational drug design. However, to achieve this, diversified approaches are required because of the complexity of metal−biomolecule interactions. Bismuth drugs in combination with antibiotics as a quadruple therapy show excellent success rates in the eradication of Helicobacter pylori, even for antibiotic-resistant strains, and in fact, they have been used in the clinic for decades for the treatment of infection. Understanding the mechanism of action of bismuth drugs may extend their medicinal application beyond the treatment of H. pylori infection. This Account describes several general strategies for mechanistic studies of metallodrugs, including system pharmacology and metalloproteomics approaches. The application of these approaches is exemplified using bismuth drugs. Through a system pharmacology approach, we showed that glutathione- and multidrug-resistance-associated protein 1-mediated self-propelled disposal of bismuth in human cells might explain the selective toxicity of bismuth drugs to H. pylori but not the human host. The development of metalloproteomics has enabled extensive studies of the putative protein targets of metallodrugs with a dynamic range of affinity. Continuous-flow GE-ICP-MS allows simultaneous monitoring of metals and their associated proteins with relatively high affinity on a proteome-wide scale. The fluorescence approach relies on unique Mn+−NTA-based fluorescence probes and is particularly applicable for mining those proteins that bind to metals/metallodrugs weakly or transiently. Integration of these methods with quantitative proteomics makes it possible to maximum coverage of bismuthassociated proteins, and the sustained efficacy of bismuth drugs lies in their ability to disrupt multiple biological pathways through binding and functional perturbation of key enzymes. The knowledge acquired by mechanistic studies of bismuth drugs led to the discovery of UreG as a new target for the development of urease inhibitors. The ability of Bi(III) to inhibit metallo-β-lactamase (MBL) activity through displacement of the Zn(II) cofactor renders bismuth drugs new potential as broad-spectrum inhibitors of MBLs. Therefore, bismuth drugs could be repurposed together with clinically used antibiotics as a cotherapy to cope with the current antimicrobial resistance crisis. We anticipate that the methodologies described in this Account are generally applicable for understanding the (patho)physiological roles of metals/metallodrugs. Our mechanism-guided discovery of new druggable targets as well as new medicinal applications of bismuth drugs will inspire researchers in relevant fields to engage in the rational design of drugs and reuse of existing drugs, eventually leading to the development of new effective therapeutics.

1. INTRODUCTION Bismuth compounds have been used in medicine for over three centuries for the treatment of various conditions, including syphilis, colitis, wound infection, and quartan malaria, but mostly for gastrointestinal disorders.1 Three bismuth drugs, namely, bismuth subsalicylate (Pepto-Bismol), colloidal bismuth subcitrate (CBS, De-Nol), and ranitidine bismuth citrate (Pylorid), have been used in the clinic to treat infection from Helicobacter pylori, which colonizes the stomachs of nearly half of the world’s population and is an etiological agent of chronic gastritis, peptic ulcers, and even stomach cancer.2 Despite the fact that H. pylori is susceptible to a range of antibiotics, the commonly used triple therapy based on a © 2018 American Chemical Society

proton pump inhibitor and antibiotics often leads to a low eradication rate of the bacterium, primarily because of the increasing prevalence of antibiotic resistance and changes in the epidemiology of the pathogen.3 Currently, bismuth-based triple therapy (tetracycline, metronidazole, and CBS) or quadruple therapy (a single capsule containing bismuth subcitrate potassium, metronidazole, tetracycline, and omeprazole) has been suggested as the first-line therapy.4−6 In addition to the existing bismuth drugs, a number of bismuth compounds with potent antimicrobial and anticancer activities Received: September 7, 2018 Published: December 31, 2018 216

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

Figure 1. (a, b) Importance of glutathione for bismuth detoxification and uptake in human cells: (a) L-BSO depleted glutathione from HK-2 cells; (b) L-BSO reduced the survival rate of HK-2 cells treated with CBS. (c, d) Glutathione- and MRP1 transporter-mediated bismuth compartmentalization. The location of bismuth was labeled using Bi-Tracer (blue fluorescence). (c) Confocal microscopic images of HK-2 cells treated with Bi-Tracer. (d) Confocal microscopy images of HK-2 treated with Bi-Tracer (50 μΜ for 36 h) and LysoTracker Red DND-99. (e) Schematic diagram depicting bismuth disposal in a human cell. Adapted from ref 16. Published by the National Academy of Sciences.

have been synthesized.1,7 However, none of these has been proved clinically. Bismuth has two major oxidation states, i.e., trivalent and pentavalent, with the trivalent state being the most common and stable form. The coordination number for bismuth varies from 3 up to 10 with diversified geometries. Bi(III) is known to have a high affinity toward sulfur, nitrogen, and oxygen ligands. In biological systems, Bi(III) preferentially coordinates to cysteine of a protein or peptide, followed by other residues including histidine, aspartic acid, and tyrosine.7,8 Comprehensive reviews regarding bismuth chemistry and biocoordination chemistry can be found elsewhere.1,7−9 The molecular mechanism of action of bismuth drugs is far from clear. Metallodrugs often exert their actions through binding and/or functional interference with metalloproteins/ metalloenzymes, thus disrupting multiple biological pathways. Extensive studies have revealed that H. pylori has developed complex systems to regulate metal homeostasis to ensure its survival, colonization, and persistence in the host.10 For example, H. pylori utilizes the host-specific transferrin/ lactoferrin to acquire iron,1,11 and the binding of Bi(III) to transferrin/lactoferrin9 may induce deprivation of iron in the pathogen. Indeed, X-ray structures of Bi-transferrin with Bi(III) bound to the N lobe and Fe(III) in the C lobe of transferrin12 shed light on the molecular mechanism. However, there is limited knowledge at the molecular level on how bismuth perturbs biological pathways that are critical for H. pylori. We have endeavored to unveil the molecular mechanism of action of bismuth drugs for the last two decades. Initially, we characterized the interactions of bismuth drugs with a variety of metalloenzymes/metalloproteins in vitro, in particular those that contain Ni(II)/Zn(II), given that Ni(II) is essential for the survival and pathogenicity of H. pylori because of the

special niche in which it resides. These studies were included in a review7 and a previous Account.8 Such studies are important in terms of understanding the biocoordination chemistry of Bi(III) but are unable to provide a holistic view of the mode of action of bismuth drugs. Recently, metalloproteomics has emerged as an invaluable tool to elucidate the molecular mechanisms of metallodrugs.13,14 There are still considerable challenges to track proteins that bind to metallodrugs or metals, particularly in live cells, as the interactions of metals/metallodrugs with proteins in vivo can be weak and even transient. In this Account, we summarize the most recent studies of the medicinal and biological chemistry of bismuth carried out mainly within last 10 years in our laboratory. We first introduce a system pharmacology approach to understand the low toxicity of bismuth drugs to the human host. We then describe several methodologies by which proteome-wide identification of metal- or metallodrug-binding proteins with diverse binding affinities can be achieved. By using bismuth drugs as an example, we demonstrate how the integrative metalloproteomic approaches can be employed to systemically identify putative protein targets of metallodrugs in pathogens and how these targets can be categorized by bioinformatics and subsequently validated by bioassays or in vitro characterization. Such studies provide a holistic view of the antimicrobial activity of bismuth drugs. The systems approach thus allows the discovery of new druggable targets (e.g., UreG) as well as new potential medicinal applications of bismuth drugs. 217

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

Figure 2. Schematic diagram illustrating the use of continuous-flow GE-ICP-MS for matching metals to proteins using Bi-binding proteins as an example. The spectrum at the right is reproduced with permission from ref 22. Copyright 2013 Wiley-VCH.

2. SYSTEM PHARMOCOLOGY APPROACH FOR UNDERSTANDING THE LOW TOXICITY OF Bi(III) TO THE HOST In contrast to other heavy-metal ions such as Pb(II) and Hg(II), Bi(III) exhibits low toxicity to humans but is highly toxic to pathogens. Low uptake of bismuth by the human gastrointestinal tract might account for its low toxicity. However, even when bismuth was repeatedly injected intramuscularly to treat syphilis, no severe toxicity to humans was found. Even at very extreme dosages or physical conditions, bismuth causes only reversible nephropathy or encephalopathy.15 We recently carried out a systemic study of the uptake and disposal of bismuth in human cells compared with bacterial cells.16 We discovered that glutathione (GSH) plays a critical role in bismuth uptake and detoxification in human cells. Depletion of GSH by supplementation of a γ-glutamylcysteine synthetase inhibitor, L-buthionine sulfoximine (L-BSO), which inhibits GSH synthesis in the first step, into HK-2 cells led to a reduced survival rate of the cells by 80% in comparison with that of cells without treatment (Figure 1a,b). In the meantime, L-BSO-treated cells developed severe cell degradation and nucleus malformation under stress with CBS, suggesting that GSH is crucial for detoxification of bismuth in human cells. It has been shown previously that over 90% of absorbed bismuth is metabolized into bismuth−sulfur particles, which are located in subcellular vesicles of many different types of mammalian cells.17 Using a homemade blue fluorescence probe, Bi-Tracer, we found bismuth inside subcellular vesicles in the perinuclear region of HK-2 cells and located inside the lysosomes (Figure 1c,d). GSH is suggested to be indispensable for bismuth compartmentalization. Interestingly, supplementation of MK571, an inhibitor of the multidrug-resistanceassociated protein 1 (MRP1), which transports a variety of metabolites or drugs out of cells, including glutathione conjugates, into CBS-treated HK-2 cells cultured in cystinesupplemented medium prevented the formation of the bismuth−sulfur particles, and the amounts of bismuth in the vesicles were reduced dramatically. Through systems studies, we found that bismuth was passively absorbed, conjugated to GSH, and then transported into vesicles via MRP1, where sequestration of absorbed bismuth consumed cytosolic glutathione and activated its de novo biosynthesis, which in turn facilitated passive uptake of bismuth (Figure 1e). Such a feedback cycle robustly removes bismuth from both intra- and extracellular space, thus protecting critical systems of the human body from acute toxicity. The introduction of system pharmacology approaches

for unraveling the mechanisms of action of metallodrugs should lead to advances in their design and use so as to avoid unwanted side effects.18

3. METALLOMICS/METALLOPROTEMICS FOR METALLODRUGS Systematic identification of metal- or metallodrug-binding proteins are particularly important toward understanding their roles in biology and medicine given that metals and metallodrugs often bind and functionally perturb the biological functions of (metallo)proteins.19−21 Because of the complexity and dynamic range of thermostability of the interactions of metal/metallodrugs with proteins/enzymes, different approaches have been developed, and the pros and cons of each method as well as their applicability are also exemplified and described in detail below. 3.1. GE-ICP-MS

A new strategy based on column-type gel electrophoresis (GE) coupled with a metal-specific detection system, e.g., inductively coupled plasma mass spectrometry (ICP-MS) was developed, i.e., continuous-flow GE-ICP-MS22 (Figure 2). The columntype gel was obtained by traditional slab gel preparation. Both native and denaturing gels could be prepared, and the gel composition could also be varied according to the protein targets of interest. A T-connection was used to split the eluate from the column gel system into two parts, one for online metal measurement by ICP-MS and the other for protein identification through biological mass spectrometry analysis. The method was employed to analyze bismuth-binding proteins from the cell lysates of H. pylori cultured in the presence of CBS. As shown in Figure 2, seven Bi-associated protein peaks were observed and identified on the basis of their peptide mass fingerprints. Among them, UreA/B, Ef-Tu, and AhpC (shown as TsaA in the figure) were also reported to bind bismuth.23,24 In vitro characterization of the interaction of bismuth with urease has also been made.20 Recently, it was also demonstrated that bismuth binds to AhpC and disrupts the peroxiredoxin and chaperone activity of the enzyme both in vitro and in bacterial cells.25 All of these results validate the reliability of the method in the identification of metallodrug- or metal-binding proteins under physiologically relevant conditions. In addition, GE-ICP-MS can also be used to investigate protein-associated metals in cells on a metallome-wide scale. For example, SlyD, a member of the FKBP family, binds to a range of metal ions via its histidine- and cysteine-rich Cterminus.26 Analysis of cell lysates of Escherichia coli overexpressing H. pylori SlyD (HpSlyD) by GE-ICP-MS revealed 218

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

Figure 3. (a) Molecular structures of fluorescence probes. (b) Schematic diagram illustrating the process of tracking the metal-associated proteome in live cells. (c) X-ray structure showing Fe3+-TRACER binding sites in the N lobe (left) and a zoom-in of the binding site (right). Residues that coordinate to Fe(III) or bind to TRACER (Asp63, Tyr95, Tyr188, and Lys206) are shown as sticks. (d) Confocal imaging of P. gingivalis bacterial cells upon incubation with Fe3+-TRACER. Live bacteria were stained as green by the LIVE/DEAD Baclight Bacterial Viability Kit. Panels (c) and (d) are reproduced with permission from ref 32. Copyright 2018 Royal Society of Chemistry.

3.2. Fluoresence-Based Approach

that HpSlyD preferentially binds Zn(II) followed by Cu(II), with 69% and 30% bound metal, respectively.26 Apart from studies of intracellular metal selectivity of metalloproteins, GEICP-MS also allows determination of the stoichiometry of metal binding to metalloproteins in vivo by the use of sulfur signals to quantify protein concentrations.27 GE-ICP-MS serves as a robust approach to allow not only tracking of metal/metallodrug-associated proteins on a proteome-wide scale but also investigation of metalloproteinassociated metals on a metallome-wide scale. The intracellular metals/metallodrugs binding proteins identified by this approach are intrinsic. However, this method is not applicable to track proteins that bind metals/metallodrugs with low affinity because of dissociation of the metals from the proteins. In addition, proteins with low abundance also might not be able to be identified unless extensive enrichment procedures are used. Its lower resolution in the protein separation can be improved by using two-dimensional gels, so that extensive coverage of metallodrug-associated proteins can be achieved. This work is currently underway in our laboratory.

Small-molecule-based fluorescent probes have been widely used for real-time visualization of proteins in live cells, which allows tracking of various cellular events with minimal perturbation of the cells being investigated.28 We have developed membrane-permeable Ni2+−NTA-based fluorescent probes by conjugation of a nitrilotriacetate (NTA) moiety with a fluorophore and an aryl azide followed by coordination with Ni2+ ions. The probes rapidly enter various live cells and plant tissues and bind specifically to a His6 tag genetically fused to a protein of interest, leading to over 13-fold fluorescence enhancement.29 A family of probes (Figure 3a) were prepared, including a coumarin-based blue-fluorescent probe (Ni2+NTA-AC),29 a BODIPY-based red-fluorescent probe (Ni2+NTA-AB),30 and a fluorescein-based green-fluorescent probe (Ni2+-NTA-AF) (unpublished). The blue-fluorescent probe Ni2+-NTA-AC has been successfully used to image His6-tagged proteins in mammalian cells and plant tissues,29 while the redfluorescent probe Ni2+-NTA-AB was utilized to visualize the proteins in live bacterial cells with the help of a surfactant.30 Incorporation of the aryl azide is essential to strengthen the 219

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

Figure 4. Cysteine- and histidine-oriented binding preference of Bi(III) in H. pylori. (a) Changes in relative abundance of various amino acids in Bi-binding motifs compared with that of the whole H. pylori proteome. (b) Amino acid sequence logos of Bi-associated CXXC, CX7C, HXXH, and HX8H sequence motifs. (c) Mapping of putative drug targets to the phenotypic features of bismuth-treated H. pylori. Copyright 2015 Royal Society of Chemistry.

approach because of either weak or transient binding or low resolution of one-dimensional GE-ICP-MS in protein separation. Certain proteins identified by this method, such as DnaK and UreG, have also been validated (vide infra). To explore the nature of the binding of the probe with proteins, using Fe3+-TRACER and human transferrin (hTF) as a showcase,32 we solved the structure of Fe3+-TRACER bound to hTF (Fe3+-TRACER−hTF) at 2.85 Å resolution. Surprisingly, the NTA moiety of the probe coordinates to Fe(III) by only two carboxylates and one nitrogen, and only one tyrosine residue (Tyr188) in the specific iron-binding site coordinates to Fe(III) (Figure 3c). The binding of the probe to hTF is further stabilized via hydrogen bonding with Lys206, Asp63, and Tyr95 (Figure 3c). The Fe3+-TRACER molecule folds like a sandwich so that the NTA group flips back onto the coumarin moiety by utilizing the flexible (CH2)4 linker. Several Fe-binding proteins in Porphyromonas gingivalis were identified32 (Figure 3d). The fluorescence-based approach has advantages in tracking endogenous metalloproteins that bind metal ions transiently or weakly. However, caution must be taken in analysis of these proteins, as the probe only serves as a “reporter” and some of the proteins might not be intrinsic. Thus, subsequent validation is necessary to rule out false information, which can be achieved through in vitro or in vivo characterization of

binding between the probe and proteins of interest as well as for fluorescence turn-on effects. The use of this family of probes was further extended to track metal-binding proteins. As shown in Figure 3b, the probes, after entering live cells, bind to proteins via the Mn+−NTA moiety. Upon UV exposure, the probe is anchored to the labeled proteins through the formation of a covalent bond between them, and the labeled proteins can be subsequently identified by conventional proteomics. Metal ions that have relatively high affinity toward NTA can also be incorporated to yield “metaltunable” probes, which are applicable to tracking of various metalloproteomes and mining of potential targets of metals or metallodrugs on a proteome-wide scale. Several coumarin-based metal-tunable fluorescence probes, denoted as Mn+-TRACER, have been employed to track different metalloproteomes in live cells.31,32 About 44 proteins in the cytosolic fraction of H. pylori were identified using Ni2+TRACER, with 11 of them, including HpUreB, HpHspA, and HpUreG, also being identified previously by immobilized metal affinity chromatography (IMAC).24 These proteins indeed play important roles in nickel homeostasis.33 Similarly, a total of 46 bismuth-binding proteins in H. pylori were identified using Bi3+-TRACER.34 Nine of these proteins were also identified by GE-ICP-MS, including UreA/B, HspA, tuf, and tsaA. The rest were tracked only by the fluorescence-based 220

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

Figure 5. (a) Network depicting Bi-influenced protein−protein interaction (BiPI) in H. pylori. Proteins are colored and shaped according to their different properties in the network. (b) UV−vis difference spectra of HpDnaK upon addition of different molar equivalents of Bi-NTA. The inset shows the titration curve plotted at 360 nm against different molar ratios of [Bi-NTA]/[DnaK]. (c) Protein thermal melting curves showing the target engagement of Bi(III) on DnaK in H. pylori intact cells. (d) ATPase activities of HpDnaK in the absence and presence of different molar ratios of CBS. (e) Glycolipid binding specificity of the N- terminal domain of HpDnaK in the absence and presence of different molar ratios of CBS. (f) Schematic diagram showing the multitargeted mode of action of CBS in eradicating H. pylori. Adapted from ref 34. Copyright 2017 Royal Society of Chemistry.

Bi(III). Both Asp and Glu exhibited high popularity in the Bibinding peptides, consistent with the high acidic property of the metal ion. Analysis of major Bi-binding motifs, e.g., CXnC, CXnH, and HXnH, reveals that Bi(III) commonly coordinates to CXnC (n = 2, 7), CXnH (n = 5, 6) and HXnH (n = 0−2, 8) (Figure 4b). In addition to identification of Bi(III)-binding motifs, Bi-IMAC also enables tracking of a large quantity of Bibinding proteins, allowing subsequent bioinformatics analysis to uncover insights into the mode of action of bismuth drugs (Figure 4c). The advantage of Bi-IMAC on-column digestion coupled with high-throughput LC−MS (or nano-LC−MS) lies in generating information-rich resources on bismuth−protein binding interfaces, providing a basis for exploration of Bi(III)-binding motifs. Moreover, the pre-enrichment capability of IMAC also allows tracking of low-abundance proteins that bind to the metallodrug. However, Bi-IMAC may also generate false information due to the binding of Bi(III) to surface-exposed residues rather than those in the active sites of enzymes. Thus, subsequent validation of the identified proteins is also needed, which can be done by in vivo or in vitro

the interactions of relevant metal ions with the identified proteins. For instance, HspB protein, which does not bind to Ni(II), was identified from H. pylori by Ni2+-TRACER. This is likely due to the fact that HspA is a Ni(II)-binding protein, which forms a complex with HspB.31 3.3. Bi-IMAC

IMAC is a useful method to mine putative metalloproteins from the complex proteomes and to identify sequence motifs relating to the cellular functional behaviors of metals.35 BiIMAC in combination with high-throughput liquid chromatography−mass spectrometry (LC−MS) enabled Bi(III)-binding proteins and peptides to be mined in H. pylori.36 In total, over 300 Bi(III)-binding peptides were identified from 166 proteins in H. pylori. Analysis of the amino acid sequences of these peptides revealed that bismuth exhibited the highest preference toward peptides containing cysteines (+274%), followed by those containing histidines (+79%) and acidic amino acids (Glu, +38%; Asp, +15%) (Figure 4a). Preferential binding of the thiolate (Sγ) of cysteines and the imidazolium nitrogens (Nδ and Nε) of histidines reflects the soft feature of 221

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

Figure 6. (a) Two-dimensional electrophoresis showing the Bi-binding proteins in live H. pylori cells. (b) Bar chart showing the activity of UreG as a GTPase. (c) Effects of metal−UreG interactions on the urease activity. Apo or metal-treated UreG was supplemented to the cell lysate of E. coli expressing the urease gene cluster, excluding the ureG gene, prior to analysis. (d) Recovery of urease activity upon supplementation of additional Ni-UreG, CBS-treated (2 molar equiv) Ni-UreG (Bi-UreG), and Ni ions alone into CBS-treated H. pylori cell lysate. (e) Urease activity of H. pylori with or without supplementation of CBS to the bacterial culture and to the extracted enzymes. (f) Structures of the representative hits, compounds 4 and 8. (g) Effect of compounds 4 and 8 on the urease activity of H. pylori. Adapted from ref 44.

binding and Bi-regulated proteins.34 A total of 63 Bi-binding proteins and 119 Bi-regulated proteins were identified from H. pylori, providing rich sources for in-depth analysis of biological pathways disrupted by bismuth drugs. Subsequently, Bi-associated proteins (i.e., Bi-binding and Biregulated proteins) were categorized by Gene Ontology (GO) enrichment analysis, and a number of protein−protein interaction (PPI) subnetworks (e.g., the tricarboxylic acid cycle, cell redox homeostasis, nickel homeostasis, protein folding, and iron homeostasis) were identified, providing evidence for multiple action of bismuth drugs. Moreover,

examination of whether the identified proteins bind to the metal ions.

4. MULTITARGETED MODE OF ACTION REVEALED BY THE SYSTEMICS APPROACH ACCOUNTS FOR THE SUSTAINED EFFICACY OF BISMUTH DRUGS AGAINST H. pylori Considering that each methodology has its limitations, we have recently integrated in-house metalloproteomics together with quantitative proteomics, allowing maximum discovery of Bi222

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

vitro studies showed that bismuth exhibited trivial effects on urease activity even at high concentration (up to several mM),23 which is unlikely to account for its in vivo activity. Given the special niches in which it resides, H. pylori requires urease for its survival and pathogenesis. The enzyme catalyzes the hydrolysis of urea in many pathogenic bacteria into ammonia and carbonic acid and has been considered as one of the most important targets in the development of antimicrobial drugs, especially for the treatment of gastric and urinary infections.39,40 The crystal structure of H. pylori urease reveals a spherical assembly of 12 catalytic units with a deeply buried dimeric nickel center and a high specificity of the substrate.41 This makes it very challenging to develop urease inhibitors by either targeting the metallocenter or mimicking the substrate of ureases.39,42 To date, only acetohydroxamic acid (AHA) is used clinically for the treatment of urinary tract infections together with antibiotics,42,43 but with adverse side effects. Previously we showed that binding of Bi(III) to urease results in trivial effects on urease activity in vitro.23 We recently demonstrated that Bi(III) inhibits urease indirectly, enabling the identification of a new druggable target, and a new class of urease inhibitors based on this target were also derived.44 As apo urease synthesized in microbes is inactive, urease exerts its function only if nickel is inserted into its active site.45 The assembly of the active site of bacterial ureases is a GTPdependent process involving the cooperative actions of several accessory proteins, such as UreE, UreF, UreG, and UreH.33,46 Homologues of UreF, UreG, and UreH/D are also found in urease-producing eukaryotes,45 indicating that the maturation of urease is relatively conserved among different species. By the use of homemade Bi3+-TRACER in combination with other biochemical techniques, UreG, a metallochaperone and GTPase, was found to be the only accessary protein that binds Bi(III) both in vitro and in vivo (Figure 6a), with an apparent dissociation constant (Kd) of 3.1 × 10−24 M. Such binding disrupts the formation of protein−protein complexes such as UreE−UreG and UreG-UreFH, which are essential for maturation of urease.46,47 Importantly, Bi(III) interacts with UreG at the Ni(II) binding site via Cys66, leading to Ni(II) release, which abolishes the GTPase activity of UreG, subsequently inhibiting the urease activity (Figure 6b,c). Consistently, supplementation of CBS to the culture medium during H. pylori growth resulted in a dramatic decline in the ureolytic activity of H. pylori urease in live bacterial cells. In contrast, incubation of CBS (up to 400 μM) with extracted H. pylori urease resulted in little inhibitory effect (Figure 6d), suggesting that Bi(III) disrupts the urease maturation process through binding to and functional disruption of UreG. This was further confirmed by the recovery of urease activity upon the addition of Ni-bound UreG into CBS-treated H. pylori cell lysate (Figure 6e). Such phenomena were also observed in other pathogenic bacteria.44 Therefore, UreG might serve as an alternative target for the design of potent urease inhibitors. With UreG from H. pylori as a showcase, by virtual screening and experimental validation, two compounds have been demonstrated to bind and functionally perturb UreG (Figure 6f), thus inhibiting urease activity in both H. pylori cell culture and a mammalian cell infection model, with IC50 values at micromolar levels, which are lower than that of AHA (Figure 6g). Although further optimization of the structures of these small molecules is needed to improve their potency and specificity toward bacterial UreG instead of human GTPases, this study opens a new opportunity for devising more effective

proteins possessing oxidoreductase activity and hydrolase activity were significantly influenced by CBS. By mapping of the bismuth-associated proteins on the protein−protein networks, a giant Bi-influenced protein interaction (BiPI) network in H. pylori was obtained (Figure 5a). The BiPI exhibits a scale-free topology with a small number of highly connected proteins participating in processes and functioning as core proteins, while most of the other proteins have only a few connections. Five central nodes are highlighted in Figure 5a: four of them are Bi-binding proteins, i.e., UreA, UreB, RpoA, and DnaK (a major heat shock protein), while NusG is downregulated by bismuth drugs. Interestingly, about 63 Bitargeting proteins could be grouped not only as highly connected hub proteins but also as bottleneck proteins in the network. Such a unique mode of action of a bismuth drug may explain its lethality and lower likelihood to develop resistance, as targeting a highly connected hub protein is more likely to be lethal to an organism37 also is less likely to develop drug resistance.38 Bioinformatics allows the possible targets or pathways that are perturbed by bismuth drugs to be derived from large quantities of data.34 As DnaK was identified to be a potential protein target and is also one of the central node proteins, in vitro characterization of the interaction of CBS with recombinant DnaK showed that Bi(III) indeed binds DnaK with a [Bi-NTA]:[HpDnaK] stoichiometry of 0.5:1 (Figure 5b), and subsequently, binding of DnaK with bismuth was also demonstrated in bacterial cells (Figure 5c). Importantly, Bi binding reduced the ATPase activity (Figure 5d), chaperone activity, and glycolipid (cerebroside sulfate) binding affinity of DnaK (Figure 5e). As glycolipid binding is a common feature of bacterium−host cell adhesion, it is likely that bismuth drugs may disrupt the interaction between H. pylori and its host gastric epithelial cells. The key biological pathways in H. pylori disrupted by bismuth drugs were validated, including the oxidative stress defense systems as well as the pH buffering system. Supplementation of CBS into the cell culture medium significantly retarded the growth of H. pylori under various oxidative stresses. The activities of several major antioxidant enzymes from H. pylori that bind to CBS, i.e., AhpC and KatA, were inhibited by 73% and 15%, respectively, by CBS. Consistent with this, the levels of reactive oxygen species (ROS) in H. pylori were elevated upon treatment with CBS. The bacterial pH buffering system was also abolished by CBS, causing bacterial growth to be inhibited. The urease activities were reduced by 40% upon CBS treatment under both acidic and neutral conditions. On the basis of this study, it is concluded that the sustainable antimicrobial activity of bismuth drugs against H. pylori and the low likelihood of H. pylori to develop resistance to bismuth drugs were attributed to the multitargeted mode of action of bismuth drugs (Figure 5f). The integrative approach might serve as a general platform for understanding the physiological and pathological roles of metals or metallodrugs.34

5. UreG SERVES AS A NEW TARGET FOR THE DEVELOPMENT OF UREASE INHIBITORS Despite the fact that the urease subdomains UreA and UreB were tracked to bind to Bi(III), it is unclear whether such binding contributes to the decreased urease activity upon treatment of H. pylori cells with CBS. However, our previous in 223

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

Figure 7. (a) Profiles for inhibition of the MBLs NDM-1, VIM-2, and IMP-4 by CBS. (b) Isobolograms of the combination of MER and CBS against different MBL-positive bacterial strains. (c) Survival curves showing efficacies in a murine peritonitis infection model. (d−f) X-ray structures showing (d) the superimposition of Bi(III)-bound NDM-1 (cyan) with native Zn-bound NDM-1 (orange), (e) the active site of native Zn-bound NDM-1, and (f) the active site of Bi-bound NDM-1. Adapted from ref 49.

(NDMs), are Zn(II)-containing enzymes that activate a nucleophilic water to cleave the β-lactam ring, conferring bacterial resistance to the last-resort carbapenems and all other bicyclic β-lactams used currently.50 Combination therapy comprising clinically used antibiotics and an inhibitor of metallo-β-lactams appears to be a more economical and effective alternative to treat infection from these so-called “superbugs”. Currently, most organic-molecule-based MBL inhibitors are designed by mimicking the substrates of specific types of MBLs51,52 and are unlikely to be developed as broadspectrum inhibitors toward all MBLs given the structural diversity of the enzyme active sites. Moreover, these inhibitors may readily encounter microbial resistance due to rapid evolution of MBLs. To date, no such inhibitors have been approved for clinical use. NDM-1 as well as other B1-type MBLs are dizinc-containing enzymes, with one Zn(II) coordinating to three histidines and another Zn(II) to a cysteine, a histidine, and an aspartic acid, with a water molecule bridging between the two Zn(II) ions (Figure 7e). Given the thiolphilic feature of Bi(III), bismuth compounds may interact with the zinc center of the enzyme. Indeed, dose-dependent inhibition of the enzymatic activities

urease inhibitors by targeting the metallochaperone UreG and is also an excellent example of mechanism-based discovery of new druggable targets.

6. NEW MEDICINAL APPLICATIONS OF BISMUTH Bismuth drugs exert their action through binding to the key enzymes, subsequently disrupting key pathological pathways in the pathogen. In particular, enzymes processing cysteine and histidine residues in the active site are likely to be targeted by Bi(III). The discovery of such enzymes may render new medicinal applications of bismuth compounds. Indeed, the antiviral activity of bismuth complexes against severe acute respiratory syndrome coronavirus (SARS-CoV) was attributed to the ability of Bi(III) to inhibit the unwinding activity of the SARS-CoV helicase, an enzyme containing a cysteine-rich Zn(II)-binding domain, which blocks virus replication.48 Our recent report demonstrated that bismuth drugs or bismuth compounds can be repurposed or developed as broadspectrum inhibitors of metallo-β-lactamases (MBLs) regardless of their type or subtype.49 MBLs, including imipenemases (IMPs), Verona integron-encoded metallo-β-lactamases (VIMs), and more recently New Delhi metallo-β-lactamases 224

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research

Among the bismuth-binding proteins that we identified, only a few have been validated (e.g., DnaK), and there are still many more potential targets that have not be investigated, such as RpoA and NusG, which are hub proteins in the BiPI network. Targeting these hub proteins is likely to be lethal for the pathogen, but whether these can be used as targets to design more potent antimicrobial drugs to treat H. pylori infection remains to be explored in the future. Clearly, the methodologies described in this Account provide a general platform for understanding the physiological and pathological roles of metals/metallodrugs or metalloagents, such as Os(III)- and Ir(III)-containing complexes, as their protein targets in biological systems are largely unknown. In addition to metalloproteomics, integration of multiomics, including transcriptomics and metabolomics, may facilitate uncovering the full system response profile of cells toward the stress of a drug.53 Repurposing a bismuth drug to overcome MBL-producing antibiotic resistance might shed light on the use of metals such as silver54 (and silver nanoparticles), copper, and gallium to cope with the current antimicrobial resistance crisis, which poses huge threats to public health globally. However, extensive research is needed to resolve the toxicity of silver in internal antimicrobial drugs, to improve the potency of metalloantimicriobal agents, and to fine-tune the compositions of cocktail therapies consisting of metalloagents and antibiotics prior to clinical use.

of NDM-1, VIM-2, and IMP-4 by CBS using meropenem (MER) as a substrate was observed, with IC50 values of 2.81, 3.54, and 0.70 μM, respectively (Figure 7a). The inhibition of MBLs by CBS can resensitize NDM-1-positive bacteria toward β-lactam antibiotics (e.g., MER), whereas CBS itself shows no or minor growth inhibition toward either NDM-1-positive or -negative bacteria. The synergy between CBS and MER was also noted with the fractional inhibitory concentration index (FICI) of 0.250 (Figure 7b). Such a synergism pattern was also observed in a panel of NDM-1-positive enterobacterial strains, i.e., E. coli, Klebsiella pneumonia, and Citrobacter freundii. Apart from the clinically used bismuth drug CBS, a series of bismuth compounds were found to exhibit more or less similar activities toward inhibition of NDM-1, VIM-2, or IMP-4. Importantly, the combination of CBS with MER effectively postponed the death of mice infected by NDM-1 -positive bacteria and led to an increase in the mice survival rate to 50%, while CBS failed to protect any of the mice from death and MER alone rescued only two of the eight mice (Figure 7c). Subsequently, we showed that Bi(III) can displace the two Zn(II) ions from NDM-1 with one Bi(III) and that the cysteine residue in the active site (Cys208) is critical for Bi(III) binding, as mutation of this residue to alanine resulted no binding either in vitro or in vivo. Importantly, binding of Bi(III) to NDM-1 is irreversible since the activity of NDM-1 could not be restored once bismuth was bound to the enzyme. Moreover, the interaction of CBS with NDM-1 enzyme also suppressed the evolution of NDM-1-positive bacteria. The Xray structure of Bi(III)−NDM-1 confirms that one Bi(III) ion replaces the two Zn(II) (Figure 7d,f). Interestingly, Bi(III) is located between the two Zn(II) ion sites, slightly closer to the Zn1 site in the native NDM-1 structure in the major conformation, coordinating to His120, H189, Cys208, Asp124, and a water molecule, forming a distorted trigonalprismatic geometry. In contrast to organic MBL inhibitors, Bi(III) compounds inhibit the enzyme through replacement of Zn(II) ion by Bi(III) ion via the critical residue of cysteine in the active site. Such an unique mechanism renders Bi(III) compounds as broad-spectrum inhibitors for the B1 and B2 classes of MBLs. In view of the use of certain bismuth compounds in the clinic for decades and their selective toxicity toward pathogens, such broad-spectrum MBL inhibitors together with antibiotics as a cotherapy will undoubtedly open a new horizon for the treatment of infection caused by MBL-positive bacteria.

■

AUTHOR INFORMATION

ORCID

Hongzhe Sun: 0000-0001-6697-6899 Notes

The authors declare no competing financial interest. Biographies Hongyan Li received her Ph.D. from the University of Edinburgh. She was a postdoctoral researcher at Hong Kong Polytechnic University prior to joining the University of Hong Kong. Her research interest is focused on the structure and function of metalloproteins. Runming Wang received his Ph.D. from the University of Hong Kong in 2018. He is now working as a postdoctoral researcher, and his research interests lie in the areas of metallodrug discovery and interaction of metals with proteins.

7. CONCLUSION AND OUTLOOK We have described several approaches, including system pharmacology and metalloproteomics, to understand the mode of action of bismuth drugs. The sustainable antimicrobial activity of bismuth drugs against H. pylori and low likelihood of resistance to bismuth are attributable to the multitargeted modes of action. On the basis of the knowledge acquired from Bi(III) inhibition of urease, UreG has been discovered as a new target for the development of urease inhibitors. Moreover, the thiolphilic feature of Bi(III) enables it to inhibit bacterial enzymes such as MBLs that contain cysteine residues in their active sites. Indeed, Bi(III) can replace Zn(II) from the active site of MBLs through binding to the critical cysteine residue, leading to inactivation of the enzyme. Such a unique mode of action makes it possible to develop bismuth compounds as the first broad-spectrum inhibitors at least for B1 MBLs.

Hongzhe Sun is a Norman and Cecilia Yip Professor at the University of Hong Kong. He has been working on the chemical biology of metals, particularly bismuth antimicrobial agents, as well as on metalloproteomics for metals in biology.

■

ACKNOWLEDGMENTS We thank the Research Grants Council (17305415, 17333616, and 17307017), the Innovation Technology Commission of Hong Kong SAR (ITS/085/14 and ITS/124/17), Livzon Pharmaceutical Group, and the Norman and Cecilia Yip Fund for support.

■

REFERENCES

(1) Sun, H. Biological Chemistry of Arsenic, Antimony and Bismuth; Wiley: Chichester, U.K., 2011.

225

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research (2) Fock, K. M.; Graham, D. Y.; Malfertheiner, P. Helicobacter pylori research: historical insights and future directions. Nat. Rev. Gastroenterol. Hepatol. 2013, 10, 495−500. (3) De Francesco, V.; Giorgio, F.; Hassan, C.; Manes, G.; Vannella, L.; Panella, C.; Ierardi, E.; Zullo, A. Worldwide H. pylori antibiotic resistance: a systematic review. J. Gastrointest. Liver Dis. 2010, 19, 409−414. (4) Laine, L.; Hunt, R.; El-Zimaity, H.; Nguyen, B.; Osato, M.; Spenard, J. Bismuth-based quadruple therapy using a single capsule of bismuth biskalcitrate, metronidazole, and tetracycline given with omeprazole versus omeprazole, amoxicillin, and clarithromycin for eradication of Helicobacter pylori in duodenal ulcer patients: A prospective, randomized, multicenter, North American trial. Am. J. Gastroenterol. 2003, 98, 562−567. (5) Dore, M. P.; Lu, H.; Graham, D. Y. Role of bismuth in improving Helicobacter pylori eradication with triple therapy. Gut 2016, 65, 870− 878. (6) Megraud, F. The challenge of Helicobacter pylori resistance to antibiotics: the comeback of bismuth-based quadruple therapy. Ther. Adv. Gastroenterol. 2012, 5, 103−109. (7) Li, H.; Sun, H. Recent advances in bioinorganic chemistry of bismuth. Curr. Opin. Chem. Biol. 2012, 16, 74−83. (8) Ge, R.; Sun, H. Bioinorganic chemistry of bismuth and antimony: target sites of metallodrugs. Acc. Chem. Res. 2007, 40, 267−274. (9) Yang, N.; Sun, H. Biocoordination chemistry of bismuth: recent advances. Coord. Chem. Rev. 2007, 251, 2354−2366. (10) Haley, K. P.; Gaddy, J. A. Metalloregulation of Helicobacter pylori physiology and pathogenesis. Front. Microbiol. 2015, 6, 911. (11) Senkovich, O.; Ceaser, S.; McGee, D. J.; Testerman, T. L. Unique host iron utilization mechanisms of Helicobacter pylori revealed with iron-deficient chemically defined media. Infect. Immun. 2010, 78, 1841−1849. (12) Yang, N.; Zhang, H.; Wang, M.; Hao, Q.; Sun, H. Iron and bismuth bound human serum transferrin reveals a partially-opened conformation in the N-lobe. Sci. Rep. 2012, 2, 999. (13) Wang, Y.; Wang, H.; Li, H.; Sun, H. Metallomic and metalloproteomic strategies in elucidating the molecular mechanisms of metallodrugs. Dalton Trans. 2015, 44, 437−447. (14) Sun, X.; Tsang, C.; Sun, H. Identification and characterization of metallodrug binding proteins by (metallo)proteomics. Metallomics 2009, 1, 25−31. (15) Leussink, B. T.; Slikkerveer, A.; Engelbrecht, M. R.; van der Voet, G. B.; Nouwen, E. J.; de Heer, E.; de Broe, M. E.; de Wolff, F. A.; Bruijn, J. A. Bismuth overdosing-induced reversible nephropathy in rats. Arch. Toxicol. 2001, 74, 745−754. (16) Hong, Y.; Lai, Y. T.; Chan, G. C.; Sun, H. Glutathione and multidrug resistance protein transporter mediate a self-propelled disposal of bismuth in human cells. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 3211−3216. (17) Stoltenberg, M.; Juhl, S.; Danscher, G. Bismuth ions are metabolized into autometallographic traceable bismuth−sulphur quantum dots. Eur. J. Histochem. 2007, 51, 53−57. (18) Romero-Canelon, I.; Sadler, P. J. Systems approach to metalbased pharmacology. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 4187− 4188. (19) Xia, W.; Li, H.; Sun, H. Functional disruption of HypB, a GTPase of Helicobacter pylori by bismuth. Chem. Commun. 2014, 50, 1611−1614. (20) Cun, S.; Sun, H. A zinc-binding site by negative selection induces metallodrug susceptibility in an essential chaperonin. Proc. Natl. Acad. Sci. U. S. A. 2010, 107, 4943−4948. (21) Meier, S. M.; Kreutz, D.; Winter, L.; Klose, M. H. M.; Cseh, K.; Weiss, T.; Bileck, A.; Alte, B.; Mader, J. C.; Jana, S.; Chatterjee, A.; Bhattacharyya, A.; Hejl, M.; Jakupec, M. A.; Heffeter, P.; Berger, W.; Hartinger, C. G.; Keppler, B. K.; Wiche, G.; Gerner, C. An organoruthenium anticancer agent shows unexpected target selectivity for plectin. Angew. Chem., Int. Ed. 2017, 56, 8267−8271.

(22) Hu, L.; Cheng, T.; He, B.; Li, L.; Wang, Y.; Lai, Y. T.; Jiang, G.; Sun, H. Identification of metal-associated proteins in cells by using continuous-flow gel electrophoresis and inductively coupled plasma mass spectrometry. Angew. Chem., Int. Ed. 2013, 52, 4916−4920. (23) Zhang, L.; Mulrooney, S. B.; Leung, A. F.; Zeng, Y.; Ko, B. B.; Hausinger, R. P.; Sun, H. Inhibition of urease by bismuth(III): implications for the mechanism of action of bismuth drugs. BioMetals 2006, 19, 503−511. (24) Ge, R.; Sun, X.; Gu, Q.; Watt, R. M.; Tanner, J. A.; Wong, B. C.; Xia, H. H.; Huang, J.; He, Q.; Sun, H. A proteomic approach for the identification of bismuth-binding proteins in Helicobacter pylori. JBIC, J. Biol. Inorg. Chem. 2007, 12, 831−842. (25) Chang, Y. Y.; Cheng, T.; Yang, X.; Jin, L.; Sun, H.; Li, H. Functional disruption of peroxiredoxin by bismuth antiulcer drugs attenuates Helicobacter pylori survival. JBIC, J. Biol. Inorg. Chem. 2017, 22, 673−683. (26) Cheng, T.; Li, H.; Xia, W.; Sun, H. Multifaceted SlyD from Helicobacter pylori implication in [NiFe] hydrogenase maturation. JBIC, J. Biol. Inorg. Chem. 2012, 17, 331−343. (27) Wang, Y.; Hu, L.; Yang, X.; Chang, Y. Y.; Hu, X.; Li, H.; Sun, H. On-line coupling of continuous-flow gel electrophoresis with inductively coupled plasma-mass spectrometry to quantitatively evaluate intracellular metal binding properties of metallochaperones HpHypA and HpHspA in E. coli cells. Metallomics 2015, 7, 1399− 1406. (28) Marks, K. M.; Nolan, G. P. Chemical labeling strategies for cell biology. Nat. Methods 2006, 3, 591−596. (29) Lai, Y. T.; Chang, Y. Y.; Hu, L.; Yang, Y.; Chao, A.; Du, Z. Y.; Tanner, J. A.; Chye, M. L.; Qian, C.; Ng, K. M.; Li, H.; Sun, H. Rapid labeling of intracellular His-tagged proteins in living cells. Proc. Natl. Acad. Sci. U. S. A. 2015, 112, 2948−2953. (30) Chao, A.; Jiang, N.; Yang, Y.; Li, H.; Sun, H. A Ni-NTA-based red fluorescence probe for protein labelling in live cells. J. Mater. Chem. B 2017, 5, 1166−1173. (31) Lai, Y. T.; Yang, Y.; Hu, L.; Cheng, T.; Chang, Y. Y.; KoohiMoghadam, M.; Wang, Y.; Xia, J.; Wang, J.; Li, H.; Sun, H. Integration of fluorescence imaging with proteomics enables visualization and identification of metallo-proteomes in living cells. Metallomics 2017, 9, 38−47. (32) Jiang, N.; Cheng, T. F.; Wang, M. J.; Chan, G. C. F.; Jin, L. J.; Li, H. Y.; Sun, H. Z. Tracking iron-associated proteomes in pathogens by a fluorescence approach. Metallomics 2018, 10, 77−82. (33) Cheng, T.; Li, H.; Xia, W.; Jin, L.; Sun, H. Exploration into the nickel ’microcosmos’ in prokaryotes. Coord. Chem. Rev. 2016, 311, 24−37. (34) Wang, Y.; Hu, L.; Xu, F.; Quan, Q.; Lai, Y. T.; Xia, W.; Yang, Y.; Chang, Y. Y.; Yang, X.; Chai, Z.; Wang, J.; Chu, I. K.; Li, H.; Sun, H. Integrative approach for the analysis of the proteome-wide response to bismuth drugs in Helicobacter pylori. Chem. Sci. 2017, 8, 4626−4633. (35) Porath, J.; Carlsson, J.; Olsson, I.; Belfrage, G. Metal chelate affinity chromatography, a new approach to protein fractionation. Nature 1975, 258, 598−599. (36) Wang, Y.; Tsang, C. N.; Xu, F.; Kong, P. W.; Hu, L.; Wang, J.; Chu, I. K.; Li, H.; Sun, H. Bio-coordination of bismuth in Helicobacter pylori revealed by immobilized metal affinity chromatography. Chem. Commun. 2015, 51, 16479−16482. (37) Jeong, H.; Mason, S. P.; Barabasi, A. L.; Oltvai, Z. N. Lethality and centrality in protein networks. Nature 2001, 411, 41−42. (38) Fraser, H. B.; Hirsh, A. E.; Steinmetz, L. M.; Scharfe, C.; Feldman, M. W. Evolutionary rate in the protein interaction network. Science 2002, 296, 750−752. (39) Follmer, C. Ureases as a target for the treatment of gastric and urinary infections. J. Clin. Pathol. 2010, 63, 424−430. (40) Flores-Mireles, A. L.; Walker, J. N.; Caparon, M.; Hultgren, S. J. Urinary tract infections: epidemiology, mechanisms of infection and treatment options. Nat. Rev. Microbiol. 2015, 13, 269−284. 226

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227

Article

Accounts of Chemical Research (41) Ha, N. C.; Oh, S. T.; Sung, J. Y.; Cha, K. A.; Lee, M. H.; Oh, B. H. Supramolecular assembly and acid resistance of Helicobacter pylori urease. Nat. Struct. Biol. 2001, 8, 505−509. (42) Kosikowska, P.; Berlicki, L. Urease inhibitors as potential drugs for gastric and urinary tract infections: a patent review. Expert Opin. Ther. Pat. 2011, 21, 945−957. (43) Fishbein, W. N.; Carbone, P. P.; Hochstein, H. D. Acetohydroxamate: bacterial urease inhibitor with therapeutic potential in hyperammonaemic states. Nature 1965, 208, 46−48. (44) Yang, X.; Koohi-Moghadam, M.; Wang, R.; Chang, Y. Y.; Woo, P. C. Y.; Wang, J.; Li, H.; Sun, H. Metallochaperone UreG serves as a new target for design of urease inhibitor: a novel strategy for development of antimicrobials. PLoS Biol. 2018, 16, No. e2003887. (45) Farrugia, M. A.; Macomber, L.; Hausinger, R. P. Biosynthesis of the urease metallocenter. J. Biol. Chem. 2013, 288, 13178−13185. (46) Fong, Y. H.; Wong, H. C.; Chuck, C. P.; Chen, Y. W.; Sun, H.; Wong, K. B. Assembly of preactivation complex for urease maturation in Helicobacter pylori crystal structure of UreF−UreH protein complex. J. Biol. Chem. 2011, 286, 43241−43249. (47) Yang, X.; Li, H.; Lai, T. P.; Sun, H. UreE−UreG complex facilitates nickel transfer and preactivates GTPase of UreG in Helicobacter pylori. J. Biol. Chem. 2015, 290, 12474−12485. (48) Yang, N.; Tanner, J. A.; Zheng, B.; Watt, R. M.; He, M.; Lu, L.; Jiang, J.; Shum, K. T.; Lin, Y.; Wong, K. L.; Lin, M. C. M.; Kung, H. F.; Sun, H.; Huang, J. Bismuth complexes inhibit the SARS coronavirus. Angew. Chem., Int. Ed. 2007, 46, 6464−6468. (49) Wang, R.; Lai, T. P.; Gao, P.; Zhang, H.; Ho, P. L.; Woo, P. C.; Ma, G.; Kao, R. Y.; Li, H.; Sun, H. Bismuth antimicrobial drugs serve as broad-spectrum metallo-β-lactamase inhibitors. Nat. Commun. 2018, 9, 439. (50) Cornaglia, G.; Giamarellou, H.; Rossolini, G. M. Metallo-βlactamases: a last frontier for β-lactams? Lancet Infect. Dis. 2011, 11, 381−393. (51) 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-β-lactamase antibiotic resistance. Nature 2014, 510, 503−506. (52) Brem, J.; van Berkel, S. S.; Aik, W.; Rydzik, A. M.; Avison, M. B.; Pettinati, I.; Umland, K. D.; Kawamura, A.; Spencer, J.; Claridge, T. D.; McDonough, M. A.; Schofield, C. J. Rhodanine hydrolysis leads to potent thioenolate mediated metallo- β -lactamase inhibition. Nat. Chem. 2014, 6, 1084−1090. (53) Han, B. J.; Zhang, Z.; Xie, Y. X.; Hu, X. Q.; Wang, H. B.; Xia, W.; Wang, Y. L.; Li, H.; Wang, Y. C.; Sun, H. Multi-omics and temporal dynamics profiling reveal disruption of central metabolism in Helicobacter pylori on bismuth treatment. Chem. Sci. 2018, 9, 7488− 7497. (54) Morones-Ramirez, J. R.; Winkler, J. A.; Spina, C. S.; Collins, J. J. Silver enhances antibiotic activity against gram-negative bacteria. Sci. Transl. Med. 2013, 5, 190ra81.

227

DOI: 10.1021/acs.accounts.8b00439 Acc. Chem. Res. 2019, 52, 216−227