Global copper response of the soil bacterial predator Myxococcus xanthus and its contribution to antibiotic cross-resistance

Global copper response of the soil bacterial predator Myxococcus xanthus and its contribution to antibiotic cross-resistance Francisco Javier Marcos-Torres , Juana P´ erez, David Torrens-Gonz´ alez, Miguel ´ Angel García-Pedrosa, Francisco Javier Contreras-Moreno, Aurelio Moraleda-Mu˜ noz * Departamento de Microbiología, Facultad de Ciencias, Universidad de Granada, Granada 18071, Spain ARTICLE INFO Keywords: Copper homeostasis Antibiotic co-resistance Bacterial predation Myxococcus xanthus Sinorhizobium meliloti ABSTRACT Copper accumulation in agricultural soils poses environmental challenges by selecting copper-resistant bacteria and also contributing to the co-selection of antibiotic-resistant bacteria. In addition, copper influences bacterial predator-prey interactions, potentially altering microbial ecosystems. Myxococcus xanthus, a soil-dwelling bacterium, preys on other microorganisms, including Sinorhizobium meliloti, a symbiotic nitrogen-fixing bacterium associated with leguminous plants. The role of copper in M. xanthus interactions remains poorly understood, although it accumulates at the predator-prey interface. In this study, we explore the transcriptomic response of M. xanthus to copper stress in both monocultures and co-cultures with S. meliloti. Our analysis identified many myxobacterial copper-regulated transcripts, and studies on mutant strains in some copper-induced genes revealed the role of two efflux pumps in cross-resistance to copper and tetracyclines. These findings provide new insights into the adaptive mechanisms of M. xanthus in response to copper, with implications for the co-selection of antibiotic resistance and the broader impact of copper on microbial community dynamics in soil ecosystems. 1. Introduction Myxococcus xanthus is a soil bacterium from the phylum Myxococcota, well-known for its complex multicellular life cycle. In the absence of nutrients, M. xanthus undergoes a unique developmental process in which individual cells aggregate and form macroscopic, fruiting bodies, which represent a key aspect of its survival strategies. These fruiting bodies are formed from three distinct cell subpopulations that differentiate and divide labor, exemplifying multicellular behavior among bacteria (Mu˜ noz-Dorado et al., 2016, 2019; Marcos-Torres et al., 2020). In the presence of other microorganisms, M. xanthus behaves as a cooperative epibiotic predator (P´ erez et al., 2016). The combination of secondary metabolites, secreted enzymes and contact-dependent killing approaches ensures that M. xanthus can successfully prey on a wide range of organisms and utilize their degradation products for its own growth and survival (Seef et al., 2021; Thiery et al., 2022; Contreras-Moreno et al., 2024b; Herrou et al., 2024). To efficiently navigate its environment, either in search of prey or when forming fruiting bodies, M. xanthus utilizes two distinct motility systems: adventurous motility (A-motility), which is associated with individual cell movement and social motility (S-motility), which governs group or coordinated displacement. Both motility systems have been shown to be required for efficient predation (P´ erez et al., 2014). Furthermore, transcriptomic analyses have evidenced that during predation, M. xanthus upregulates genes related to killing, lysing, and consuming prey, while it downregulates genes involved in the developmental cycle (P´ erez et al., 2022), which suggests a shift in metabolic and regulatory priorities, allowing M. xanthus to maximize its predatory efficiency. Several studies have indicated that metals commonly present in the soil environment play significant roles in modulating predatory interactions. Thus, while iron is well-studied for its effects on microbial competition and predation (Lee et al., 2020; Contreras-Moreno et al., 2024b), the role of copper remains more elusive, although previous studies have reported the accumulation of copper inside the phagosome of predatory protozoans to kill bacteria (German et al., 2013; Hao et al., 2016), as well as the participation of this metal in the predatory activity of the bacteria Cupriavidus necator (Seccareccia et al., 2016) and M. xanthus (Contreras-Moreno et al., 2020, 2024a). During the interaction of M. xanthus with Sinorhizobium meliloti, copper accumulates at the predator-prey interface and, although its role in this interaction is not fully understood, this copper accumulation induces melanin production in S. meliloti (Contreras-Moreno et al., 2020). * Corresponding author. E-mail address: [email protected] (A. Moraleda-Mu˜ noz). Contents lists available at ScienceDirect Microbiological Research journal homepage: www.elsevier.com/locate/micres https://doi.org/10.1016/j.micres.2025.128357 Received 29 July 2025; Received in revised form 22 September 2025; Accepted 29 September 2025 Microbiological Research 302 (2026) 128357 Available online 1 October 2025 0944-5013/© 2025 The Author(s). Published by Elsevier GmbH. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). Copper is an essential micronutrient with important biological functions, since its redox activity makes it both a critical cofactor for cuproenzymes and a potential toxin when present in excess (Festa and Thiele, 2011). Consequently, bacteria, like all living organisms, have evolved sophisticated mechanisms to manage the dual effects of copper. To maintain copper homeostasis, bacteria utilize various strategies, including copper-efflux pumps and detoxification systems, such as proteins that can convert toxic Cu⁺ back to less reactive Cu 2+ , and copper-binding proteins and chaperones that sequester copper to prevent it from interacting with sensitive cellular targets. Despite these protective mechanisms, the harmful oxidation produced by copper still induces the cell repair systems (Moraleda-Mu˜ noz et al., 2005; Rensing and McDevitt, 2013; P´ erez et al., 2018). The copper response in M. xanthus involves a sophisticated regulatory network that comprises many genes, the majority of which are clustered into two distinct chromosomal regions, copper region 1 (CR1) and copper region 2 (CR2) (Moraleda-Mu˜ noz et al., 2010a, 2010b; P´ erez et al., 2018). CR1 includes genes encoding for two HME-RND heavy metal efflux systems that are induced by copper, while CR2 houses many genes related to copper homeostasis that are under the control of two regulatory elements: the two-component system (TCS) CorSR and the copper-dependent extracytoplasmic function (ECF) sigma factor CorE (Fig. S1A). The histidine kinase CorS detects Cu 2+ in the periplasmic space and, along with the response regulator CorR, modulates a nine-gene operon involved in the maintenance response to copper (S´ anchez-Sutil et al., 2013, 2016). CorE senses the copper redox state in the cytoplasm and CorE-dependent genes are activated quickly upon exposure to copper (G´ omez-Santos et al., 2011), initiating an immediate detoxification response. This phased response helps M. xanthus to maintain copper homeostasis under varying levels of copper stress. In addition to the genes located in CR1 and CR2, the M. xanthus genome also contains other genes encoding HME-RND systems, which are induced by copper (Moraleda-Mu˜ noz et al., 2010a). Understanding how M. xanthus manages copper at the molecular level during predation could provide valuable insights into bacterial survival strategies and metal homeostasis during predatory interactions. This complex response is critical to survive in its natural habitat, the soil, where besides natural copper fluctuations, the concentration of this metal is increasing due to human activities, particularly in agricultural areas because of its use as a biocide. Copper accumulation in soils has been shown to significantly alter microbial communities and act as a selective pressure favoring the prevalence of certain genes, such as antibiotic resistance genes (ARGs), in microbial populations (Ballabio et al., 2018; Pan et al., 2023). In fact, a growing number of studies have demonstrated that the propagation of ARGs in different environments is driven by the presence of metals (Pal et al., 2017; Zhao et al., 2024). In particular, copper pollution has been shown to promote the development and propagation of antibiotic resistance, mainly by co-selection phenomena and by horizontal gene transfer (HGT) of resistance genes (Pal et al., 2017; Edet et al., 2023). Co-selection occurs when the presence of copper exerts a selective pressure on microbial communities, leading to the simultaneous selection of metal resistance genes and ARGs. This process can arise through three primary mechanisms: co-resistance (resistance genes to metals and antibiotics are physically located together on the same mobile genetic element), cross-resistance (resistance genes can confer resistance to both metals and antibiotics), and co-regulation (a regulatory element can simultaneously regulate the expression of genes for metals and antibiotics resistance). Moreover, copper pollution can also accelerate HGT of ARGs, further facilitating the dissemination of ARGs across different bacterial species (Vats et al., 2022). Furthermore, HGT can be facilitated by microbial interplays such as predator-prey interactions. M. xanthus prey on a broad array of organisms, including clinically relevant pathogens, pathogenic bacteria in biofilms and soil bacteria (P´ erez et al., 2011, 2014; Livingstone et al., 2017; Kamada et al., 2023; Arakal et al., 2025), which could act as a donor of genes that could change prey susceptibility to antibiotics. In fact, a recent study reported the exchange of ARGs across members of the Myxococcota and Pseudomonadota phyla (Brown et al., 2024). Moreover, some evidence point to a correlation between the presence of M. xanthus in natural soil communities and the abundance of antibiotic-resistant bacteria (Saha et al., 2025). As described above, M. xanthus has many copper-responsive genes involved in metal homeostasis, and their abundance could contribute to copper and antibiotic co-selection by cross-resistance or co-regulation events. In this study, we delved into the global response to copper in M. xanthus monocultures and during predation, and determined which copper-resistance genes of this bacterium contribute to bacterial resistance to antibiotics by co-selection. These results not only expand our knowledge of the copper response in soil bacteria but also shed light on the role of this metal in bacterial predation and identify new genes involved in metal-induced antibiotic co-selection that can help to combat antibiotic resistance, one of the biggest health challenges of the XXI century. 2. Materials and Methods 2.1. Bacterial strains, plasmids, and culture conditions Bacterial strains and plasmids used in this work are listed in Table S1. M. xanthus and S. meliloti strains were grown in CTT and TY broth, respectively, as previously described (P´ erez et al., 2022) with vigorous shaking at 30◦C. CTT agar plates containing Difco agar (Becton Dickinson, Sparks, MD, United States) were supplemented with kanamycin (80 µg/ml) or galactose (10 mg/ml), when necessary. 2.2. Construction of the in-frame deletion mutants Since previous studies of M. xanthus copper homeostasis were performed on the DZF1 strain, in this work we obtained in-frame deletion mutants in the DK1622 background using the previously reported methodology (Marcos-Torres et al., 2016). Briefly, plasmids listed in Table S1 were introduced into M. xanthus DK1622 by electroporation to obtain integration into the chromosome by homologous recombination. Kanamycin-resistant merodiploid strains were grown on 1 % galactose CTT agar plates without kanamycin to facilitate loss of the plasmid through a second homologous recombination. Kanamycin-sensitive and galactose-resistant strains were analyzed by Southern blotting to check for in-frame deletions. 2.3. Phenotypic characterization of the mutant strains Each phenotypical analysis has been performed at least two times to ensure the soundness of the outcomes. To analyze the response to the copper limitation of the ΔcopC strain, we monitored the OD 600 of three replicates of the wild-type (WT) and mutant strains inoculated at an initial optical density at 600 nm (OD 600 ) of 0.05 in the presence of different concentrations of tetrathiomolybdate (TTM), a Cu 2+ -specific chelator, after 24 h of growth. Similarly, copper resistance was assayed for the Δczc2 strain by analyzing the growth of the WT and mutant strains in the presence of different concentrations of CuSO 4 . Since M. xanthus has been reported to exhibit different copper resistance capability when pre-incubated with copper (S´ anchez-Sutil et al., 2013), growth was also tested with cells pre-adapted to 300 μ M CuSO 4 . Motility of the Δczc2 strain was tested by tracking the colony diameter of three replicates of 10 μ l drops at OD 600 of 15 of the WT and mutant strains in CTT agar plates with different concentrations of CuSO 4 . As previously described, adventurous motility was assayed in CTT plates with 1.5 % agar, while social motility was monitored in CTT plates with 0.5 % agar (G´ omez-Santos et al., 2012). The colony diameter was measured every 24 h for 120 h and used to calculate the colony expansion rates for each strain. Predatory assays to study the interaction between M. xanthus and F.J. Marcos-Torres et al. Microbiological Research 302 (2026) 128357 2 S. meliloti were performed as previously described (Contreras-Moreno et al., 2020), using cultures at OD 600 of 15 and 5, respectively, in media with and without 900 µM CuSO 4 supplementation. Plates from 8 replicates for each condition were incubated at 30ºC and pictures from representative samples were taken. A modified version of the Kirby-Bauer disk diffusion susceptibility method was used to assess antibiotic resistance using three replicates. Briefly, 2 ml of M. xanthus liquid cultures at OD 600 of 0.3 were mixed with 5 ml of CTT soft agar (0.8 % agar) and poured on CTT agar plates with or without 300 μ M CuSO 4 (or 50 μ M CuSO 4 for the assays with the copper-sensitive Δcus2 strain). After solidification of the overlays, 6 mm antibiotic disks (Oxoid Holdings Ltd, Thermo Fisher Scientific) were placed on top, and the plates were incubated at 30ºC for 72 h before measuring the diameter of the inhibition halo. The antibiotics tested and the amounts of each one are collected in Table S2. In addition, a minimum inhibitory concentration (MIC) assay was performed using three replicates to determine the susceptibility of WT and Δcus3 strains to nalidixic acid. Briefly, the myxobacterial strains were grown overnight in CTT medium and diluted at OD 600 of 0.05 in CTT medium supplemented with 300 µM CuSO 4 and 0, 25, 50, 100, 200, 400, or 800 µg/ml. Bacterial growth was assessed for each antibiotic concentration after an incubation period of 24 h at 30◦C and the MIC value for nalidixic acid was identified as the lowest concentration of the antimicrobial required to inhibit visible growth of the tested strains. 2.4. Preparation of samples for RNA-Seq To obtain the samples for transcriptomic analysis, M. xanthus DK1622 and S. meliloti Rm1021 WT strains were grown to OD 600 of 1. Then, the cultures were concentrated in TM buffer (10 mM Tris–HCl [pH 7.6], 1 mM MgSO 4 ) to a final OD 600 of 15. For each of the two replicates of the co-culture samples, three 10-µl drops of the rhizobial suspension were deposited on the surface of CTT agar plates with and without 300 µM CuSO 4 supplementation and allowed to dry. This metal concentration was selected as it does not affect physiological processes, such as cell viability or motility, during the growth of the DK1622 M. xanthus WT strain (G´ omez-Santos et al., 2012), but induces the expression of copper-responsive genes (Moraleda-Mu˜ noz et al., 2010a, 2010b). Next, 10-µl drops of the M. xanthus suspension were deposited on top of each of the rhizobial colonies of a subset of the plates to obtain samples of predatory and prey interacting cells (samples Mx_Sm and Mx_Sm_Cu), while for the monoculture samples, another subset of samples of M. xanthus were kept growing alone (samples Mx and Mx_Cu). After 2, 6, and 10 h of incubation, two replicates from each of the four conditions were harvested from the plates. Pellets from each sample were resuspended immediately into 0.5 ml of RNA Protect Bacteria Reagent (Qiagen, Hilden, Germany), incubated at room temperature for 5 min, and harvested by centrifugation at 5000x g for 10 min (4◦C). Next, after removal of the supernatant, the pellets were stored at −80◦C. 2.5. RNA extraction To purify RNA, the frozen pellets were thawed, and the cells were lysed for 10 min at room temperature with 250 μ l of 3 mg/ml lysozyme (Roche Diagnostic, Mannheim, Germany) and 0.4 mg/ml proteinase K (Ambion, Carlsbad, CA, United States) prepared in TE buffer (10 mM Tris-HCl; 1 mM ethylenediaminetetraacetic acid [EDTA], pH 8.0). RNeasy Mini Kit (Qiagen, Hilden, Germany) was used for RNA extraction, performing on-column DNase digestion with the RNAse-free DNase set (Qiagen, Hilden, Germany), eluting each sample in 50 µl of RNasefree water. 2.6. Library preparation, sequencing, and transcriptomic data analysis Total RNA samples were processed by Novogene [Novogene Europe, Cambridge, United Kingdom], including rRNA depletion from total RNA samples with the Illumina Ribo-Zero Plus rRNA Reduction Kit (Illumina, Inc.). The remaining RNA was processed according to the procedures described in P´ erez et al. (2022) providing, on average, 23.53 million raw reads and a genome coverage of 386.58x. This coverage dropped to 380.50x after raw reads filtering, which implies removing reads with adaptor contamination, with more than 10 % of uncertain nucleotides and/or with more than 50 % of low-quality nucleotides. FPKM (fragments per kilobase of transcript per million fragments mapped) normalization was used for comparison of samples (Table S3). The average FPKM values of the two replicates were used to calculate the Log2 fold change (Log2FC) and those genes with a padj<0.05 were considered differentially expressed genes (DEGs) (Table S4). 2.7. Statistical analysis Graphical data are expressed as the average ±standard deviation of three independent replicates. Statistical evaluations were conducted using GraphPap Prism 8.0.2. (GraphPad Software Inc.), and statistical significance between groups was assessed using an independent twotailed Student’s t-test, with a P-value of <0.05 indicating a significant difference. Graphical and heat map representations were created using Microsoft® Excel® for Microsoft 365 MSO v.2506. 2.8. Data availability Transcriptome sequencing data (raw-reads) can be found at https://www.ncbi.nlm.nih.gov/ under the Bioproject accession number PRJNA1164694. 3. Results and discussion 3.1. M. xanthus copper response in monocultures Copper response in M. xanthus was analyzed by comparing transcripts after culture in the absence and presence of 300 μ M CuSO 4 at three different time points (2, 6, and 10 h). In our studies, all DEGs with | Log2FC| >1 (Table S4) were manually analyzed. 3.1.1. Upregulated genes in the presence of copper are mostly involved in copper efflux and detoxification Among the most upregulated DEGs are those in CR1 and CR2, as well as the HME-RND efflux systems previously reported as being involved in copper homeostasis (Moraleda-Mu˜ noz et al., 2010a, 2010b) (Fig. S1A and B; Table S4). CR2 includes the CorE-regulated genes encoding the multicopper oxidase (MCO) CuoB, the P 1B -type ATPase CopB, an outer membrane protein OmpB, and a protein with a heavy metal-associated domain (MXAN_3427) (Fig. S1A). Transcriptomic data indicate that genes in the CorE regulon show an expression pattern typical of the immediate response to copper, exhibiting an early peak of expression that drops at later times (Fig. 1A). In CR2, the nine-gene operon curA is also located, which is under the control of CorSR and includes the genes encoding the MCO CuoA, the P 1B -type ATPase CopA, CorSR, and a protein that resembles the subunit III of the Cbb3-type cytochrome C oxidase (MXAN_3421) (S´ anchez-Sutil et al., 2013). Transcriptomic data confirm that genes regulated by CorSR are implicated in the maintenance response because the expression of the curA operon is maintained over time (Fig. 1A). The gene for a third MCO, cuoC, is also located near to CR2 and shows upregulation by copper as previously reported (S´ anchez-Sutil et al., 2007; Moraleda-Mu˜ noz et al., 2010a) (Fig. 1A and B). Our analyses also corroborate the strong upregulation of the genes coding for the efflux systems Cus2 and Cus3, and a minor induction for cus1 and czc2 genes (Fig. 1A). Cus1 was reported to be mainly involved in Zn 2+ homeostasis (Moraleda-Mu˜ noz et al., 2010a); therefore, copper induction of the cus1 cluster is likely to be unspecific. To further explore F.J. Marcos-Torres et al. Microbiological Research 302 (2026) 128357 3 the role of the Czc2 efflux pump in copper homeostasis, we obtained a Δczc2 in-frame mutant using the M. xanthus DK1622 WT strain as the genetic background. To analyze the survival to copper toxicity, we monitored the growth of the Δczc2 strain in liquid CTT medium supplemented with increasing copper concentrations. The results showed that the Δczc2 mutant exhibited the same survival phenotype as the WT strain in both copper-adapted and non-adapted cells (Fig. S2A). Given that high copper concentrations affect motility in M. xanthus (G´ omez-Santos et al., 2012), Aand S-motility were also evaluated. While A-motility was unaffected in the Δczc2 strain (Fig. S2B), this mutant displayed decreased S-motility when exposed to high copper concentrations (Fig. S2C). Furthermore, Δczc2 strain shows a slight Fig. 1. Genes coding proteins involved in M. xanthus copper homeostasis. A: Heatmap of the upregulated DEGs in M. xanthus monocultures and in co-cultures with S. meliloti at times 2 h, 6 h and 10 h. Shaded red gradient represents Log2FC in gene expression. B: Function and sub-cellular location of the upregulated proteins implicated in copper detoxification. OM: outer membrane; IM: inner membrane; HME_RND efflux pumps: Cus1, Cus2, Cus3 and Czc2; multicopper oxidases: CuoA, CuoB and CuoC; P 1B -type ATPases: CopA and CopB; outer membrane protein: OmpB. The TCS CorSR regulates CopA and CuoA (in green). The ECF sigma factor CorE controls CopB, CuoB, and OmpB (in blue). F.J. Marcos-Torres et al. Microbiological Research 302 (2026) 128357 4 defect in predation that becomes more apparent in the presence of copper (Fig. S2D and E). This suggests that, although not essential for copper survival, the Czc2 system contributes to the fitness of M. xanthus in the presence of this metal, potentially alleviating the negative effects of copper on physiological processes. 3.1.2. Downregulated genes unraveled a new system within CR1 likely involved in the acquisition of copper and two TCSs involved in redox homeostasis Among the downregulated genes in the presence of copper were those of the cluster MXAN_0976-MXAN_0979 located in CR1 (Figs. 2A and C, S1A and B). Bioinformatic analyses of the genes in this cluster revealed that MXAN_0976 and MXAN_0977 resemble mbnP and mbnH, respectively. These two-partner genes typically co-occur in bacteria and are referred to as the “metallo-mystery pair” (Manesis et al., 2021). MbnP belongs to a copper-binding proteins family, while MbnH is a member of the bacterial diheme cytochrome c peroxidases that has been shown to be located in the periplasm of M. xanthus (Bhat et al., 2011; Manesis et al., 2021). The MbnPH pair has been proposed to play a role in copper redistribution and acquisition, especially in methanotrophs where it is involved in copper release from the chalkophore methanobactin. Genes encoding for MbnPH pairs are commonly associated with genes encoding copper-binding periplasmic chaperones and membrane transporters (Manesis et al., 2021). In fact, two of the downregulated genes in M. xanthus are those coding for the P 1B -type ATPase CopC (MXAN_0979) and a chloride channel protein (MXAN_0978) (Fig. 2A). We previously obtained three different translational lacZ fusions in the copC promoter and none of them exhibited expression upon the addition of copper. Moreover, a ΔcopC strain did not show any phenotype in the presence of copper. We speculated at that time that CopC could be involved in copper uptake, but no expression was detected either after the addition of metal chelators (Moraleda-Mu˜ noz et al., 2010b). The current transcriptomic data have shown that, in fact, copper downregulates the expression of the transporter copC at 6 and 10 h (Fig. 2A). These results suggest an unknown mechanism of copper acquisition in M. xanthus in which MbnPH proteins would hypothetically act to sequester copper together with the transporters copC and/or MXAN_0978 (Fig. 2A and C; Table S4). To further investigate, a ΔcopC in-frame deletion mutant was obtained from the DK1622 strain and its phenotype was analyzed under copper-deficient conditions. After 48 h of incubation in the presence of tetrathiomolybdate (TTM), a specific Cu²⁺ chelator, the ΔcopC strain was slightly more sensitive than the WT strain to copper chelation (Fig. 2D). This would indicate that while copA and copB are involved in copper efflux, copC might be hypothetically involved in the response to copper starvation. Although there are some reports on ATPase-catalyzed copper uptake (Hassani et al., 2010), copper acquisition by P 1B -type ATPases has never been directly demonstrated and requires further investigation. Alternatively, CopC might be procuring copper to MnbPH in the periplasm, so they can reallocate the metal under copper starvation Fig. 2. Downregulation of M. xanthus genes in the presence of copper. Heatmap of downregulated DEGs in the copC cluster (A) and genes encoding two paralogous regBA-like TCS (B) in M. xanthus monocultures and in co-cultures with S. meliloti at times 2 h, 6 h and 10 h. Shaded blue gradient represents Log2FC in gene expression. C: Predicted function and sub-cellular location of proteins encoded by copC cluster genes that seem to be implicated in copper acquisition in the absence of this metal. Proteins have been colored as in Fig. S1 to better locate them in their genomic environments. D: Growth of WT and ΔcopC strains after 48 h in CTT media supplemented with different concentrations of the Cu 2+ chelator TTM. Data represent the average ±standard deviation of three replicates. F.J. Marcos-Torres et al. Microbiological Research 302 (2026) 128357 5 (Fig. 2D). Besides the genes in the CR1, two additional gene clusters (MXAN_6223-MXAN_6227 and MXAN_6976-MXAN-6981), both including genes that resemble the RegBA TCS from other bacteria, stand out among the most downregulated genes in the presence of copper (Fig. 2B; Fig. S1A and B). These TCSs are highly conserved global regulatory systems that control the redox state in a variety of energygenerating and energy-utilizing biological processes, such as electron transport and aerotaxis (Wu and Bauer, 2008). A set of genes coding for oxidoreductases and cytochrome c biogenesis proteins were also downregulated in the presence of copper (Table S4) and upregulated during predation, when both regBA clusters were also upregulated (P´ erez et al., 2022). Thus, it is tempting to speculate that these TCSs are regulating proteins implicated in electron transport under changed copper levels. 3.1.3. Oxidative stress response and copper toxicity When copper enters the bacterial cell, the reducing environment in the cytoplasm changes Cu 2+ to Cu + , which can then participate in Fenton-type reactions and the Haber-Weiss cycle to produce highly reactive oxygen species (ROS) that, in turn, react with lipids, proteins, and nucleic acids. Copper-induced oxidative stress can also lead to thiol oxidation, causing the inactivation of thiol-containing proteins and the depletion of glutathione (GSH), which plays a major protective role against heavy metal toxicity. However, one of the main copper-toxicity mechanisms is the displacement of iron from iron-sulfur clusters and protein mismetallation, resulting in their inactivation (Andrei et al., 2020; Li et al., 2021). Our transcriptomic data indicate that proteins containing thioredoxin domains, which are involved in the reduction of a number of enzymes maintaining thiol-based redox homeostasis (Tanifuji and Kimura, 2024), are the main defensive response against copper-induced oxidative stress. Of the M. xanthus genes encoding proteins with thioredoxin domains, MXAN_6156 and MXAN_7304 were strongly upregulated, while MXAN_0289, MXAN_0290, MXAN_2652 and MXAN_6848 were slightly induced (Fig. 3A and C; Table S4). To further alleviate thiol stress, in the presence of copper, M. xanthus also induces the genes coding for a GSH disulfide reductase (MXAN_2318) and a GSH S-transferase (MXAN_5471) (Fig. 3A and C), proteins that are commonly involved in the detoxification of oxidized GSH (Lu and Holmgren, 2014). The GSH response during copper-induced oxidative stress seems to be limited to maintaining the redox balance of the preexisting GSH rather than increasing the synthesis of new GSH (Okada and Kimura, 2022), since we did not observe significant changes in the expression of the Fig. 3. M. xanthus response to copper-induced oxidative stress. Upregulated DEGs involved in antioxidant systems, including genes of the thioredoxin system (Trx) and the glutaredoxin antioxidant system (Gor) (A), and ROS-scavenging enzymes (B) that contribute to redox homeostasis. C: Heatmap of DEGs involved in oxidative stress caused by copper in M. xanthus monocultures and in co-cultures with S. meliloti at times 2 h, 6 h and 10 h. Shaded red or blue gradients represent Log2FC upregulation or downregulation, respectively, in gene expression. F.J. Marcos-Torres et al. Microbiological Research 302 (2026) 128357 6 main genes of GSH synthesis, such as the glutamate-cysteine ligase genes gcl1 and gcl2 (MXAN_1276 and MXAN_5806), nor the gluthathione synthetase MXAN_0017. In the maintenance of redox homeostasis, we also observed the upregulation of M. xanthus genes encoding for ROS-scavenging enzymes, such as the catalase KatE (MXAN_6188) (Fig. 3B and C). Although copper Fenton-like reactions do not produce H 2 O 2 , the detoxification of superoxide ions by superoxide dismutases (such as those coded by the upregulated genes sodB and sodC) results in the production of H 2 O 2 . The gene cluster MXAN_1563–1564, which encodes for two alkyl hydroperoxide reductases (AhpD and AhpC, respectively), is also upregulated under copper-induced stress, and support catalases in reducing H 2 O 2 and organic hydroperoxides (Thomas et al., 2008) (Fig. 3B and C). As mentioned above, one of the main mechanisms of copper toxicity is the displacement of iron ions from the iron-sulfur clusters. This is especially the case for dehydratase enzymes that are part of many metabolic pathways (Macomber and Imlay, 2009). In fact, in the presence of copper, several genes encoding for dehydratases implicated in the production of M. xanthus secondary metabolites, such as myxovirescin (MXAN_3931–3950), myxochromide (MXAN_4077–4079), DKxanthene (MXAN_4293–4304) or myxalamide (MXAN_4525–4530), were downregulated (Table S4), suggesting that copper inhibits the synthesis of these products, although how copper affects these gene clusters at the transcriptional level is still unknown. Sulfur metabolism accomplishes essential cell functions, and it has been frequently reported to be involved in copper resistance and in the maintenance of redox homeostasis in cooperation with other detoxification systems (Huang et al., 2019; Yan et al., 2021). The thioredoxin-dependent adenylylsulfate reductase MXAN_2340, involved in assimilatory sulfate reduction to sulfide, as well as the gene cluster MXAN_0177–0179, encoding a sulfite exporter, a rhodanese-like domain-containing protein, and a metallo-hydrolase, respectively, were downregulated in M. xanthus in the presence of copper (Table S4). On the other hand, the cluster MXAN_4309–4316, including genes encoding a sulfate exporter and a DsrE family protein involved in sulfur oxidation, was upregulated in the presence of copper (Table S4). 3.2. M. xanthus copper response during predation Previously, we demonstrated that copper plays a role during the interaction between M. xanthus and S. meliloti (Contreras-Moreno et al., 2020). Here, we investigated how this metal influences this interaction by comparing myxobacterial DEGs in co-cultures at different time-points in the presence of copper with data obtained from the M. xanthus monocultures. As in monocultures, the highest induced genes were those directly implicated in copper efflux and detoxification, involving genes in CR1 and CR2, as well as the genes for the HME-RND systems Czc2 and Cus3, and the ferritin genes MXAN_6156 and MXAN_7304 (Figs. 1A, 3C, and S1; Table S4). When compared to the FPKM values of monocultures, the data reveal that the expression of most genes is not significantly altered by the presence of prey. However, the three genes coding for the Cus3 efflux system had lower FPKM values during predation in the presence of the metal (Fig. S3A; Table S3). Since predation has been shown to affect the spatial distribution of copper in M. xanthus colonies and the oxidative stress response (Contreras-Moreno et al., 2018), it is possible that the transcriptional mechanisms behind cus3 copper induction are affected by these conditions. On the other hand, the expression of several genes seems to be altered by copper only under predation conditions (Table S4). To discern whether these changes are caused by a mere alteration of the predatory expression pattern by the presence of copper, or if they are part of a more specific synergistic response, we compared the differential expression of these genes in M. xanthus predatosomes during predation on S. meliloti in the absence of copper (Fig. S4). One of the most interesting observations under these conditions was the upregulation after 2 h of the gene cluster MXAN_5264-MXAN_5265 coding for a protein with a cation efflux domain and a protein with a glyoxal oxidase domain, respectively (Fig. S4A; Table S4). This gene cluster is involved in metal handling and is regulated by the CorE-paralog CorE2 (MXAN_5263), which was previously found to be activated by zinc and cadmium rather than copper (Marcos-Torres et al., 2016). As stated above, metals such as copper or iron play an important role during predation, and both partners in this interaction can either compete to acquire them or defend themselves from metal toxicity, which is bound to alter intracellular metal fluctuations. This would explain why copper, either directly or indirectly, is able to activate CorE2 under these conditions. Whether copper-dependent expression of the CorE2 regulon exclusively during predation is a consequence of an intracellular metal imbalance, or if these genes are required to be expressed under these conditions, hinting at a more specific role during predation is still to be determined. After six hours, a gene cluster coding for a glycine betaine/L-proline ABC transport system (MXAN_2248–2251) was also specifically upregulated by copper during predation (Figs. S3B and S4A; Table S3 and S4). Both betaine and proline are osmotolerant molecules (Tripathi et al., 2022), and the upregulation of this transporter suggests a role in maintaining osmotic balance during predation, which seems to be of more relevance when interaction occurs in copper-rich environments. The transporter’s role in maintaining osmotic balance could be vital not only for survival but also for optimizing the predatory behavior of M. xanthus in these harsh conditions. Among the transcripts induced by copper exclusively during predation stand out two glycosyltransferases (MXAN_1422 and MXAN_2921), probably involved in modifying cell surface components, such as polysaccharides or lipopolysaccharides, which could play a role in maintaining structural integrity during predation or facilitating interactions with the prey. The selective induction of two metalloproteins (MXAN_2692 and MXAN_6714) in co-cultures with copper, which could play roles in metal binding, detoxification, or catalytic functions under metal stress, reinforces the idea that certain metalloproteins are essential for managing copper during predation. A cluster of genes involved in lipid biosynthesis (MXAN_6398–6404) is downregulated in monocultures but upregulated by copper during predation, reflecting how reactive lipid composition is to changes in environmental conditions. This result aligns with previous findings by P´ erez et al. (2022) and Jain et al. (2025), who reported similar metabolic shifts in response to predatory interactions. This upregulation of lipid biosynthesis genes only during predation could be indicative of several adaptive responses, such as increased membrane remodeling, energy storage, or signaling pathways. On the other hand, two transcriptional regulators (MXAN_6233 and MXAN_6479) and a cluster that includes the hybrid histidine kinase (MXAN_0095) and a methyltransferase domain-containing protein (MXAN_0096) were only induced in the presence of both copper and prey, and could be part of the signaling pathways involved in modulating stress response, metabolism adjustments, or interaction with S. meliloti under copper conditions (Fig. S4A; Table S4). Regarding the transcripts that are downregulated by copper during predation, it is interesting to note the downregulation of the catalase gene katE (MXAN_6188), which, as indicated above, is upregulated by copper in monocultures. KatE is the main catalase during both exponential growth and the stationary phase, as well as during fruiting body formation, and appears to be essential for M. xanthus viability (Kimura et al., 2022). The downregulation of the katE gene, along with that of many oxidative stress response genes in the context of predation and copper (Fig. 3C), suggests lower intracellular levels of H 2 O 2 , indicating that the combined response to both conditions enhances the oxidative stress fitness of M. xanthus during predator-prey dynamics. The only upregulated gene related to the oxidative stress response in this context is MXAN_4389 (Fig. S4A), which codes for the catalase KatB, which has been reported to be rapidly induced by H 2 O 2 (Kimura et al., 2022). The induction of this catalase at a later stage of predation in the presence of F.J. Marcos-Torres et al. Microbiological Research 302 (2026) 128357 7 copper suggests that it may play a protective role by facilitating the fast degradation of external H 2 O 2 permeating through the cell membrane, as suggested by Kimura et al. (2022). We previously reported that S. meliloti responds actively to predation by upregulating H 2 O 2 producing genes (Soto et al., 2023), and it is possible that the presence of copper could further enhance the production of this ROS in the prey, which in turn may activate katB. Similarly, the cluster MXAN_2948-MXAN_2951, encoding for a metal ABC transporter system, is downregulated by copper only while preying on S. meliloti. This, together with the abovementioned lower expression levels of the Cus3 system and the catalase KatE, could suggest that activation of the copper response during predation may improve the predator physiological conditions, resulting in lower internal oxidative stress and a more efficient metal handling than predating in the absence of copper. 3.3. Copper interference with iron uptake and trafficking The expression profile of the iron uptake machinery in M. xanthus under copper stress, in both monocultures and co-cultures, reveals intriguing dynamics that highlight the intricate relationship between copper and iron metabolism. Most of the genes involved in iron competition during myxobacterial predation (Contreras-Moreno et al., 2024b) show a distinct pattern of downregulation in monocultures at 2 and 6 h, followed by an increase in expression at 10 h (Fig. S5), which suggests that the presence of copper significantly impacts iron homeostasis. This observation can be explained by several mechanisms based on the synergistic iron-copper toxicity: i) Cross-talk between iron-regulatory proteins and copper. The major iron homeostasis regulator FurA in M. xanthus responds to iron availability and controls the expression of iron-uptake genes (Contreras-Moreno et al., 2024b). However, high copper levels can cause the mismetallation of FurA and/or other iron-sensing proteins, disrupting their function and leading to the improper regulation of iron homeostasis genes. ii) Copper and iron are known to catalyze the formation of ROS, causing significant oxidative stress. In response to the excess of copper, M. xanthus must tightly regulate the levels of both metals to avoid oxidative damage. Additionally, copper-induced ROS (e.g., superoxide and hydrogen peroxide) can also disrupt the proper metallation of iron-dependent enzymes and transcriptional regulators, further complicating iron metabolism (Imlay, 2014). The temporary downregulation of iron uptake at earlier stages (2 and 6 h) may serve as a protective mechanism against iron-copper synergistic toxicity. iii) Copper and siderophores linkage. M. xanthus produces myxochelin, a catecholate-type siderophore that participates in iron acquisition. Interestingly, some siderophores, such as yersiniabactin from Escherichia coli and enterobactin from Salmonella enterica, can also bind Cu 2+ , enhancing copper resistance and preventing its reduction to the more toxic Cu + form (Hyre et al., 2021). Similar to other bacteria, such as Pseudomonas aeruginosa, which increases its requirement of iron to survive under copper stress (Chillappagari et al., 2010), M. xanthus may also face a higher demand for iron during copper exposure. Moreover, the presence of copper during predation on S. meliloti further alters the expression pattern of iron uptake genes, shifting toward an earlier upregulation of these mechanisms, denoting how critical iron competition becomes for M. xanthus during predation (Fig. S5). This ensures that M. xanthus can not only withstand copper stress but also outcompete S. meliloti for limited iron resources (Contreras-Moreno et al., 2024b). In fact, the competition for iron intensified as predation progressed, requiring an increase in iron uptake at later time points (Fig. S5). Moreover, it has been recently demonstrated that high iron concentrations prevent cellular copper accumulation and toxicity in Caulobacter, where some proteins involved in iron transport provide protection against copper excess (Cherry et al., 2025). 3.4. Copper and antibiotic co-selection Since soil environments can reach high copper concentrations from the use of this metal in agriculture as a biocide and from natural sources (up to 496 and 1817 mg Cu kg −1 has been reported in the topsoil of some European [Ballabio et al., 2018] and southern Brazilian [Korchagin et al., 2020] agricultural sites, respectively), we wanted to investigate how the copper response in M. xanthus could contribute to the selection of ARGs that could spread among soil populations. For this purpose, the M. xanthus WT strain was grown in liquid medium and the susceptibility of the cells to a panel of 11 antibiotics (Table S2) with different mechanisms of action was analyzed in media supplemented with 300 μ M CuSO 4 . The results obtained indicate that, in the presence of copper, M. xanthus WT cells increased their susceptibility to kanamycin, and also slightly to tetracycline and nalidixic acid (Fig. 4A). These results imply that copper, even below toxic levels, enhances the effectiveness of these antibiotics against M. xanthus, suggesting a synergistic effect of copper with antibiotics as biocides. This, however, can also promote the expression and selection of ARGs at lower antibiotic concentrations to survive under these conditions. As discussed above, the copper response is a multifactorial process involving the differential expression of hundreds of genes and, while some of those genes may increase antibiotic sensitivity, other genes may contribute to increase resistance and thus be susceptible to spreading antibiotic resistance among soil populations. Since the identified genes involved in the copper response of M. xanthus are not located in mobile genetic elements, copper and antibiotic resistance co-selection can only be mediated by crossresistance and co-regulation events in this bacterium. To approach the co-selection of copper and antibiotic resistance genes, we studied the contribution to antibiotic resistance of the 7 efflux systems (for crossresistance) and 2 transcriptional regulators (for co-regulation) that have been shown to be induced by copper in M. xanthus. For this purpose, we utilized deletion mutants for the genes encoding the P 1B -type ATPases CopA, CopB, and CopC, the efflux systems Cus1, Cus2, Cus3 and Czc2, and the transcriptional regulators CorSR and CorE (Table S1), and analyzed their resistance in the presence of 300 µM CuSO 4 against the different antibiotics used previously with the M. xanthus WT strain (Table S2). The analysis of the antibiotic susceptibility of M. xanthus mutant strains for the copper-responsive regulators CorSR and CorE revealed small differences from the WT strain (Fig. S6A). Thus, the ΔcorSR strain was more sensitive to kanamycin than the WT strain. Similarly, a mutant in the ATPase gene copA, regulated by CorSR, also displayed a slight increase in sensitivity to kanamycin compared with the WT strain (Fig. S6B), suggesting that CorSR affects resistance to this antibiotic via copA induction. In a similar manner, the ΔcorE strain and the mutant for the CorE-regulated ATPase gene copB were more sensitive to erythromycin (Fig. S6A and B). The ΔcorE mutant was also slightly more sensitive to moxifloxacin (Fig. S6A), which would point to an additional gene in the CorE regulon involved in the resistance to this antibiotic. As for the third ATPase gene, copC, the mutant showed no notable changes in antibiotic sensitivity, indicating that this efflux protein is not involved in copper-antibiotic cross-resistance (Fig. S6B). On the other hand, the four efflux pumps involved in copper homeostasis showed a more significant role in antibiotic resistance. Among them, the Δczc2 mutant only displayed a low increase in resistance to nalidixic acid, moxifloxacin, and erythromycin (Fig. 4D). As discussed above, this efflux pump plays a minor role in copper resistance, mainly affecting social motility in the presence of copper. This slight increase in the antibiotic resistance of the Δczc2 mutant may be due to the upregulation of other copper-resistance mechanisms that may confer resistance to these antibiotics. The Cus1 and Cus2 efflux pumps exhibited a striking copper-antibiotic cross-resistance, as Δcus1 and Δcus2 strains are more susceptible to tetracycline antibiotics (doxycycline, tetracycline, and oxytetracycline) compared with the WT strain (Fig. 4B and C). These results are consistent with previous reports of copper and F.J. Marcos-Torres et al. Microbiological Research 302 (2026) 128357 8 tetracycline cross-resistance, and with efflux being one of the most common mechanisms for tetracycline resistance, with RND transporters commonly mediating the intrinsic efflux of tetracyclines (Grossman, 2016). Additionally, the Δcus2 strain showed increased sensitivity to kanamycin as well, highlighting the potential involvement in broader antibiotic resistance mechanisms for the Cus2 efflux system. It should be noted that the assays with the Δcus2 strain were performed in media supplemented with a lower copper concentration (50 µM CuSO 4 ) because of the higher sensitivity of this mutant to the presence of the metal (Moraleda-Mu˜ noz et al., 2010a). An unexpected finding involved the highly copper-induced Cus3 pump. Contrary to the Δcus2 strain, although the Δcus3 strain is sensitive to high copper concentrations (Moraleda-Mu˜ noz et al., 2010a), it does not display altered susceptibility to the copper concentrations assayed here. Surprisingly, this mutant was fully resistant to nalidixic acid under these conditions (Fig. 4B). To quantifiy this striking result, MIC for nalidixic acid was determined to compare the level of resistance in the WT and Δcus3 strains in the presence of copper, obtaining MIC values of 50 and 800 µg/ml, respectively. Since in the absence of copper, when cus3 is not induced, M. xanthus shows susceptibility to nalidixic acid (Fig. 4A), this result suggests that Cus3 disruption leads to changes that inadvertently provide resistance to nalidixic acid. Although to a lesser extent, our results also show that the Δcus3 strain is also more resistant to erythromycin, suggesting that higher antibiotic resistance is related to the physiological state of the strain rather than specifically counteracting the mode of action of these antibiotics. This mechanism has been observed in other bacterial species, where disruptions in certain cellular systems can enhance resistance to antibiotics by altering membrane permeability or energy dynamics. In previous studies, similar disruptions in membrane-associated systems led to antibiotic resistance through indirect pathways (Wales and Davies, 2015). This may explain the unexpected resistance of the Δcus3 mutant, highlighting a possible compensatory mechanism within M. xanthus cells. In fact, we previously demonstrated that the deletion of some copper homeostasis genes, such as those for the CopA or CopB ATPases, induces a compensatory mechanism that alters the expression profiles of other genes, such as those for the Cus2 and Cus3 systems (Moraleda-Mu˜ noz et al., 2010b). In summary, this analysis highlights the complex interplay between copper homeostasis and antibiotic resistance in M. xanthus, with the metal-efflux pumps Cus1 and Cus2 being the main players influencing cross-resistance to antibiotics. Further research is needed to elucidate the precise mechanisms by which these Cus-type pumps contribute to copper-antibiotic interactions. 4. Conclusions The complex M. xanthus copper response includes an unusual number of inducible elements involved in efflux (two ATPases and four RNDtype systems), copper oxidation (three MCOs), and several copperbinding proteins and outer membrane proteins. Transcriptomic data allowed the identification of a cluster, that includes the gene for the ATPase CopC, which may be involved in copper redistribution and/or uptake under copper-restricted conditions. Additionally, the data unveiled a multifaceted response to oxidative stress, elicited mainly through the activation of several thioredoxins and interferences in iron metabolism. Fig. 4. Copper-antibiotic cross-resistance in M. xanthus strains. A: Antibiotic sensitivity of the WT strain in the absence or presence of 300 µM CuSO 4 . B: Antibiotic sensitivity of the WT, Δcus1 and Δcus3 strains in media supplemented with 300 µM CuSO 4 . C: Antibiotic sensitivity of the WT and Δcus2 strains growing in the presence of 50 µM CuSO 4 . D: Antibiotic sensitivity of the WT and Δczc2 strains in the presence of 300 µM CuSO 4 . Antibiotic tested: NA, nalidixic acid; CIP, ciprofloxacine; LEV, levofloxacin; MXF, moxifloxacin; DO, doxycycline; TE, tetracycline; OT, oxytetracycline; K, kanamycin; E, erythromycin; AZM, azithromycin; SP, spiramycin. See Table S2 for the concentrations assayed. Three independent replicates were analyzed, and data represent average ±standard deviations. The twotailed Student’s t-test was used to determine significant differences (*: P<0.05; ** : P<0.01; *** : P<0.005; **** : P<0.001). F.J. Marcos-Torres et al. Microbiological Research 302 (2026) 128357 9