The complex global response to copper in the multicellular bacterium Myxococcus xanthus

876 |Metallomics, 2018, 10, 876--886 This journal is ©The Royal Society of Chemistry 2018 Cite this: Metallomics, 2018, 10,876 The complex global response to copper in the multicellular bacterium Myxococcus xanthus Juana Pe ´rez, * Jose ´Mun ˜oz-Dorado and Aurelio Moraleda-Mun ˜oz The complex copper response of the multicellular proteobacterium M. xanthus includes structural genes similar to those described in other bacteria, such as P 1B -type ATPases, multicopper oxidases, and heavy metal efflux systems. However, the two time-dependent expression profiles of the different copper systems are unique. There are a number of genes responsible for an immediate response, whose expression increases after the addition of copper, but rapidly decreases thereafter to basal levels. The regulatory element that controls this early response is CorE, a novel extracytoplasmic function sfactor that is activated by Cu 2+ and inactivated by Cu + . Other genes are part of a maintenance response. These genes show a profile that slows up after the copper addition and reaches a plateau at 24–48 h incubation. Most of the genes involved in this response are encoded by the operon curA, which is regulated by the two-component system CorSR. Moreover, other genes involved in the maintenance response are regulated by different regulatory elements that remain unknown. Additionally, copper activates the transcription of the structural genes for carotenoid synthesis through a mechanism that requires the activation of the sfactor CarQ. Bearing in mind that M. xanthus is not very resistant to copper, it is speculated that the complexity of its copper response might be related to its complex life cycle. Introduction Copper is a micronutrient transition metal with important structural and metabolic roles as a prosthetic group in key metalloproteins involved in essential processes, such as respiration Departamento de Microbiologı ´a, Facultad de Ciencias, Universidad de Granada, E-18071 Granada, Spain. E-mail: [email protected] Left Jose ´Mun ˜oz Dorado, and right Juana Pe ´rez Prof. Jose ´Mun ˜oz Dorado studied Biology at the University of Granada (Spain) where he also completed his PhD on Myxobacteria. He then moved to the University of Medicine and Dentistry of New Jersey, United States, as a post-doctoral fellow from 1988 to 1992. Afterwards, he moved to Yokohama, Japan, where he worked on yeast for Kirin Brewery Co., from May 1992 to September 1993, when he returned to the University of Granada to work as Assistant Professor and later as Associate Professor. Since 2009 he is Full Professor of Microbiology at the University of Granada. He is the leader of the Prokaryotic Development Group, which research is focused on the complex multicellular lifestyle of Myxococcus xanthus, mainly on the roles of paralogous genes to cope with changing environments, such as nutrient abundance or the presence of metals. We are also interested in different aspects of the cooperative predatory activity of this bacteria. Dr Juana Pe ´rez studied Pharmacy at the University of Granada where he also completed his PhD. In 1989 she received a Fulbright Fellowship for working in enzymes involved in lignin degradation at the Forest Products Laboratory in Madison (WI). In 1993, she returned to Spain to work as postdoctoral researcher at the Institute of Parasitology and Biomedicine of the Spanish National Research Council (CSIC) and later to the University of Granada, where she worked as Assistant and Associate Professor. Since 2011 she is Full Professor of Microbiology at the University of Granada. Received 30th May 2018, Accepted 21st June 2018 DOI: 10.1039/c8mt00121a rsc.li/metallomics Metallomics MINIREVIEW Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025 This journal is ©The Royal Society of Chemistry 2018 Metallomics, 2018, 10, 876--886 | 877 and photosynthesis. 1 This is possible because copper is a redoxactive metal that fluctuates between an oxidized (Cu 2+ ) and a reduced (Cu + ) state. However, this high reactivity is also the reason why copper can be toxic to cells. This metal is able not only to generate reactive oxygen species (ROS), but it also binds tightly to the active sites of many metalloenzymes, leading to their inactivation. 2–4 To deal with this paradoxical role and to modulate the delicate cellular levels, all organisms, including bacteria, have developed a complex network of copper handling and trafficking mechanisms. 5,6 For maintaining copper homeostasis, bacteria activate the expression of specific genes that include efflux pumps, such as P 1B -type ATPases and heavy-metal efflux systems that belong to the resistance-nodulation-cell division (HME-RND) superfamily. 7–9 Since copper cannot be degraded, bacteria also activate metallochaperones for its sequestration and periplasmic multicopper oxidases (MCOs) to decrease the copper toxicity in the cells through the oxidation of toxic Cu + to less toxic Cu 2+ . All of these structural genes must be controlled by metal-sensing regulatory transcription factors, that is, by metalloregulatory proteins. 4,5,10,11 Myxococcus xanthus is a deltaproteobacterium with a distinctive biological cycle among the prokaryotes and it is considered one of the few bacteria that exhibits clear multicellular behavior. 12,13 On solid surfaces, rod-shaped vegetative cells move in coordinated groups to cooperatively prey on a wide variety of microorganisms. 14 In the absence of nutrients, the groups of cells move together with organized gliding movements and, at certain points, form macroscopic structures called fruiting bodies (Fig. 1). During this stage, three different subpopulations of cells show division of labor. A small percentage of cells differentiate from rods into spherical resistant myxospores, another fraction of cells remain as peripheral rods, and the majority of cells die, presumably to provide nutrients for aggregation and spore differentiation (Fig. 1). 13 M. xanthus is mainly found in soil and its life cycle must therefore be accomplished in this variable environment with many fluctuating elements such as metals. Copper is found naturally in soil at concentrations that vary from 13 to 24 mg kg 1 , although traffic and agricultural, industrial, mining and other domestic activities are contributing to the increase of this metal in the soil. 15 The M. xanthus copper response differs from other bacteria in many aspects. First, reflecting the metabolic and structural differences of the cells during their life cycle, developing and growing cells show different resistance to copper (60 mM and 900 mM, respectively). However, cells from both stages exhibit similar increased levels of metal resistance (up to 2000 mM) Fig. 1 Myxococcus xanthus multicellular life cycle. (A) Vegetative growth. On solid surfaces, rod-shaped vegetative cells (yellow rods) move in search of the prey in a coordinated manner, forming dynamic multicellular groups, which feed cooperatively preying on a wide variety of microorganisms (here represented as orange rods). Cells secrete hydrolytic enzymes and secondary metabolites that kill and lyse the prey (light grey rods). (B) Developmental cycle. In the absence of nutrients, groups of cells organize their movements and form macroscopic fruiting bodies with three different subpopulations. The majority of cells die (light yellow rods). A small fraction of cells remains as peripheral rods and the rest of the rod cells develop into resistant round myxospores (yellow balls). This complex cycle must be accomplished in the variable soil environment, where one of the fluctuating elements is copper. Dr Aurelio Moraleda Mun ˜oz studied Biology at the University of Granada (Spain) where he also received his PhD under the guidance of Professor Jose ´ Mun ˜oz-Dorado in 2003. He subsequently pursued post-doctoral research with Professor Lawrence J. Shimkets at the University of Georgia from March 2005 to April 2007, when he returned to the University of Granada to work as postdoctoral researcher. He became an Assistant Professor in the Department of Microbiology at the University of Granada in 2012. Aurelio Moraleda-Mun ˜oz Minireview Metallomics Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025 878 |Metallomics, 2018, 10, 876--886 This journal is ©The Royal Society of Chemistry 2018 when cultures are pre-incubated with copper, indicating an induction of a resistance phenotype by adaptation. 16 Second, although M. xanthus is not especially resistant to this metal compared to other bacteria, its genome encodes a large number of genes involved in copper homeostasis (Fig. 2). 16–18 Third, this extensive copper regulon is orchestrated by diverse and specific regulatory elements that allow M. xanthus to react to copper in a hierarchical way, with some of the genes involved in an immediate response and others in sustaining such a response. 19,20 And fourth, this bacterium accumulates carotenoids, probably to quench singlet oxygen generated by this metal, a mechanism not reported for any other bacterium. 21 In this review we focus on the most interesting aspects of the complexity of the M. xanthus copper response. Searching for systems involved in the M. xanthus global copper response An in silico analysis of the M. xanthus genome revealed a high number of genes involved in copper and/or other metal homeostasis, including three P 1B -type ATPases named copA,copB, and copC, 18 three MCOs named cuoA,cuoB, and cuoC, 16 and six HME-RND heavy metal efflux systems. 17 In accordance with the analysis of sequence signatures in transmembrane domains described previously, 22,23 three of those tripartite efflux complexes were classified as Cus systems (Cus1, Cus2, and Cus3), thought to be involved in the export of Cu + and other monovalent metals, while the other three were classified as Czc systems (Czc1, Czc2, and Czc3), thought to be involved in cobalt, zinc, and cadmium efflux. 17 Most of these twelve systems are encoded in two main regions of the M. xanthus genome. 17 One of the strategies used to study the role of these genes in copper homeostasis was to follow their expression profiles in the absence and presence of copper and other metals. These experiments revealed that many gene expressions were copper dependent and differentially regulated, suggesting a sophisticated network not only of structural proteins but also of metal sensors (Fig. 2). In most cases, developing cells reached the same levels of copper-dependent expression with 10-fold lower metal concentrations, 16–18 most likely due to metabolic and structural differences between cells in the two phases of the life cycle. Actually, the physiological stage of cells is an important factor in copper tolerance, as reported for differences between the aerobic and anaerobic copper response in Escherichia coli. 24 Two major expression profiles are clearly observed in genes up-regulated by copper in M. xanthus. One is exhibited by the MCO cuoB and the P 1B -type ATPase copB (Fig. 3A). These two genes are rapidly up-regulated upon the addition of copper, reaching a maximum at 2 h. Afterwards the expression decreases to a very basal level. The fact that deletion mutants in these two genes are markedly more sensitive to copper than the wild-type strain indicates that they represent the first line of defense Fig. 2 M. xanthus global copper response. The genomic environment of the genes and the sub-cellular localization of the proteins are represented. The MCOs (brown) CuoA, CuoB, and CuoC are periplasmic proteins that oxidize Cu + (grey balls) to the less toxic Cu 2+ (blue balls) through their cuprous oxidase activity. The P 1B -type ATPases (light green) CopA and CopB extrude Cu + from the cytoplasm to the periplasm. The HME-RND pumps, the Cus-like systems Cus2 and Cus3, detoxify the cytoplasm and the periplasm by pumping copper to the exterior. MXAN_3414 (purple) is a protein resembling the subunit III of the cbb3-type Cytochrome c oxidases. MXAN_3427 (pink) encodes a protein with a heavy metal-associated domain. The two regulatory elements (red) are the two-component system CorSR and the copper-dependent ECF sigma factor CorE. See text for details. Metallomics Minireview Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025 This journal is ©The Royal Society of Chemistry 2018 Metallomics, 2018, 10, 876--886 | 879 against the stress generated by copper, where they are involved in what has been called an immediate response (IR). In contrast, five other genes, cuoA,cuoC,copA,cus2, and cus3, show a very different profile during growth, in which expression goes up slowly after the copper addition, reaching maximum levels at 24 h (Fig. 3B). For this reason, these genes participate in the maintenance of the response (MR) to this metal. As with the IR genes, the developmental copper-responsive expression profiles for cuoA,copA,cus2, and cus3 are similar to those shown during growth, although with only 1/10 of the metal concentration in the medium. However, the MCO cuoC exhibits an induction during development which is independent of copper. 18 The phenotypes of the mutants indicate that all of these five genes are involved in conferring copper tolerance to M. xanthus. 16–18 Regarding the rest of the genes (copC,cus1,czc1,czc2, and czc3), they are not considered to participate in the copper response since they are not up-regulated by this metal and the deletion mutants do not exhibit a significant increase in copper sensitivity compared with the wild-type strain. 17,18 In summary, seven genes were initially identified as participating in conferring copper tolerance to M. xanthus. Two of them (copB and cuoB) participate in the IR, while the others (copA,cuoA, cuoC,cus2,andcus3) participate in the MR. The participation of three different periplasmic MCOs, which are able to oxidize Cu + to Cu 2+ , in the M. xanthus copper response might reflect the need for this bacterium to achieve a fine adjustment of the oxidation state of copper in the periplasm. 16 This may be related to the unique structure of the M. xanthus peptidoglycan, which is required to permit the conversion of the long vegetative rods into coccoid myxospores during development. 13 In fact, the accumulation of Cu + in the M. xanthus periplasm of the MCO mutants induces the morphology of the growing cells to change from rods to spherical forms. 16 In both responses (IR and MR), the coincidence in the expression profiles of a P 1B -type ATPase with a MCO (CuoB with CopB and CuoA–CuoC with CopA) (Fig. 3) indicates that they function together, with the Cop protein extruding Cu + from the cytosol to the periplasm, where the cognate Cuo proteins oxidize Cu + to the less toxic form Cu 2+ through their demonstrated cuprous oxidase activity. 16 CorE, a copper-dependent RNA polymerase sigma factor with a surprising mechanism of action, regulates the M. xanthus copper immediate response The analysis of the genomic environment of the IR genes cuoB and copB (Fig. 2) revealed a gene, MXAN_3426, with strong similarities to extra cytoplasmic function (ECF) sigma factors, which could be responsible for the IR. This gene has been named corE. RT-PCR assays have proved that corE forms an operon with cuoB and a gene (MXAN_3424) encoding a protein with similarities to outer membrane efflux proteins (OEPs). 19 OEPs are channels that allow the export of a variety of substrates and that may also be involved in the extrusion of copper from the periplasm to the exterior. 22 Moreover, gene MXAN_3427 encodes a protein with a heavy metal-associated domain (Fig. 2). Studies on CorE have demonstrated that it regulates the expression not only of cuoB and MXAN_3424, but also that of copB and MXAN_3427. Furthermore, CorE partially regulates the expression of the MR gene copA. 19 The rest of the genes analyzed (cuoA,cuoC,copC,cus1,cus2,cus3,czc1,czc2, and czc3) are not regulated by CorE. Therefore, a total of six genes have so far been found to participate in the IR, all of which are regulated by CorE (Table 1). One intriguing question is how CorE functions. CorE resembles ECF sfactors, which are small alternative sfactors, rather divergent in sequence, that contain only two regions (s2ands4) Fig. 3 The M. xanthus copper response is organized in a time-dependent mannerinanimmediateandamaintenance response. Specific b-galactosidase activities were measured in the cell extracts of strains harboring fusions between the promoter of each gene and lacZ, which were grown in the presence of 600 mM copper. (A) IR is represented by cuoB (blue) and copB (red) profiles. (B) MR is exemplified by cuoA (blue) and copA (red). Redrawn from 20 Sa ´nchez-Sutil et al., 2013 20 and Go ´mez-Santos et al., 2011 19 (with permission of the authors). Minireview Metallomics Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025 880 |Metallomics, 2018, 10, 876--886 This journal is ©The Royal Society of Chemistry 2018 out of the four found in the housekeeping sfactor. These two regions are required for interaction with the RNA polymerase core enzyme and recognition of the promoter. 25,26 ECF sfactors represent the third pillar of bacterial signal transduction, and they modulate transcription, alternating between an active and an inactive state. Canonical ECF sfactors are frequently regulated by a membrane-anchored protein known as an anti-s factor, which sequesters the sfactor in the absence of stimuli, thereby keeping it inactive. Upon receiving a proper signal, the anti-sis inactivated and releases the sfactor, which can now direct the expression of the target genes after the recruitment of the RNA polymerase. 25 Although many ECF sfactors are regulated through an anti-sfactor, different modes of regulating ECF sfactors’ activities have been described. Currently, ECF sfactors are classified into 50 groups. 26,27 Most ECF sfactors are co-expressed with their cognate antisfactor, 26 but in the case of CorE, there was no good candidate that might function as an anti-sfactor in the neighborhood region. Nevertheless, the anti-sfactor could be encoded in any other region of the M. xanthus chromosome. To answer the question of whether CorE works in cooperation with an unidentified anti-sfactor, CorE was over-expressed in a DcorE cuoB-lacZ strain, and the expression of cuoB was analyzed. It is assumed that sand anti-sfactors must be in a 1 : 1 ratio to properly function. Therefore, if the sfactor is over-expressed, this protein would be present in a higher quantity than the cognate anti-sfactor and, in consequence, the remaining free sfactor would function in the absence of stimulus, which in the case of CorE would be in the absence of copper. Qualitative and quantitative analyses demonstrated that over-expression of CorE did not lead to expression of cuoB in the absence of copper. Moreover, copper addition yielded very similar expression profiles of cuoB in the wild-type strain and in the strain where CorE was over-expressed, indicating that CorE functions without an anti-sfactor. 19 This observation raised the question of how CorE responds to copper addition. Several in vitro and in vivo analyses demonstrated that CorE undergoes a sensitization/desensitization process whereby the regulator is active during a limited period of time after the contact with copper, which is dependent on the redox state of copper (Fig. 4). These data allowed the elaboration of a model in which CorE is activated by Cu 2+ , which most likely changes its conformation to the active state to bind the DNA and to promote the transcription (Fig. 4C and D). Once the immediate response has been triggered by CorE, the reducing conditions of the cytosol would favor the formation of Cu + , inactivating the sfactor and terminating the immediate response (Fig. 4E). 19 According to this model, CorE must specifically recognize copper. The detailed study of the sequence of CorE revealed that it exhibits a C-terminus extension called CRD (for Cys rich domain) and a conserved CxC motif located between the s2 and the s4 regions (Fig. 4A and B). Studies using in-frame deletion mutants of the CRD and site-directed mutagenesis in different Cys residues demonstrated that both the CxC and CRD are essential for CorE activity. 19,28 Moreover, it was also demonstrated that Cys189 (Fig. 4A) is the main residue of the CRD involved in the sensitization of CorE, because this single mutation abolishes 93% of the metal response. Similarly, Cys184 (Fig. 4A) seems to be essential to the desensitization process because the expression profile of the genes regulated by the CorE C184A strain exhibited a sustained copper response over time, with no peak at 2 h. 19 In conclusion, CorE is a copper-dependent ECF sfactor that functions without an anti-sfactor, whose active and inactive states are regulated by a CxC motif and a CRD that detect the redox state of copper (Fig. 4). Based on the above results, CorE is not only the founder member of a new group of ECF sfactors, named CorE-like or ECF44, 27 but it also represents a new class of metal-sensing regulator. 11 The M. xanthus genome codes for a second ECF sfactor that also contains a CRD, which is located in a genetic environment containing genes that code for proteins with similarities to metaldetoxification proteins. 28 This second member of the group was named CorE2 (Fig. 4A). CorE2 exhibits several similarities Table 1 Predicted proteins of the locus involved in the immediate response regulated by CorE, and the curA operon involved in the maintenance response and controlled by CorSR. The domains in the database Pfam are in parentheses MXAN_ Protein Predicted function Predicted localization Immediate response (IR) 3415 CopA Copper-translocating P 1B -type ATPase (PF04945, PF00122, PF02954) Membrane protein 3422 CopB Copper-translocating P 1B -type ATPase (PF00403, PF00122, PF00702) Membrane protein 3424 3424 Outer membrane efflux protein (PF02321) External membrane protein 3425 CuoB Multicopper oxidase (PF07732, PF07731) Periplasmic 3426 CorE ECF sfactor (PF04542, PF08281) Cytoplasmic 3427 3427 Heavy-metal-associated domain (PF00403) Cytoplasmic Maintenance response (MR) 3413 3413 Hypothetical protein Periplasmic 3414 3414 Cytochrome c oxidase, cbb3-type, subunit III (PF13442) Periplasmic 3415 CopA Copper-translocating P 1B -type ATPase (PF04945, PF00122, PF02954) Periplasmic 3416 3416 Lipoprotein containing DUF305 domain (PF03713) Membrane protein 3417 3417 Lipoprotein containing two PQQ_2 domains (PF13360) Probably external membrane lipoprotein 3418 CorR s54-dependent DNA-binding response regulator (PF00072, PF00158, PF02954) Probably external membrane lipoprotein 3419 CorS Signal transduction histidine kinase (PF02518, PF00512, PF13581, PF00672, PF13589) Cytoplasmic 3420 CuoA Multicopper oxidase (PF07732, PF07731, PF00394) Membrane protein 3421 3421 Protein with His-rich region, cupredoxin1 (PF13473) Periplasmic Metallomics Minireview Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025 This journal is ©The Royal Society of Chemistry 2018 Metallomics, 2018, 10, 876--886 | 881 with CorE, such as functioning without an anti-sfactor and being regulated by metals. However, CorE2 also exhibits clear differences from CorE. Thus, cadmium and zinc, but not copper, activate this sfactor. Moreover, although genes regulated by this sfactor also reach a maximum expression at 2 h, these levels decrease slowly thereafter, remaining quite close to the maximum even at 48 h. This difference in the expression profile can be explained by the fact that, in contract to copper, cobalt and cadmium are redox-inactive metals. Accordingly, once they enter the cytoplasm and activate CorE2, the activation will be maintained through time because the redox state of these metals will remain unaltered. 28 These data obtained from CorE2 reinforce the idea that the copper redox state is responsible for the activation and inactivation of CorE. 19,28 In order to look more closely at the mechanism of action of the two members of this new group, and especially to investigate how these two metal-sensing proteins distinguish between different metals, all the cysteines and other residues on CorE2 CRD were mutated. Although the Cys motif is well conserved within the two sfactors, there is one Cys (Cys174 in CorE2) that is not present in CorE, where an Ala is found (Fig. 4A). Interestingly, the mutant CorE2 C174A did not respond to cadmium, but only to copper. Moreover, the opposite change, that is, the replacement of the Ala185 with a Cys in the CRD of CorE, did not change the CorE affinity for copper. However, upregulation of the expression of genes regulated by CorE in the presence of cadmium was three times higher in the mutant A185C than in the wild-type strain. These data suggest that the Cys arrangement of the CRD determines the metal specificity. 28 Moreover, it is worth mentioning that unpublished results of our group in collaboration with S. Atrian and M. Capdevilla have revealed that the binding of Cu/Zn to CorE or Cd/Zn to CorE2 induces important conformational changes in the a-helices of these two sfactors, which most likely are responsible for modulating their signaling mechanism. A bioinformatic search has allowed the identification in different phyla of Bacteria 67 sfactors that exhibit a CorE-like architecture with a highly conserved Cys arrangement located at the C-terminus. Although most of the proteins belong to the order Myxococcales, they are also present in alfa, beta, and gamma Proteobacteria,Acidobacteria,Chlamydia/Verricomicrobia, and Spirochaeta. 28 The maintenance response is mainly modulated by the two-component system CorSR The maintenance response (MR) of M. xanthus originally comprised five genes (cuoC,cus2,cus3,cuoA, and copA). However, as the MCO cuoA and the P 1B -ATPase copA form an operon (curA) with another seven genes (Table 1), a total of twelve genes must be considered as part of the MR. The different profiles exhibited by genes of the MR compared to genes involved in the IR (Fig. 3) clearly show that this MR must be modulated by different metallosensor proteins. In fact, there are two genes in the curA operon that show similarities to a two-component system, called CorSR, where CorS is the histidine kinase (HK) and CorR is the response regulator (RR), which is responsible for the expression of the curA operon, but not of cuoC,cus2, or cus3. 20 Three lines of evidence reflect that although curA confers copper resistance during growth, it is especially important for the detoxification of the cells during development. First, DcorSR cells accumulate 13-fold more copper than those of the wildtype strain during development, but only 1.3-fold more during growth. Second, the effect of this mutation on fruiting body formation is more dramatic than on growth. Thus, adapted cells of the mutant tolerate half as much copper as the wildtype strain during growth. Conversely, although the mutant developing cells could tolerate similar copper concentrations than growing cells, they could build fruiting bodies with 25-fold less copper than the wild-type strain. Higher concentrations blocked aggregation but did not kill the cells. And third, the expression of developmental genes that are expressed late in development is clearly impaired in the DcorSR mutant. 20 Sequence analyses using different bioinformatic approaches revealed that CorS has a cytoplasmic region exhibiting modular organization with three conserved domains: a HAMP domain, Fig. 4 Mechanism of action of the new group of CorE-like metal-dependent ECF sfactors. (A) Domain architecture of the two ECF sfactors of M. xanthus. CorE and CorE2 have the domains s2(PF04542)ands4 (PF08281) distinctive of all ECF sigma factors. Between both domains, there is a conserved sequence containing a CxC motif that is essential for the activity. A characteristic CRD (for Cys rich domain) involved in metal detection is located at the C-terminal domain. (B) In the absence of metal, CorE remains inactive and it is unable to bind to DNA. (C and D) In the presence of Cu 2+ (blue balls), the protein turns active, allowing binding to the core RNA polymerase and DNA at the 35 and 10 regions, and starting transcription. (E) In the reducing environment of the cytoplasm, the redox state of copper rapidly changes to Cu + (grey balls) and the protein again changes its configuration to become inactive. Minireview Metallomics Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025 882 |Metallomics, 2018, 10, 876--886 This journal is ©The Royal Society of Chemistry 2018 a kinase core consisting of a dimerization and a histidine phosphotransfer domain (DHp), and the catalytic and ATPase domain (CA) (Fig. 5). The protein is anchored to the membrane through two predicted transmembrane-spanning domains and contains a periplasmic sensor domain that consists of 158 residues. The sensor domain only exhibits similarities with other periplasmic domains of HKs from different bacteria of the order Myxococcales. 29 The predicted secondary structure of this domain allowed it to be included in the ‘‘all-alpha’’ protein group, and it therefore forms part of the ‘‘all helical fold’’ sensor class. 30 This result is very interesting because this class includes HKs which are not involved in copper sensing, such as NarX from E. coli, 31 as well as TorS from E. coli and Vibrio parahaemolyticus. 32 In contrast, other sensor domains of HKs involved in bacterial copper homeostasis, such as CinS from Pseudomonas putida, 33 CopS from Synechocystis sp. 34 and CusS from E. coli 35 are included in the ‘‘mixed alpha-beta fold’’ class. 29 These observations indicate that the mechanism of signal transmission of CorS must differ from that of other copper sensor HKs. The difference could be due to the fact that the other sensors recognize Cu + , while CorS detects Cu 2+ . 16 To look more closely at this mechanism of signal transmission of CorS, the periplasmic sensor domain was dissected by constructing a series of systematic in-frame deletion mutations. A small region of 28 residues located at the N-terminal sequence of the a1-helix and 9 residues at the C-terminal sequence of the a4-helix of the periplasmic region have been found to be essential for copper-regulated gene expression (Fig. 5). 29 The analysis of site-directed mutagenesis mutants of the two His (His38 and His171) present in those regions and a third, well conserved, His (His34) that remained in all deletion mutants and that is also located in the a1-helix, showed that these three residues are essential for activation of this kinase by copper. In vivo interaction studies using a bacterial two-hybrid system revealed that although CorS forms a homodimer in the absence of copper, the addition of the metal favors dimerization. These results suggest that copper plays a role in both dimerization and transmission of the signal. The current hypothesis is that copper binding is achieved in the context of a dimeric sensor, where the His34, His38, and His171 of both monomers are implicated in copper coordination (Fig. 5). The changes induced by the copper binding might reinforce or stabilize the sensor dimerization, and such conformational changes would be transduced through the HAMP domain into the cytoplasmic domain. In those circumstances, the kinase activity mediated by the catalytic domain would be activated and CorS would be autophosphorylated. The phosphorylated CorS would allow the phosphotransfer to its cognate RR, CorR, which could then promote the transcription of the curA operon. 29 CorR is a DNA-binding protein that shows the typical domain architecture of Group I of the bacterial enhancer binding proteins (bEBPs). 20 For this reason, it may be predicted that CorR activates the expression of genes from promoters recognized by the core RNA polymerase associated with s54. 20,36 Four different sequences with similarities to promoters for s54 were identified just upstream of four genes within the curA operon, indicating that transcription activated by CorR/s54 could be initiated at different sites. 20 This fact would explain why the expression profiles observed for genes of this operon differ not only in the levels obtained for each one, but also in the timing of when the plateau is reached. 16,18,20 The complexity of the expression of curA is further supported by the fact that copA is regulated not only by CorSR, but also by CorE, 20 so this ATPase participates in both the MR and the IR. These results reflect the fact that M. xanthus needs a very precise adjustment of all the copper homeostasis mechanisms not only over time, but also throughout its entire multicellular life cycle. A noteworthy application of these studies, which has had great acceptance in the international scientific community studying myxobacteria, is the use of the strong copper-dependent promoter of cuoA toconstructasetofplasmidvectorsthatallowconditional copper-inducible gene expression in M. xanthus. 37,38 The biosynthesis of carotenoids and interconnections between different mechanisms involved in copper homeostasis Carotenoids are red colored isoprenoids frequently produced by photosynthetic bacteria to protect against photooxidative damage since they are able to quench singlet oxygen ( 1 O 2 ) and Fig. 5 M. xanthus CorS signal-transduction model. Schematic representation of the modular organization of the cytoplasmic portion of CorS, with the HAMP domain (red rectangle), and the DHp and CA domains typical of HKs. The four ahelices of the periplasmic sensor domain are represented as rectangles, colored in orange, light grey, dark grey, and green (helices 1, 2, 3, and 4, respectively). In the absence of copper CorS is a homodimer. His34 and His38, within the a1-helix, and His171, in the a4-helix, are involved in copper coordination. The presence of Cu 2+ (blue balls) enhances the tendency of the domain to dimerize and these changes help to transmit the signal through the HAMP domain to the catalytic domain. Metallomics Minireview Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025 This journal is ©The Royal Society of Chemistry 2018 Metallomics, 2018, 10, 876--886 | 883 other reactive oxygen species that can damage cell components. 39 Although M. xanthus is a non-phototrophic bacterium, blue light triggers a complex transcriptional response leading to the biosynthesis of carotenoids. 40–42 This process is regulated by a canonical ECF s/anti-sfactor system (CarQ–CarR), which works in coordination with the photoreceptor CarF acting as an anti–anti-sfactor. Genes coding for CarQ, CarS, and CarR form the operon carQRS, which is controlled by the light-inducible promoter P QRS . 43,44 Briefly, in the presence of light, CarR is somehow inactivated in a process that requires the action of the anti–anti-sfactor CarF, liberating the sfactor CarQ. Free CarQ recruits the core RNA polymerase, starting the transcription of the carotenoid biosynthesis structural genes. These genes are located at two unlinked loci, the carB operon (with six genes) and the gene crtI, which are regulated by the promoters P B and P I , respectively. 45–47 The P B promoter is down-regulated in the dark by the repressor protein CarA, which is encoded in the constitutively expressed carA operon, located just downstream of carB. 48,49 CarS mediates the derepression of the carB operon, which along with the CarQ-mediated activation of crtI leads to the accumulation of carotenoids (Fig. 6). However, it has been demonstrated that copper also induces carotenogenesis in the dark. 21 Genetic studies using a collection of mutants in structural and regulatory genes revealed that copper-induced carotenoid synthesis is carried out through transcriptional activation of the same structural genes activated by light (Fig. 6). All of the regulatory elements involved in the light-induced carotenogenesis also participate in the copper response, with the only exception being the carF product. In other words, copper seems to directly inactivate CarR. 21 However, some aspects of the carotenoid response to copper remain to be elucidated. For instance, how copper activates the biosynthesis of carotenoids, bearing in mind that it is neither induced by other metals nor by other compounds that generate oxidative stress such as hydrogen peroxide, t-butylhydroperoxide and paraquat. Although 1 O 2 was originally proposed as the most feasible candidate responsible for triggering this response, it has been reported that CarF mediates signaling by the 1 O 2 generated via photoexcited protoporphyrin IX. 50 However, copper-mediated carotenoid synthesis is independent of CarF, and therefore the mechanism used to transmit the signal to the CarQ–CarR system must be different. Another unanswered question is why the biosynthesis of carotenoids induced by copper is more effective when cells are at suboptimal growth conditions. The reasons for this physiological modulation of the carotenogenic response could be related to the activity of some of the proteins described above as being involved in copper homeostasis. In this respect, it is worth mentioning that the mutant DcorSR accumulates carotenoids at lower copper concentrations than the wild-type strain due to the lack of expression of MXAN_3414 (Fig. 2 and Table 1). The DMXAN_3414 strain not only induces the biosynthesis of carotenoids at lower copper concentrations during growth, but it is the only strain that builds red fruiting bodies during development. 20 One reasonable explanation for this phenotype is that this gene could be required for energy production under copper stress. Therefore, the absence of this protein in the mutant could decrease energy levels, imitating suboptimal growth conditions and consequently inducing the biosynthesis of carotenoids at lower copper concentrations. In support of this hypothesis is the high homology exhibited by MXAN_3414 with the subunit III of cbb3-Cox (that belongs to the Pfam PF13442). Several authors have previously reported that in other bacteria, such as Pseudomonas fluorescens 51 and Salmonella typhimurium, 52 there is a connection between copper resistance, Cox biogenesis, and competitive fitness. Since it is usual in some bacteria to have several respiratory oxidases to allow cells to adapt their respiratory systems to the demands of a variety of environmental conditions, including different concentrations of metal ion cofactors, it is possible that the MXAN_3414 gene Fig. 6 Model summarizing some of the described elements of the regulatory network required for activation of the carotenogenesis in M. xanthus. Light excites protoporphyrin IX to a high-energy state, which then interacts with molecular oxygen to give 1 O 2 (yellow balls). 1 O 2 functions as the signal for the anti–anti-sCarF to interact with the anti-sCarR, which in turn liberates the ECF sfactor CarQ. After recruitment of the core RNA polymerase, CarQ activates the expression of crtI and the regulatory operon carQRS. CarS counteracts repression of P B by CarA, leading to the expression of the structural operon carB, which is responsible, along with the gene product of crtI, for the synthesis of carotenoids. The carotenogenesis induction by copper is CarF-independent, but genetic studies have demonstrated that the genes responsible for the synthesis of carotenoids are the same. Arrows indicate positive regulation; blunt-ended lines indicate negative regulation. See text for details. Minireview Metallomics Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025 884 |Metallomics, 2018, 10, 876--886 This journal is ©The Royal Society of Chemistry 2018 product could function as an alternative Cox subunit that promotes M. xanthus fitness in the presence of copper. Those results would explain why this metal induces carotenogenesis only in suboptimal growth conditions. This interconnection between the carotenogenesis response and curA again reveals the complexity of the adaptation mechanisms to copper in M. xanthus. The third question that arises is why non-carotenogenic mutants are as resistant (or even more resistant) to copper as the wild-type strain. This suggests that the synthesis of carotenoids itself does not contribute to copper tolerance in M. xanthus, in spite of the fact that the ability of the carotenoids to attenuate copper-induced oxidative damage has been clearly demonstrated. 53 However, another feasible explanation could be that the absence of carotenoids overstimulates the expression of some other mechanisms involved in copper tolerance. The interdependence between different systems is well documented in the case of high up-regulation of copA in a DcopB mutant or of cus2 and cus3 systems in the double mutant DcopA–DcopB. 18 In this latter case, the more than 10and 3-fold increase in the expression of Cus2 and Cus3 respectively not only compensates for the deficiency of the two ATPases, but it also yields a mutant strain that is more resistant to copper than the wild-type strain. 18 Concluding remarks and future perspectives The complex copper response of M. xanthus involves the participation of many genes, several of which are paralogs, that are finely controlled by different regulatory elements that sense copper in the cytoplasm or the periplasm and are able to differentiate between the two redox states of copper to act in a time-dependent manner. So far, only two regulatory elements, CorE and CorSR, have been identified and characterized. However, taking into consideration that cus2,cus3 and cuoC are up-regulated by copper, but are not under the control of either of the two identified elements, and that cuoC is not only up-regulated by copper, but also by starvation, there is evidence that at least two other transcriptional regulators must participate in this complex response. One intriguing question is why M. xanthus encodes such a high number of genes involved in conferring copper tolerance when it is not especially resistant to copper. It is tempting to speculate that such a high number of copper-detoxification systems could be necessary to successfully complete the multicellular lifestyle of this bacterium. Several data, such as the fact that growing and developing cells show different copper sensitivities and that defects in aggregation and sporulation are observed in some mutants, indicate that several of these elements are required for normal development in an environment where copper may be abundant. Nevertheless, it cannot be ruled out that M. xanthus may also use some of these elements during growth when cells prey on some other microorganisms, in the same way that eukaryotic predators and macrophage do. 54–56 This is an interesting question that needs to be addressed. In summary, all of these answered and unanswered questions indicate that M. xanthus’ copper response is carried out by a network of interconnected mechanisms that are precisely regulated by unique elements. Unraveling this tangle of proteins used by M. xanthus to adapt the different cell types to changing environments will contribute to our understanding of copper signaling in living cells. Conflicts of interest There are no conflicts to declare. Acknowledgements This work has been supported by the Spanish Government, grants CSD2009-00006 and BFU2012-33248 to Jose ´Mun ˜ozDorado, and BFU2016-75425-P to Aurelio Moraleda-Mun ˜oz (70% funded by FEDER). We would like to thank Silvia Atrian and Merce `Capdevilla for contributing to CorE metal binding studies (unpublished). References 1 C. Rensing and S. F. McDevitt, The copper metallome in prokaryotic cells, Met. Ions Life Sci., 2013, 12, 417–450. 2 S. Chillappagari, A. Seubert, H. Trip, O. P. Kuipers, M. A. Marahiel and M. Miethke, Copper stress affects iron homeostasis by destabilizing iron-sulfur cluster formation in Bacillus subtilis,J. Bacteriol., 2010, 192, 2512–2524. 3 L. Macomber and J. A. Imlay, The iron-sulfur clusters of dehydratases are primary intracellular targets of copper toxicity, Proc. Natl. Acad. Sci. U. S. A., 2009, 106, 8344–8349. 4 K. J. Waldron, J. C. Rutherford, D. Ford and N. J. Robinson, Metalloproteins and metal sensing, Nature, 2009, 460, 823–830. 5 P. Chandrangsu, C. Rensing and J. D. Helmann, Metal homeostasis and resistance in bacteria, Nat. Rev. Microbiol., 2017, 15, 338–350. 6 R. A. Festa and D. J. Thiele, Copper: an essential metal in biology, Curr. Biol., 2011, 21, 877–883. 7 J. M. Argu ¨ello, D. Raimunda and M. Gonza ´lez-Guerrero, Metal transport across biomembranes: emerging models for a distinct chemistry, J. Biol. Chem., 2012, 287, 13510–13517. 8 D. H. Nies, Efflux-mediated heavy metal resistance in prokaryotes, FEMS Microbiol. Rev., 2003, 27, 313–339. 9 J. A. Delmar, C. C. Su and E. W. Yu, Bacterial multidrug efflux transporters, Annu. Rev. Biophys., 2014, 43, 93–117. 10 K. J. Waldron and N. J. Robinson, How do bacterial cells ensure that metalloproteins get the correct metal?, Nat. Rev. Microbiol., 2009, 6, 25–35. 11 C. Rademacher and B. Masepohl, Copper-responsive gene regulation in bacteria, Microbiology, 2012, 158, 2451–2464. 12 P. Cao, A. Dey, C. N. Vassallo and D. Wall, How myxobacteria cooperate, J. Mol. Biol., 2015, 427, 3709–3721. Metallomics Minireview Downloaded from https://academic.oup.com/metallomics/article/10/7/876/5952465 by Universidad de Granada - Biblioteca user on 31 July 2025