Metal Messengers: Communication in the Bacterial World through Transition-Metal-Sensing Two-Component Systems

Abstract

Bacteria survive in highly dynamic and complex environments due, in part, to the presence of systems that allow the rapid control of gene expression in the presence of changing environmental stimuli. The crosstalk between intra- and extracellular bacterial environments is often facilitated by two-component signal transduction systems that are typically composed of a transmembrane histidine kinase and a cytosolic response regulator. Sensor histidine kinases and response regulators work in tandem with their modular domains containing highly conserved structural features to control a diverse array of genes that respond to changing environments. Bacterial two-component systems are widespread and play crucial roles in many important processes, such as motility, virulence, chemotaxis, and even transition metal homeostasis. Transition metals are essential for normal prokaryotic physiological processes, and the presence of these metal ions may also influence pathogenic virulence if their levels are appropriately controlled. To do so, bacteria use transition-metal-sensing two-component systems that bind and respond to rapid fluctuations in extracytosolic concentrations of transition metals. This perspective summarizes the structural and metal-binding features of bacterial transition-metal-sensing two-component systems and places a special emphasis on understanding how these systems are used by pathogens to establish infection in host cells and how these systems may be targeted for future therapeutic developments.

Graphical Abstract

INTRODUCTION

Microbial organisms are some of the most diverse, prolific, and versatile life-forms that are capable of surviving in virtually all planetary niches including within the human body.^1,2 Many of these locations, such as deep-sea waters, toxic waste sites, hydrothermal vents, frozen glaciers, and volcanic ash, are inhospitable and may be constantly changing.^3,4 Because these environments are often harsh, diverse, and dynamic, most unicellular organisms can only survive in these locations because of their ability to respond and to adapt rapidly to environmental changes. For microbes, this crosstalk between intra- and extracellular environments is primarily mediated by two-component signal transduction systems.^5,6

Two-component signal transduction systems are widespread among bacteria and play crucial roles in critical cellular processes, such as motility, chemotaxis, and the production of virulence factors.^7–14 Two-component systems typically function by combining two proteins to serve three functions efficiently: environmental recognition, signal transduction, and gene regulation. Generally, the two proteins in bacterial two-component systems are a membrane-bound sensor histidine kinase (HK) and a cytosolic response regulator (RR) (Figure 1). The sensor HK receives the environmental stimulus by interacting directly with either a signaling ligand or a receptor that binds to a signaling ligand, and the signal transduction begins. Stimulus binding then induces an autophosphorylation reaction in which the γ-phosphate of an ATP molecule is transferred to a conserved His residue on the HK. The signal now exists as a covalently attached phosphoryl moiety that may be readily transferred to the receptor RR. The phosphoryl group is then delivered via a phosphotransferase reaction to a conserved Asp residue on the RR.^15,16 The RR typically consists of a regulatory domain and a DNA-binding domain, although variations do exist. Phosphorylation occurs at the regulatory domain and triggers conformational and dynamical changes in the overall protein, altering the affinity of the RR to DNA, and allowing the RR to affect downstream transcription (Figure 1).¹⁷ Specific HKs are typically paired with genepecific RRs that either activate or, less frequently, inhibit transcription of DNA when bound.^5,18

Figure 1.

Cartoon of a prototypical two-component system. In a typical two-component system, a transmembrane sensor histidine kinase (HK), embedded in the lipid bilayer, senses a signal in its periplasmic/extracellular domain. Upon signal detection, an autophosphorylation event occurs that causes conformational and dynamical changes throughout the HK. These changes functionally alter the HK to become a phosphotransferase, and the phosphoryl moiety is then transferred to its cognate soluble response regulator (RR). The phosphorylated RR generally has a higher affinity to its cognate DNA sequence(s), where it then binds to regulate gene expression.

While these systems generally facilitate the ability of all bacteria to live across multiple environments, two-component systems may also function as critical virulence factors associated with problematic pathogens. As just a few examples, the Vibrio cholerae VxrA/B two-component system controls the production of toxins that cause cholera; the Escherichia coli PhoP/Q two-component system regulates the phoPQ operon that alters the ability of E. coli to survive within macrophages and when challenged with antibiotics; and the Salmonella typhimurium EnvZ/OmpR two-component system is necessary for the expression of several virulence genes, as mutants have a reduced ability to infect hosts.^19–22 While a more comprehensive list of virulence-associated two-component systems is listed in Table 1, it is clear that numerous problematic and antibiotic-resistant microbes use these two-component systems to survive in several environmental niches. In many cases, these systems sense the presence of trace metal ions in the environment, such as transition metals. Transition metals are necessary for the normal metabolic functions of all bacteria, and the presence of critical metals (such as Fe, Zn, Mn, and Cu) often regulate microbial virulence.^23–26 This perspective summarizes the general structural and mechanistic knowledge of these metal-sensing two-component systems, which could be targets of future therapeutic developments to tackle antibiotic resistance of certain pathogens.

Table 1.

List of Select Two-Component Histidine Kinase (HK) Signal Transduction Systems in Pathogenic Bacteria and Their Known (or Assumed) Environmental Signals

Two-Component Sensor/Regulator	Pathogen	Environmental Signals/Conditions	Refs
VxrA/VxrB	Vibrio cholerae	cell wall damage	19
PhoP/PhoQ	Escherichia coli, Erwinia chrysanthemi	[Mg2+]	27, 28
EnvZ/OmpR	Salmonella typhimurium	osmolarity	20
CovS/CovR	Streptococcus pyogenes	[Mg2+]	29
PmrB/PmrA	Pseudomonas aeruginosa, Salmonella typhimurium, Erwinia carotovora	[Fe3+], pH	30, 31
TorS/TorR	Shewanella oneidensis	acid stress	32
BvgS/BvgA	Bordetella pertussis	[Mg2+], temperature, nicotinic acid	10, 33
ArcA/ArcB	Escherichia coli, Haemophilus influenzae	anaerobiosis	34
GacS/GacA	Pseudomonas aeruginosa	glycans	35, 36
BT2391	Bacteroides thetaiotaomicron	mucins	37
WspR/WspH	Burkholderia cenocepacia	chemosensory	38
BasS/BasR	Bacillus subtilis, Escherichia coli	[Fe3+]	39
FirR/FirS	Haemophilus influenzae	[Fe2+]	40
BqsR/BqsS	Pseudomonas aeruginosa	[Fe2+]	41
WalK/WalR	Bacillus subtilis, Staphylococcus aureus	[Zn2+]	42
BaeS/BaeR	Escherichia coli, Salmonella enterica	[Zn2+]	43, 44
ZraS/ZraR	Escherichia coli, Salmonella enterica	[Zn2+]	45
CzcS/CzcR	Pseudomonas aeruginosa, Pseudomonas putida	[Zn2+]	46
CopR/CopS	Pseudomonas aeruginosa, Corynebacterium glutamicum	[Cu+], [Cu2+]	47, 48
CusR/CusS	Escherichia coli, Klebsiella pneumoniae	[Cu+], [Cu2+], [Ag2+]	49
DsbR/DsbS	Pseudomonas aeruginosa, Escherichia coli	[Cu2+]	50
ManS/ManR	Synechocystis sp. PCC 6803	[Mn2+]	51, 52
ColS/ColR	Pseudomonas putida, Pseudomonas aeruginosa, Xanthomonas citri	[Mn2+], [Cd2+], [Fe3+], [Zn2+]	53

THE STRUCTURE OF HISTIDINE KINASES

In two-component systems, the sensor HKs may be divided into three groups based on their domain composition and mode of phosphoryl transfer (Figure 2). The classical two-component HK senses a specific signal at the periplasmic/extracellular face of the membrane-embedded HK; after signal sensing, the HK autophosphorylates at a conserved His residue, and the phosphoryl moiety (now the signal) is transferred to a conserved Asp residue on the receiver domain of the cognate RR (Figure 2). In addition to the classical HK, unorthodox and hybrid HKs also exist. In the unorthodox signal transduction system, a more complex phosphoryl-relay mechanism exists wherein the HK has distinct transmitter, receiver, and phosphotransferase domains all tethered to the input domain, necessitating multiple consecutive phosphorylation events. The phosphoryl moiety is then relayed to the RR for downstream gene regulation (Figure 2). The unorthodox HK is less common than the classical HK, but this system is still used in pathogens, such as Shewanella oneidensis (TorS/R), Bordetella pertussis (BvgA/S), and even E. coli (ArcA/B).^10,32,33,54 The hybrid HK is very similar to the unorthodox HK except that there is a lack of a second tethered His acceptor domain; instead, a stand-alone, intermediary protein is utilized that then catalyzes the transfer of the phosphoryl moiety to its respective RR (Figure 2).⁵⁵ Hybrid HKs are present in several pathogens such as the Pseudomonas aeruginosa (GacS/A), Bacteroides thetaiotaomicron (BT2391), and Burkholderia cenocepacia (WspR/H).^35,37,38

Figure 2.

There are three major classes of two-component histidine kinase (HK) systems: classic, unorthodox, and hybrid. Classic membrane-bound HKs tether the input domain (anchored in the lipid bilayer but extended into the periplasmic/extracellular space) to a combined transmitter and phosphotransferase domain in the cytosol. Unorthodox membrane-bound HKs tether the input domain to three distinct, sequential domains in the cytosol: transmitter, receiver, and phosphotransferase. Hybrid membrane-bound HKs tether the input domain to two distinct sequential domains (transmitter and receiver) that ultimately transfer the phosphoryl group to an independent, stand-alone phosphotransferase in the cytosol. In all cases, after the phosphoryl cascade across the HK, the phosphoryl moiety is transferred to a conserved Asp residue on the receiver (response regulator; RR) that then produces an output such as transcriptional regulation. The classes of HKs are also distinguished based on whether their phosphoryl relay events are two-step (classic) or four-step (unorthodox and hybrid).

In a two-component system, a necessary function of HKs is their ability to sense a variety of external stimuli and to transmit this information to a cognate RR via a phosphorylation event, which is made possible by the architecture of the HK itself.⁵ HKs are generally composed of the same basic signaling components including a dimerization and histidine phosphotransfer (DHp) domain and a catalytic ATP-binding (CA) domain, while the variety of environmental stimuli that are perceived is derived from a modular periplasmic/extracellular sensing domain.⁵ The sensing domain functions as the location of signal input, upon which ATP-dependent autophosphorylation of the HK occurs.^56,57 The overall transcriptional control of a two-component system is dependent on the sensing and phosphorylation activity of the HK, as the phosphorylation of the RR happens almost exclusively from its cognate HK.⁵⁸ The general structural features of the HK are described below.

The Kinase Core: The DHp and CA Domains.

A common structural feature of HKs is the kinase core, comprising the DHp and CA domains, both of which exhibit strong structural homology across organisms. The Pfam database of DHp domains recognizes four common sequences that compose the His kinase A (phosphoacceptor) domain,⁵⁹ although additional diversity does exist. X-ray crystallography has revealed that DHp domains typically form long antiparallel α-helices connected by a hairpin loop region that generate a four-helix bundle upon homodimerization (Figure 3).^5,60 This architecture is well conserved across organisms; for example, the superpositioning of a random selection of DHp domains shows high structural similarity (Cα RMSD of 1.072 Å) as illustrated in Figure 3. Interestingly, the handedness of the DHp helical bundle arrangement (whether left- or right-handed) defines whether autophosphorylation is cis (intra-molecular, i.e., on the same HK protomer) or trans (intermolecular, i.e., on the neighboring HK protomer).⁶¹ Within these domains is a conserved phosphoryl-accepting His residue, and the initial catalytic step in signal transduction is the phosphorylation of the Nε atom in this conserved His residue by the γ-phosphoryl group of ATP (Figure 3).⁶²

Figure 3.

Structural conservation of the histidine phosphotransfer (DHp) domains of HKs. (a) Superpositioning of crystallized DHp domains reveals strong conservation of the DHp domain architecture (Cα RMSD value of 1.072 A). (b) DHp domain of the Cu+/Ag+-binding E. coli HK CusS (PDB ID 7ZP0). (c) DHp domain of the Zn2+-binding Lactobacillus plantarum HK WalK (PDB ID 4U7N). (d) DHp domain of the pH-sensing Thermotoga maritima HK TM0853 (PDB ID 2C2A). (e) DHp domain of a T. maritima HK HK853 (PDB ID 6RFV); the stimulus that is sensed via this HK is unknown. In panels b−e, the side chains of the phosphoryl-accepting His residues are explicitly shown. The N- and C-proximal termini are represented by N* and C* labels, respectively, while the second protomer proximal termini are represented by ‘N* and ‘C* labels, respectively.

The DHp domain is located spatially adjacent to the CA domain, which catalyzes the transfer of the γ-phosphoryl group from ATP to the His residue on the DHp domain.^5,62 The CA domain functions as a monomer and binds ATP by using a fold that consists of an α/β-sandwich comprising three to four α-helices and five mixed β-strands (Figure 4). This fold is distinct from known Ser/Thr/Tyr kinases but is structurally homologous to the ATPase domains of the glycoside hydrolase (GHL) ATPase superfamily.¹⁷ GHL proteins are defined by the presence of a conserved ATPase domain that, upon nucleotide binding, undergoes conformational changes to “trap” the substrate.⁶³ Similar to the structural conservation of the DHp domain, the α/β-sandwich of the CA domain is well conserved across two-component system kinase cores; for example, the superpositioning of a set of CA domains shows high structural similarity (Cα RMSD of 0.950 Å) as illustrated in Figure 4. These CA domains contain a highly conserved ATP-binding cavity, coined the ATP lid, that is highly flexible in the absence of nucleotide but becomes ordered and covers the nucleotide in its presence.^64,65 In other GHL ATPase family members, the structural changes of the ATP-lid also appear to serve a regulatory function, and a similar conclusion has been made for HK CA domains.⁶⁶ Upon sensing signal, the autophosphorylation event induces rotation of the DHp helices in such a manner that brings the CA domain closer to the HK, which could amplify signal detection.¹⁷

Figure 4.

Structural conservation of the catalytic ATP-binding (CA) domains of HKs. (a) Superpositioning of crystallized CA domains reveals strong conservation of the CA domain architecture (Cα RMSD value of 0.950 A). (b) CA domain of the Cu+/Ag+-binding E. coli HK CusS (PDB ID 7ZP0). (c) CA domain of the Zn2+-binding L. plantarum HK WalK (PDB ID 4U7N). (d) CA domain of the pH-sensing T. maritima HK TM0853 (PDB ID 2C2A). (e) CA domain of a T. maritima HK HK853 (PDB ID 6RFV); the stimulus that is sensed via this HK is unknown. The N- and C-proximal termini are represented by N* and C* labels, respectively.

The Periplasmic/Extracellular Sensor Domain.

Bacterial organisms use two-component HKs to sense a diversity of environmental signals,^5,7,8,10,17 and the ability to do so with such a broad range of potential external stimuli is a function of the periplasmic/extracellular sensor domain. Because of the assortment of signals that are perceived, the sensory domains have high selectivity, share very little primary sequence similarity, and have highly variable structures.^5,67 Underscoring the latter point, the structural diversity of the sensor domain is highlighted in Figure 5, which demonstrates large structural divergence of several randomly chosen HK sensor domains (Cα RMSD > 15 Å) despite the strong structural homology of their cognate DHp and CA domains (see Figures 3 and 4, for example). Unfortunately, the broad sequence and structural diversities of the sensor domains make the determination of the specificity of each HK difficult a priori; in some cases, the presence of the HK within a gene cluster can be used to infer a probable stimulus, but experimental evidence may be lacking. However, a few common topologies have been observed. For example, some sensory domains consist of an extra-periplasmic/cellular sensory region that resides between two transmembrane α-helices (Figures 5 and 6). It is believed that this type of sensory domain allows for extra-periplasmic/cellular signals, such as metal ions, to be detected and transduced throughout the membrane to regulate kinase activity.^17,68 Other sensory domains are believed to detect more complex stimuli (such as metal chelates) and consist of a larger binding pocket buried within multiple helices (see Figure 6c, for example);⁶⁸ these sensory domains can even sense ion gradients across the membrane. However, due to the large sequence variability among sensor domains, the specific stimuli remain to be uncovered in many cases, and only a limited number of structures of HK sensor domains have been solved, representing a continued challenge in the field that is ripe for discovery.⁶⁸

Figure 5.

Structural divergence of HK sensory domains. (a) Superpositioning of crystallized HK sensory domains demonstrates strong structural diversity (Cα RMSD value >15 A). (b) Sensor domain of the Cu+/Ag+-binding E. coli HK CusS (PDB ID 5KU5). (c) Sensor domain of the Zn2+-binding S. aureus HK WalK (PDB ID 4YWZ). (d) Sensor domain of the nitrate-sensing E. coli HK NarQ (PDB ID 5IJI). (e) Sensor domain of the enterobactin-sensing P. aeruginosa HK PfeS (PDB ID 3KYZ). The N- and C-proximal termini are represented by N* and C* labels, respectively, while the second protomer proximal termini are represented by ‘N* and ‘C* labels, respectively.

Figure 6.

Structural diversity of the stimulus-binding pockets of the HK sensory domains. (a) Sensor domain of the Cu+/Ag+-binding E. coli HK CusS (PDB ID 5KU5). Ag+ ions (gray spheres) are coordinated in a linear geometry via His44 and His176 residues (one from each protomer). (b) Sensor domain of the nitrate-sensing E. coli HK NarQ (PDB ID 5IJI). Nitrate, bound within the α-helices that comprise the sensory domain, is coordinated electrostatically by interactions with Arg50 (one from each protomer). (c) Sensor domain of the enterobactin-sensing P. aeruginosa HK PfeS (PDB ID 3KYZ). Several bound Cl⁻ ions (yellow spheres) make electrostatic contacts with the sensor domain that would otherwise be occupied by enterobactin. The N- and C-proximal termini are represented by N* and C* labels, respectively, while the second protomer proximal termini are represented by ‘N* and ‘C* labels, respectively.

THE RESPONSE REGULATOR (RR)

The cognate HK response regulator (RR) proteins are known as the “on−off” switches of two-component systems, as they represent the terminal component of the phosphoryl-based signal transduction pathway.⁵⁸ The RR catalyzes the phosphoryl transfer from the HK to a conserved Asp residue on itself via a phosphotransferase reaction.^5,17,58 Additionally, many RRs may catalyze auto-dephosphorylation, which allows the RR to limit the duration of its activated state and to attenuate intrinsically its own transcriptional response.⁶⁹ Like other components of the two-component HK systems, the three-dimensional architecture of the RR is linked to its function.

Most two-component RRs are composed of two domains: a highly conserved N-terminal regulatory domain and a variable C-terminal effector domain.⁵ The N-terminal domain of most RRs is also known as the regulatory domain, and a number of structures of these domains have been solved via X-ray crystallography and/or NMR, revealing strong structural conservation;^19,69–71 for example, the superpositioning of a random selection of RR N-terminal domains shows high structural similarity (Cα RMSD of 0.791 Å) as illustrated in Figure 7. The N-terminal regulatory domain of most RRs adopts a canonical (βα)₅ assembly in which a central five-stranded parallel β-sheet is surrounded by five α-helices (Figure 7). Crucially, the regulatory domain also hosts the conserved Asp residue that accepts the phosphoryl group from the HK (Figure 7). In most N-terminal regulatory domains, this Asp is located on a solvent-exposed loop between β3 and α3 that is adjacent to other acidic residues believed to be involved in the coordination of Mg²⁺, which is required for phosphoryl transfer and dephosphorylation.^{5,17,58,71–73} Phosphorylation of the RR N-terminal domain ultimately causes conformational and dynamical changes that affect a large portion of the RR; these changes facilitate an array of molecular interactions that cause the desired transcriptional response, which occurs at the C-terminal effector domain.^5,17,73

Figure 7.

Structural conservation of the N-terminal regulatory domains of the response regulators (RR). (a) Superpositioning of crystallized regulatory RR domains reveals strong conservation of the domain architecture (Cα RMSD value of 0.791 A). (b) Regulatory domain of the Francisella tularensis biofilm formation RR QseB (PDB ID 5UIC). (c) Regulatory domain of the K. pneumoniae lipopolysaccharide modification and polymyxin resistance controlling RR PmrA (PDB ID 3W9S). (d) Regulatory domain of the E. coli virulence controlling RR PhoP (PDB ID 2PL1). (e) Regulatory domain of the S. pneumoniae cell wall metabolism and fatty acid biosynthesis controlling RR YycF (PDB ID 2A9O). In panels b−e, the side chain of the phosphorylated Asp residue is explicitly shown. The N- and C-termini are represented by N and C labels, respectively.

The C-terminal domain of most RRs is also known as the effector domain, and these domains have an array of structural diversity. Most C-terminal RR domains are DNA-binding domains that either activate or repress the transcription of specific genes. Much like the HK sensor domains, the C-terminal RR effector domains are incredibly varied in both sequence and structure as a number of diverse genes need to be targeted.^58,73 So far, three major subfamilies of C-terminal effector domains have been identified: osmotic pressure response regulator-like (OmpR), nitrate/nitrite metabolism regulator-like (NarL), and nitrogen regulatory response regulator-like (NtrC) domains.⁵ The largest subfamily of the effector domains are the OmpR-like^74–76 that comprise a conserved fold containing a recognition helix that is capable of interacting with the major groove of DNA and flanking loops that are proposed to interact with the minor groove of DNA.⁷⁷ NarL-like effector domains comprise a conserved fold containing a helix-turn-helix motif believed to interact directly with DNA.⁷⁸ NtrC-like effector domains are the most structurally complex subfamily of the C-terminal domains of RRs and comprise two subdomains: an ATPase domain and a helix-turn-helix DNA-binding domain.⁷⁹ It is believed that NtrC-like response regulators oligomerize after phosphorylation into octamers that hydrolyze ATP and regulate transcription. ATP hydrolysis is thought to provide energy for both open complex formation and the activation of transcription.^79,80 Thus, similar to the architecture of the HK, the RR of the two-component system mixes both conserved and variable structural features to afford a common function (activation via phosphorylation) while being able to elicit variable outputs (gene regulation).

After phosphorylation, most RRs limit the duration of their activated state (and thus their signal output) by a process known as auto-dephosphorylation.⁷³ This intrinsic phosphatase activity attenuates the signal transduction pathway, although the rate may vary dramatically among RRs. For example, the phosphorylated CheY response regulator has a half-life of just a few seconds,⁸¹ the phosphorylated VanR response regulator has a half-life of ca. 10 h,⁸² and the phosphorylated yeast SSK1 has a half-life of nearly 48 h.⁸³ Because the rates of the intrinsic RR phosphatase activities vary so greatly, many HKs are bifunctional and are able to dephosphorylate their cognate RRs directly instead of relying on intrinsic RR phosphatase activity.^84–86 This HK bifunctionality is thought to be afforded by important conformational changes that allow the HK to switch between phosphotransfer and phosphatase states. Studies of the cytoplasmic region of the B. subtilis DesK HK in complex with its cognate RR DesR have provided some structural insight into these different states.^87,88 As part of its phosphatase activity, the cytoplasmic portion of DesK (including both the CA and DHp domains) exhibits a rigid, symmetric structure in which the DHp α-helices form a conserved coiled-coil motif that induces the movement of the phosphoryl-accepting His residue away from the catalytic region, preventing phosphor-transfer. Once autophosphorylated, this coiled-coil motif breaks and the DesK adopts a substantially more flexible and asymmetric conformation that is able to transfer the phosphate moiety to DesR.^88,89 These data suggest that both the orientation and the distance of the phosphoryl-accepting His residue may be controlling factors for phosphatase/phosphotransferase discrimination,⁸⁸ but more studies are necessary to unravel this complex mechanism.

THE HK/RR COMPLEX

Protein−protein interactions between the HK and the RR are essential for signal transduction, and structural studies have begun to shed light onto the intricacies of the HK/RR complex, although many questions remain. Due to the complexities of trapping HK (an integral membrane protein) in complex with its cognate, soluble RR, most structural research on bacterial two-component systems has focused on isolated, soluble, and sometimes fragmentary domains. Regarding the HK, while structures of the entire CA and DHp domains have been solved,^90–94 the domains appear to be in inactive conformations as the ATP-binding pocket and conserved phosphoryl-accepting His residue are too far apart for catalysis to occur; the active conformations of these domains remain to be revealed. A few complex structures between the soluble HK and RR domains have been solved to modest resolution,^95,96 expanding our understanding of the signal-transduction process. The first example occurred in 2006 when researchers used small-angle X-ray scattering and X-ray crystallography to characterize biophysically the T. maritima ThkA/TrrA HK/RR complex (resolved to 4.2 Å).⁹⁵ In 2009, the same ThkA/TrrA system was crystallized and its structure was determined to an improved 3.8 Å resolution, allowing the authors to build a model that exposed the interdomain interactions of the complex between the soluble portions of the HK and the RR. This structure revealed that the HK ThkA forms a four-helix homodimer and that the phosphoacceptor Asp residue of the RR TrrA faces the conserved HK His residue, presumably to allow phosphorrelay.⁹⁶ Only a few additional truncated HK/RR complex structures since then have been solved such as the B. subtilis HK DesK complexed with its cognate RR DesR, which has been studied more recently in the context of dephosphorylation (vide supra).^88,89 Thus, future studies should be undertaken to decipher the atomic-level details of the interactions between the intact HK in complex with its RR, which could be made possible given recent major advances in cryo-EM techniques.

TRANSITION-METAL-SENSING TWO-COMPONENT SYSTEMS

Transition metals are essential for critical biological processes such as enzyme catalysis, electron transfer, and even the structural stabilization of biological macromolecules.^97–100 However, transition metals can also be toxic to cells at high concentrations, as they can disrupt cellular functions through protein and/or nucleic acid misfolding and/or via the generation of reactive oxygen species (ROS).^101–104 Thus, all organisms have developed sophisticated mechanisms to regulate transition metal ion homeostasis such as metal import, metal chaperones, metal storage, metal efflux, and even metal-sensing two-component systems. Of particular interest, bacteria use the modular architecture of the two-component system to sense changes in numerous metal ion concentrations and to adjust their cellular responses and metabolism accordingly. These responses are employed particularly during infection where transition metal ions may be strongly withheld (i.e., nutritional immunity) and in which transition metal ions may be used to combat invading pathogens (i.e., the oxidative burst).^105,106 During these scenarios, bacteria often leverage these transition-metal-sensing two-component systems to enhance their survival and to increase their virulence.

Bacterial transition-metal-sensing two-component systems are typically of the classical form and are composed of a membrane-bound dimeric HK that detects a metal ion concentration outside the cytoplasmic membrane while a cytosolic RR controls gene expression related to the particular transition metal that is sensed. In general, in the absence of transition metal ion binding, the sensor HK exists in a basal state with low kinase activity;^107,108 transition metal ion binding induces conformational and dynamical changes that activate the kinase activity of the sensor HK, leading to autophosphorylation at the conserved His residue. Once phosphorylated, a transfer of the phosphoryl group to the conserved Asp residue on the RR occurs, inducing separate conformational and dynamical changes that allow the RR to bind to DNA and to regulate gene expression.⁵ As a direct response to changes in metal ion concentrations in the extracytosolic region of the cell, the intracellular “output” by the RR may lead to changes in cellular metabolism, the upregulation of import/efflux pumps, and even the expression of genes to combat oxidative stress. After the signal is perceived, the “output” of the RR is typically terminated via dephosphorylation by either an intrinsic or a HK-induced autophosphatase activity.⁵

It is well-known that pathogenic prokaryotes require a range of transition metals to maintain homeostasis and to establish infection within hosts.^23,26,105 Due to the number of required metals and the varied set of oxidation and/or coordination states that a transition metal may adopt, numerous bacterial metal-sensing two-component systems exist. The remainder of this perspective focuses on the two-component systems used to sense the most abundant transition metals that are also strongly linked to bacterial pathogenesis (especially in mammalian hosts): Mn, Fe, Cu, and Zn.^5,98,109 In general, the metal ion selectivity of each system is determined by the metal-binding site built by the periplasmic/extracellular sensor domain of the HK. In some cases, there is either a confusion or a lack of knowledge as to the precise oxidation state that is sensed of each metal, and there may even be promiscuity in which metals are sensed and even which genes are targeted. Our current knowledge is summarized below.

MANGANESE SENSING TWO-COMPONENT SYSTEMS

Manganese (Mn) is an essential redox-active transition metal and a required cofactor in all domains of life,¹¹⁰ playing a critical role in the synthesis and activation of various enzymes, in lipid, protein, and carbohydrate metabolism, and in the virulence of certain bacterial pathogens.^110,111 Examples of Mn metalloenzymes that aid in these processes include arginase, which catalyzes the conversion of L-arginine to L-ornithine and urea,¹¹² glutamine synthetase, which catalyzes the formation of glutamine from glutamic acid and ammonia,¹¹³ phosphoenol-pyruvate decarboxylase, which is involved in anaplerotic metabolism,¹¹⁴ and even Mn-superoxide dismutase, which is an important enzyme for defending against oxidative stress.¹¹⁵ Mn is particularly critical in its contribution to the catalytic detoxification of reactive oxygen species (ROS) and may even replace Fe as the active site of certain enzymes in order to avoid the deleterious oxidative damage by the Fe²⁺/Fe³⁺ redox cycle.^116–119 However, high levels of Mn can be toxic due to this element’s redox activity and the possibility of mismetallation of essential enzymes (in some cases).^120,121 Because of this potential toxicity, bacteria control Mn handling to regulate Mn homeostasis. Surprisingly, however, only a few Mn-specific two-component systems have been discovered and characterized.

The ManSR System.

The manganese sensor/manganese regulator (ManSR) system, composed of a membrane-bound sensor HK, ManS, and a cytosolic RR, ManR, has been found to regulate Mn homeostasis in the cyanobacterium Synechocystis.⁵¹ The ManS protein is predicted to have both periplasmic and cytosolic domains, the former of which is presumably used to sense Mn and the latter of which is the kinase domain. In this bacterium, the mntCAB operon encodes a high affinity Mn transporter that is transcriptionally dependent on Mn concentrations.⁵¹ By using a strain in which the promoter of this operon directs the transcription of luxAB reporter genes, it was discovered that the inactivation of ManS resulted in high levels of lux luminescence, indicating that the ManSR system binds to this region in order to control the transcription of mntCAB.⁵¹ Because mntCAB is only transcribed in Mn-starved conditions, it was hypothesized that ManS senses high Mn levels and phosphorylates ManR (ManR^P); ManR^P then binds to the mnt promoter region to halt transcription.⁵² To explore this possibility, site-directed mutagenesis was used in order to modify the HK domain of ManS to remove the phosphorylatable HK His (H205L).⁵¹ This approach completely abolished ManSR activity and demonstrated lowered rates of Mn acquisition, further underscoring that the ManSR system is critical for Mn homeostasis.⁵¹ Interestingly, results have also shown that ManS may have some regulatory effects on Fe and Zn homeostasis, as the levels of feoB (the primary Fe²⁺ transporter) and znuA (a Zn²⁺ transporter) were also disrupted due to expression of the H205L ManS variant.⁵¹ It is possible that the ManSR system may target more than one RR, one of which may act on mnt (ManR), whereas others may act on the genes of other metal transport systems (such as feo and/or znu). Nevertheless, the ManSR system remains one of the few Mn-specific two-component systems that have been characterized, while other Mn-dependent two-component systems demonstrate promiscuity.

The ColSR System.

An additional two-component system capable of sensing Mn is ColSR, which is known to exist in various species such as P. putida, P. aeruginosa, and Xanthomonas.^53,122,123 ColSR consists of two proteins: a membrane-bound sensor HK, ColS, and a cytosolic RR, ColR. The ColSR system has been previously shown to contribute to membrane functionality via LPS production control, stress tolerance, and virulence of various pathogens, but the conditions that activate ColSR are unclear.^53,122–124 However, experiments show that the ColSR system is necessary for Fe, Zn, Cd, and Mn tolerance in P. putida, suggesting ColSR may be metal-promiscuous.¹²⁴ To explore this possibility, the MIC values of different transition metals for wild-type (WT), ΔcolR, and ΔcolS P. putida strains were investigated. In liquid culture, ΔcolR and ΔcolS P. putida clearly showed an increased sensitivity to Zn, Fe, Cd, and Mn, while their resistance to Co, Cu, and Ni resembled that of the WT strain, which was corroborated in solid media except for the Cd sensitivity.¹²⁴ These sensitivities could be reversed through complementation approaches.¹²⁴ Furthermore, supplementation of ΔcolR P. putida with a phosphorylation-deficient ColR (D51A) mutant plasmid failed to reverse the phenotype, highlighting the importance of the ColSR signal transduction axis for this organism to survive in the presence of excess transition metals.¹²⁴ Additionally, variants of ColS in which the periplasmic sensor domain of ColS (particularly the E¹²⁶xxE¹²⁹ motif) implicated these residues to be important at least for Fe and Zn binding; however, the mechanism for sensing of Mn and Cd remain unclear in this system.

The VicK/GcrR System.

A two-component system known as VicK/GcrR is important in sensing Mn by S. mutans, a major organism living in dental biofilms.^125,126 This Mnsensing two-component system is unique in that there are two cognate RRs of VicK: VicR and GcrR.¹²⁷ The VicRK two-component system was previously demonstrated to modulate biofilm formation, oxidative stress, and acid tolerance responses in S. mutans.¹²⁷ However, subsequent experiments discovered that VicK is alternatively capable of transphosphorylating GcrR in the presence of Mn.¹²⁵ GcrR (also coined CovR) has been shown previously to be essential in the virulence and acid tolerance of S. mutans via SloR modulation, which is a diphtheria toxin repressor (DtxR) homologue that controls the expression of various genes in response to metal availability, particularly Mn.^128,129 GcrR was thought to be an orphan response regulator, meaning it was not linked to a cognate HK, although its ortholog in S. pyogenes (CovR) is linked to CovS.¹³⁰ In S. mutans, GcrR has been shown to modulate sucrose-dependent bacterial adherence by regulating glucosyltransferases B and C, connecting this system to biofilm formation.^131,132 Moreover, the autophosphorylation of VicK is facilitated specifically by Mn²⁺ based on in vitro phosphorylation assays.¹²⁵ VicK is capable of catalyzing phosphor-relay to both its cognate response regulators VicR and GcrR. However, and importantly, VicK phosphorylates VicR at much lower levels in the presence of Mn while it phosphorylates GcrR at much higher levels in the presence of Mn.¹²⁵ These findings suggest that, in the presence of Mn, the VicK system specifically targets GcrR in order to regulate biofilm formation, but how Mn imparts this selectivity is unclear.

IRON SENSING TWO-COMPONENT SYSTEMS

Iron (Fe) is the most abundant transition metal on earth and is an essential nutrient for almost all organisms due to its participation in critical cellular and metabolic processes, such as deoxyribonucleic acid (DNA) biosynthesis, cellular respiration, amino acid biosynthesis, electron transport, O₂ transport, and even N₂ fixation.^133–138 This impressive range of activity is due to the versatility of Fe as a cofactor. For example, Fe can span multiple oxidation states, such as ferrous (Fe²⁺), ferric (Fe³⁺), and ferryl (Fe⁴⁺), can accommodate several coordination states, and its redox potential can span nearly a 1 V range.¹³⁹ However, despite its necessity, Fe is toxic at high concentrations, and if uncontrolled, its redox activity can cause oxidative stress and damage cellular components.^104,140 As a result, most bacteria have evolved a suite of mechanisms to control Fe homeostasis, including two-component systems to sense and to regulate intracellular Fe levels.

The PmrBA System.

One well-known Fe-sensing two-component system is the polymyxin-resistant mutant B/A (PmrBA) system, which exists in various pathogens such as S. enterica and P. aeruginosa.^30,31 The PmrBA system consists of two proteins: PmrB, the sensor HK, and PmrA, the RR, that work in tandem to control the expression of genes involved in virulence and Fe homeostasis.³⁰ PmrB senses extracytoplasmic Fe³⁺ and autophosphorylates itself, after which PmrA catalyzes the transfer of the phosphoryl moiety. Once phosphorylated, PmrA undergoes conformational and dynamical changes that alter the affinity of PmrA for its target DNA regions.¹⁴¹ Examples of promoter regions targeted by PmrA include those upstream of ugd, pmrC, and pmrG genes, which encode for proteins involved in polymyxin resistance and the synthesis of lipopolysaccharides (LPS).^27,142,143 The PmrB protein was found to bind Fe³⁺ via its periplasmic domain that contains several copies of an ExxE motif similar to what is present in other Fe homeostasis proteins.^144,145 PmrBA also senses environments containing toxic levels of Fe, and a PmrA mutant in S. enterica was hypersensitive to death via Fe saturation.¹⁴¹ PmrBA is also necessary in S. enterica for both the acquisition of Fe and the survival in the presence of other cationic antimicrobial peptides.^141,142,146 Interestingly, PmrBA also works in tandem with another two-component system, PhoPQ, which is the major regulator of cell surface properties, antibiotic susceptibility, and stress responses in Salmonella.¹⁴² In this scenario, PmrBA surprisingly appears to sense low extracytoplasmic levels of Mg²⁺, which in turn promotes the transcription of PhoP-activated PmrA-dependent genes.²⁸ However, it is unclear how PmrBA is capable of being selective for both Fe³⁺ and Mg²⁺ under different environmental conditions.

The BasSR System.

Another Fe-sensing two-component system is the BasSR system, which exists in various pathogens such as B. subtilis and E. coli.^39,147 This system consists of two proteins: BasS, the sensor HK, and BasR, the RR, that work together to sense changes in Fe availability and to regulate the expression of genes involved in Fe uptake, Fe utilization, and even biofilm formation.^39,148,149 The mechanism by which the BasSR system senses changes in Fe availability is not well understood. However, like PmrBA, it is thought that BasS senses environmental Fe availability.^39,148 Like PmrB, BasS contains a conserved ExxE motif in its periplasmic domain that is thought to be an analogous Fe-binding motif.¹⁴¹ The Glu residues are hypothesized to coordinate Fe ions, although the oxidation state of Fe that is bound is unclear.^141,150 After Fe binding, BasS undergoes autophosphorylation, which begins the signal transduction pathway and allows BasR to modulate transcription. It is thought that the BasSR system may regulate the expression of the FeoB transporter, which is the chief prokaryotic Fe²⁺ importer in bacteria.¹⁵¹ The FeoB transporter is upregulated in response to low-Fe conditions, and mutants lacking the BasSR system are defective in FeoB expression and iron uptake.^111,152 However, regulation of the feo operon is complicated, as feo is thought to be under control of both FUR (operon repression) and FNR (operon promotion) in most bacteria.^134,151 The BasSR system can override the repression of FUR and induce the expression of FeoB under low-iron conditions, which interestingly also alters biofilm formation.¹⁵² Biofilms are sessile, multicellular bacterial communities that live within an excreted extracellular matrix that is maintained through a tight regulation of bioavailable Fe, as too little Fe prevents biofilm formation, while excess Fe causes biofilm dispersion.¹⁵³ The BasSR system is thought to regulate biofilm formation by modulating the expression of genes involved in extracellular polysaccharide synthesis and cell-surface adhesion such as the ybf operon, which is a transcription factor that regulates the modification of LPS, polymyxin tolerance, and biofilm formation.¹⁵⁴ E. coli mutants lacking the BasSR system are defective in biofilm formation.^150,155 This connection between Feo-mediated Fe²⁺ uptake and biofilm formation via two-component systems is intriguing and warrants future exploration.

The FirSR System.

Another studied Fe-sensing two-component system is the ferrous iron responsive sensor/regulator system (FirSR), which is present in H. influenzae and regulates the expression of genes involved in Fe homeostasis and virulence.^40,156 The FirSR system consists of two proteins: FirS, the sensor HK, and FirR, the RR. FirS is predicted to be a membrane-bound homodimer with a periplasmic sensor domain and a cytoplasmic kinase domain, while FirR is predicted to be a cytoplasmic protein comprising a receiver domain and an effector domain.⁴⁰ The FirSR system has been implicated in regulating the expression of genes involved in both cold shock survival and virulence in H. influenzae. FirSR itself is subject to autoregulation, as the FirR protein can bind to ygiW, a gene overlapping the promoter region of the FirSR operon. The function of ygiW is unknown, but binding of FirR at this region lowers the levels of FirSR 6-fold.⁴⁰ In contrast, cold shock experiments showed that the levels of ygiW and firR transcripts increased over 60-fold when cultures were exposed to colder temperatures, which was not the case in deletions of either firR or firS, indicating that both FirR and FirS are needed for ygiW−firSR expression in colder temperatures.⁴⁰ Moreover, upon Fe²⁺ supplementation, promoter activity increased, while the same was not true upon Fe³⁺ supplementation, suggesting that the FirSR system is specific for the sensing of Fe²⁺.⁴⁰ However, this protein has only been studied in H. influenzae, and it is unclear whether this same system is present in other bacteria.

The BqsSR System.

A very intriguing and more abundant Fe-sensing two-component system is the biofilm quorum sensing sensor/regulator system, or BqsSR, which was first discovered in P. aeruginosa but has homologues in other species such as V. cholerae that are also termed CarSR.⁴¹ The BqsSR system consists of two proteins: BqsS, the sensor HK, and BqsR, the OmpR-like RR, which work together to sense Fe²⁺ and to regulate the expression of genes involved in biofilm formation and antibiotic resistance.⁴¹ Experiments have shown that the deletion of either bqsS or bqsR in P. aeruginosa results in a significant increase in biofilm formation, and these mutants were also defective in the production of short chain quorum sensing signals such as the N-butyryl-L-homoserine lactone, the Pseudomonas quinolone signal, and even rhamnolipid production.⁴¹ Additionally, reverse-transcription experiments suggest that BqsS recognizes Fe²⁺ by utilizing an RExxE motif in its periplasmic domain; once Fe²⁺ is sensed, the signal is transduced via BqsR, which binds to a tandem repeat DNA sequence (5′-TTAAG(N)₆-TTAAG-3′) to respond to elevated Fe²⁺ concentrations and cationic stressors.^157,158 Based on this sequence, over 100 genes are potentially under the control of BqsR, including several genes that are involved in cationbinding, metal transport, lipopolysaccharide modulation, polyamine synthesis and transport, and antibiotic resistance.^157–159 Given the large number of genes that are predicted to be targeted by BqsR, as well as its stronger conservation across many bacteria, the characterization of the BqsSR system warrants further studies.

COPPER SENSING TWO-COMPONENT SYSTEMS

Copper (Cu) is one of the most abundant transition metals and has incredibly high affinity for metalloproteins based on the Irving−Williams series.^111,160,161 The generally accepted biologically relevant oxidation states of Cu are cuprous (Cu⁺) and cupric (Cu²⁺), and Cu is found in one of these oxidation states within the active sites of several important enzymes, such as cytochrome c oxidase (an essential electron acceptor in the mitochondrial electron transport chain that is required to catalyze the reduction of oxygen to water for aerobic ATP production), particulate methane monooxygenase (pMMO) (an important enzyme that catalyzes the hydroxylation of methane to methanol), and Cu−Zn superoxide dismutase (SOD) (an enzyme found in all living cells that is critically important in defending against oxidative stresses).^162–165 However, like Mn and Fe, Cu can be toxic at high concentrations, as its facile redox properties allow it to participate in Fenton-like reactions that may irreversibly oxidize proteins, nucleic acids, and lipids, if unregulated.¹⁶⁶ Because of these properties, pathogens must maintain tight control over Cu homeostasis in order to prevent its excess accumulation; unsurprisingly, Cu-sensing two-component systems are part of a suite of proteins that are used to regulate intracellular Cu levels.

The CopSR System.

A now well-known Cu-sensing two-component system is CopSR that monitors extracytoplasmic Cu in various pathogens such as P. aeruginosa, L. lactis, C. glutamicum, and Synechocystis.^47,48,167 CopSR consists of two proteins: CopS, the sensor HK, and CopR, the RR, that work in tandem to regulate Cu homeostasis. CopS is predicted to be a homodimeric membrane protein with a periplasmic sensor domain and a C-terminal cytoplasmic catalytic transfer domain, while CopR is predicted to be a two-domain protein that, when phosphorylated, allosterically modifies transcriptional activity of genes involved in Cu homeostasis.¹⁶⁸ Data have revealed that the CopSR system may even regulate itself, as well as the expression of copBAC, a heavy-metal effluxresistance nodulation and division (RND) type copper efflux system.⁴⁸ The expression of the copBAC operon is critical for resistance to excess Cu, as pathogens with defects in this operon have Cu hypersensitivity and significant growth defects.¹⁶⁹ CopS not only has kinase activity but also a Cudependent phosphatase activity, and the transcriptional control of the CopSR regulon is dependent on the activity of the latter.¹⁶⁸ In L. lactis, CopR controls the expression of at least 11 genes and may even have a self-regulatory role, as CopR was shown to bind to the promoter of the CopR regulon in the absence of Cu.¹⁷⁰ Intriguingly, CopR binds to a “COP box” consensus sequence (5′-TACANNTGTA-3′) via its N-terminal domain, which is conserved at least in Firmicutes.¹⁷⁰ The solution structure of the N-terminal regulatory domain of CopR has been solved via NMR, and this domain surprisingly contains a winged helix structure that is capable of binding DNA (PDB ID 2K4B), a structural feature that is normally part of the C-terminal domain of most RRs.¹⁷⁰ This structural divergence from other two-component RR proteins is intriguing, and as the CopSR system promotes the expression of the periplasmic Cu-homeostasis network in several pathogens, it warrants further investigation.

The CusSR System.

A well-known and structurally characterized Cu-sensing two-component system is the CusSR system, which is present in a number of Gram-negative bacteria such as E. coli and K. pneumoniae.^171,172 The CusSR systems consists of two proteins, CusS, the sensor HK that is capable of binding both Cu and Ag, and CusR, the RR, that work together in order to drive the expression of Cu efflux transporters and redox enzymes responsible for alleviating Cu toxicity.¹⁷¹ Transcriptional studies have shown at least a 2-fold increase in the transcription of cusS and cusR genes after induction with Cu, indicating that CusSR is Cu-regulated.¹⁷³ Genes that are targeted by CusSR include cusCFBA, that encode the CusCFBA complex, the primary bacterial Cu response system including the CusA RND Cu efflux pump.^174–176 Disruption of either cusS or cusR leads to a reduced tolerance for extracellular Cu, increased intracellular Cu levels, and a lack in the transcriptional activation of cusCFBA genes in E. coli.⁴⁹ X-ray crystal structures of the periplasmic sensor domain of CusS have been determined with the O₂-stable Cu⁺ proxy Ag⁺ bound to the protein (Figure 6). This structure revealed linear coordination of Ag⁺ via two His residues (His⁴⁴ and His¹⁷⁶) located at the homodimer interface, consistent with other HK sensor domains.¹⁷⁶ Presumably, Cu⁺ coordinates at this same site in a similar manner. This sensor domain is a mixed α/β fold with five central antiparallel β-sheets that are flanked by α-helices on either side. The N-terminal α-helix is much longer than the C-terminal α-helix and may extend into the transmembrane domain.¹⁷⁶ The structure of the CusS kinase core (comprising the CA and DHp domains) has also been determined (Figures 3 and 4). The catalytic CA domain consists of the canonical topology: an α/β-sandwich fold comprising three α-helices and five mixed β-strands with an ATP-lid.⁹¹ This catalytic CA domain is proposed to hydrolyze ATP via its intrinsic ATPase activity. The DHp domain of CusS consists of two long, antiparallel α-helices that are linked by a hairpin loop that form a four-helix bundle upon homodimerization (Figure 4).⁹¹ Given the importance of this system to bacterial Cu homeostasis, additional structural information in the presence of the cognate metal, Cu⁺, would be beneficial.

The DsbSR System.

The last Cu-sensing two-component system that has been studied in some detail is the protein disulfide bond (DsbSR) formation system. The DsbSR system consists of two proteins: DsbS, the sensor HK, and DsbR, the RR. DsbSR helps regulate cell virulence and Cu sensitivity in various pathogens such as P. aeruginosa and E. coli by inducing the transcription of genes involved in protein disulfide bond formation (the dsbDEG operon) in the presence of Cu to avoid Cu toxicity.⁵⁰ Like the CopS system, DsbS has phosphatase activity toward DsbR in the absence of Cu, which blocks the transcription of dsb genes.⁵⁰ Cu binds to the periplasmic sensor domain of DsbS using a critical Cys⁸² residue, and this binding inhibits the phosphatase activity of DsbS, allowing for the accumulation of phosphorylated DsbR (DsbR^P).⁵⁰ Deletion of the dsbSR genes drastically increased P. aeruginosa Cu susceptibility, while complementation restored Cu resistance.⁵⁰ Interestingly, deletion of dsbDEG resulted in a similar Curesistant phenotype; ectopic expression of the dsbDEG operon in the ΔdsbSΔdsbR mutant was capable of restoring Cu resistance to that of WT, indicating that dsbDEG is a major target of the DsbSR two-component system.⁵⁰ To further support this assignment, DsbR was found to target a consensus 5′-TTA-N₈-TTAA-3′ DNA sequence that exists within the promoter region of dsbDEG, and electrophoretic mobility shift assays (EMSAs) showed that DsbR does indeed bind to this consensus sequence.⁵⁰ However, the precise mechanism by which the dsbDEG genes control Cu toxicity is still unclear.

ZINC SENSING TWO-COMPONENT SYSTEMS

After Fe, zinc (Zn) is the second-most important transition metal in biological systems.¹³⁹ However, unlike Fe, Zn is redox-inert under physiological conditions and is found only in the divalent oxidation state (i.e., Zn²⁺).¹⁷⁷ Nevertheless, Zn is still an essential transition metal for organisms due to its incredibly important role as both a cofactor and a structural component of macromolecules. Zn is necessary for DNA repair, the oxidative stress response, enzymatic reactions, and several cellular processes.¹⁷⁸ Some examples in which Zn is critical include Zn finger proteins that commonly bind to and regulate DNA and RNA (although these are more commonly found in eukaryotes), amino acid biosynthesis, transcriptional activation, lipid binding, and more.^99,111,179 Although Zn is essential, its excess can also be toxic to bacteria; however, unlike Mn, Fe, and Cu where this toxicity is linked typically to redox activity and the generation of ROS, since Zn is redox-inert, it more commonly inhibits key enzymes and competes with the delivery of other transition metal ions to biological macromolecules.¹⁸⁰ Thus, like other metals, pathogens both sequester Zn excess and regulate Zn uptake,¹⁸¹ the latter of which can be controlled by Zn-sensing two-component systems.

The WalKR System.

One well-studied Zn-sensing two-component system is the WalKR system, which is highly conserved within low G+C content Gram-positive pathogens such as B. subtilis, S. aureus, S. pneumoniae, and E. faecalis.^182–185 The WalKR system consists of two proteins: WalK, the sensor HK containing a Per-Arnt-Sim (PAS) domain both in its sensing and in its cytoplasmic regions, and WalR, the RR.⁴² WalKR is involved in the regulation of several genes that are important for virulence, biofilm formation, and cell viability.^42,186,187 Interestingly, while the sensor domain has not been crystallized in the presence of Zn²⁺, structural investigations of the cytoplasmic WalK PAS domain revealed Zn²⁺ bound tetrahedrally to four amino acids (His²⁷¹, Asp²⁷⁴, His³⁶⁴, Glu³⁶⁸); mutations of this binding site negatively impact the WalK kinase activity and WalR phosphorylation, indicating that intracellular Zn²⁺ may negatively regulate the WalKR signal axis.¹⁸⁸ WalKR is known to be essential for survivability of both S. pneumoniae and S. aureus.^42,186 For example, in S. pneumoniae, the expression of several cell-wall biosynthesis genes including pcsB, a cysteine, histidine-dependent amidohydrolase/peptidase (CHAP) domain protein, is dependent on the WalKR system.^187,189 In another example, cell death of S. aureus is induced by WalKR deletion, as the WalKR system regulates peptidoglycan synthesis and positively controls autolytic activity in the two major S. aureus autolysins, AtlA and LytM.⁴² In fact, WalR appears to target at least nine genes in S. aureus involved in cell wall degradation, and the deletion of walKR led to a significant decrease in peptidoglycan biosynthesis and to cell wall modifications.⁴² Additionally, WalKRdepleted cells were observed to be more resistant to lysostaphin-induced lysis, further underscoring the involvement of WalKR in maintaining cell wall structure. Finally, in B. subtilis, WalR has been shown to bind to the promoter regions of yocH and ykvT, which encode cell wall hydrolases, and tagAD, which produces key enzymes of teichoic acid biosynthesis, further linking the WalKR system to cell viability.¹⁹⁰ Given the strong connection between the WalKR two-component system and pathogenic viability, more studies into the mechanistic functions of this system are strongly warranted.

The BaeSR System.

A second characterized Zn-sensing two-component system is the BaeSR system, which has been studied in the context of E. coli and S. enterica.^43,44,191 The BaeSR system is a classical two-component system comprising two proteins: BaeS, the sensor HK, and BaeR, the RR. BaeS may sense a variety of environmental stimuli, including Zn²⁺, after which it autophosphorylates and then transfers the phosphoryl moiety to a conserved Asp residue on BaeR. Phosphorylated BaeR (BaeR^P) has increased affinity for DNA, and a known consensus sequence to which BaeR^P binds is 5′-TTTTTCTCCATDATTGGC-3′ (where D could be T, G, or A) that is known to be present in several promoter regions.^191,192 The genes downstream of these promoters are involved in drug resistance, stress response, and metal resistance/homeostasis such as three RND efflux pumps (MdtABC, AcrD, MdtD), one periplasmic protein (Spy, function unknown),¹⁹¹ and the baeSR operon itself.^44,193 S. enterica strains lacking baeSR exhibit significant growth defects in high Zn²⁺ conditions, confirming this system’s important role in Zn²⁺ regulation.¹⁹³ Interestingly, in addition to responding to high levels of extracytoplasmic Zn²⁺,⁴³ the BaeSR system also responds to genes in a more promiscuous manner.¹⁴⁷ For example, some experiments have shown a response from BaeSR in the presence of spheroplasts, indole, tannin, sodium tungstate (Na₂WO₄), and even Cu.^44,147 Thus, while the BaeSR system appears to sense extracytoplasmic Zn²⁺ levels and to activate several multidrug efflux mechanisms, BaeSR may be a more promising therapeutic target due to its larger promiscuity in sensing and gene regulation.

The ZraSR System.

A third characterized Zn-binding two-component system is the zinc resistance associated system, or ZraSR, which is known to be present in several pathogens including E. coli, S. enterica, and Proteus mirabilis.^194–196 Unlike the WalKR and BaeSR systems, the ZraSR system is unique, as it contains a third periplasmic repressor partner, ZraP, in addition to its sensor HK, ZraS, and its RR, ZraR.⁴⁵ The ZraSR system is responsive to Zn²⁺ concentrations;⁴⁵ high Zn²⁺ concentrations are sensed by the ZraSR system that then contributes to the activation of metal tolerance genes and, once stable, represses itself by activating the expression of zraP.¹⁹⁵ The third protein of this system, ZraP, is an important Zn-binding soluble protein expressed under Zn²⁺ stress conditions and has been reported to act as a Zn²⁺-dependent molecular chaperone.^196,197 The binding of Zn²⁺ to ZraP causes an ATP-independent oligomeric change (from monomer to octamer) that significantly increases ZraP’s stability and function.^45,196 The structure of Zn²⁺-bound ZraP reveals the presence of four interfacial Zn²⁺-binding sites that contribute to its oligomeric stability.¹⁹⁸ It has been proposed that this two-component system participates in cellular Zn²⁺ balance as well as Zn²⁺ overload protection via metal ion sequestration via ZraP.⁴⁵ In addition to sensing Zn²⁺, experiments have also found that ZraSR regulates hydrogenase 3 formation, a portion of the formate hydrogen lyase complex that is involved in fermentative H₂ production,¹⁹⁵ which is a significant product in microbial metabolism.¹⁹⁹ This intriguing duality of the ZraSR system remains to be explored further.

The CzcSR System.

The final Zn-sensing two-component system that has been significantly characterized is the CzcSR two-component system, which has been studied in P. putida, P. aeruginosa, and Cupriavidus metallidurans.^46,200,201 The CzcSR system consists of two proteins, CzcS, the sensor HK, and CzcR, the RR, that work in tandem to regulate virulence and to attenuate the toxicity of several transition metals.⁴⁶ Like many proteins of the czc operon, the CzcSR system is a promiscuous metal binder and has been shown to be induced in the presence of not only Zn²⁺ but also Cd²⁺, Co²⁺, and Cu²⁺.²⁰² After signal transduction via the CzcSR axis, there is an increase in expression of the CzcCBA efflux pump machinery that induces resistance to high concentrations of Zn²⁺, Cd²⁺, Co²⁺, and/or Cu²⁺ through metal-export mechanisms.²⁰² Importantly, CzcR also simultaneously regulates carbapenem resistance by down-regulating the expression of the OprD porin, the main route of entry for this important class of antibiotics that are used against numerous drug-resistant Gram-negative and Gram-positive bacteria.^202,203 In fact, CzcR may even be involved in the virulence of these pathogens, as a czcR deletion strain showed an alteration in quorum-sensing metabolites, such as rhlA, rhlI, rhlR (involved in rhamnolipid biosynthesis), as well as virulence genes lasB and lasI (involved in biofilm formation).^202,204 Given the dual connection between metal homeostasis and antibiotic resistant imparted by the CzcSR system, further study is strongly warranted.

CONCLUSIONS AND FUTURE PERSPECTIVES

Bacterial transition-metal-sensing two-component systems are critical means by which pathogens sense and respond to environmental levels of transition metals. The utilization of two-component sensing systems by bacteria is necessary, as the right concentration of intracellular metals is crucial: too little and metabolic homeostasis is disrupted, too much and toxicity ensues. Transition metals are also critical for the establishment of infection by pathogens and, as this perspective highlights, several transition-metal-sensing two-component systems are critically connected to virulence. Thus, certain two-component systems could be good targets for future therapeutic developments. In this regard, a positive aspect of this approach is the fact that two-component systems are rare in eukaryotes, which should help preclude off-target effects. However, many features of bacterial two-component systems are conserved, which could be problematic for the targeting of pathogenic bacteria while leaving commensal bacteria unharmed. Studies aimed at characterizing distinctive structural and functional features that delineate pathogenic metal-sensing two-component systems from commensal metal-sensing two-component systems would propel this strategy forward in the future. To understand the signal relay process fully, high-resolution structures of HK/RR complexes are critical. Fewer than three intact HK structures have been solved, which is perhaps unsurprising since HKs are integral membrane proteins. Perhaps more surprising, however, is that very few intact RR structures have been characterized; the paucity of these structures may be due to the dynamic nature of the RR DNA-binding domains, which may preclude crystallization. As such, future NMR studies on intact RRs may provide critical structural information that is not easily obtained from X-ray crystallography. Finally, an intact, full-length HK/RR complex has yet to be structurally characterized. However, with the rising accessibility of cryo-EM and microED techniques, which have been applied to numerous integral membrane proteins and membrane protein complexes,²⁰⁵ an intact HK/RR complex structure may be just beyond the horizon. Additional structural and mechanistic information derived from these studies would represent a boon to the study of transition-metal-sensing two-component systems and would provide future avenues for therapeutic developments to target infectious pathogens.

ABBREVIATIONS

CA

catalytic ATP-binding domain

CHAP

cysteine, histidine-dependent amidohydrolase/peptidase

DHp

dimerization and histidine phosphotransfer domain

DtxR

diphtheria toxin repressor

GHL

glycoside hydrolase

HK

histidine kinase

LPS

lipopolysaccharide

NarL

nitrate/nitrite metabolism response regulator

NtrC

nitrogen regulatory response regulator

OmpR

osmotic pressure response regulator

PAS

Per-Arnt-Sim domain

RND

resistance nodulation and division

ROS

reactive oxygen species

RR

response regulator