Extracellular vesicles as an alternative copper-secretion mechanism in bacteria

Journal of Hazardous Materials 431 (2022) 128594 Available online 26 February 2022 0304-3894/© 2022 Published by Elsevier B.V. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/). Extracellular vesicles as an alternative copper-secretion mechanism in bacteria Steeve Lima a , b , c , Jorge Matinha-Cardoso a , b , Joaquín Giner-Lamia d , e , f , Narciso Couto g , Catarina C. Pacheco a , b , Francisco J. Florencio d , h , Phillip C. Wright i , Paula Tamagnini a , b , j , Paulo Oliveira a , b , j , * a i3S – Instituto de Investigaç˜ ao e Inovaç˜ ao em Saúde, Universidade do Porto, Rua Alfredo Allen, 208, 4200-135 Porto, Portugal b IBMC – Instituto de Biologia Molecular e Celular, Universidade do Porto, Rua Alfredo Allen, 208, 4200-135 Porto, Portugal c MCbiology Doctoral Program, ICBAS – Instituto de Ciˆ encias Biom´ edicas Abel Salazar, Universidade do Porto, Rua de Jorge Viterbo Ferreira, 228, 4050-313 Porto, Portugal d Instituto de Bioquímica Vegetal y Fotosíntesis, Universidad de Sevilla-CSIC C, Am´ erico Vespucio, 49, 41092 Sevilla, Spain e Centro de Biotecnología y Gen´ omica de Plantas, Universidad Polit´ ecnica de Madrid (UPM), Instituto Nacional de Investigaci´ on y Tecnología Agraria y Alimentaria (INIA), Parque Científico y Tecnol´ ogico, UPM Campus de Montegancedo, Ctra, M-40, km 38, 28223 Madrid, Spain f Departamento de Biotecnología-Biología Vegetal, Escuela T´ ecnica Superior de Ingeniería Agron´ omica, Alimentaria y de Biosistemas, Universidad Polit´ ecnica de Madrid (UPM), Campus, Av. Puerta de Hierro, nº 2, 4, 28040 Madrid, Spain g Department of Chemical and Biological Engineering, University of Sheffield, Mappin St, Sheffield City Centre, Sheffield S1 4NL, United Kingdom h Departamento de Bioquímica Vegetal y Biología Molecular, Facultad de Biología, Universidad de Sevilla, Avenida Reina Mercedes s/n, 41012 Sevilla, Spain i University of Southampton, Office of the President and Vice Chancellor B37, University Rd, Highfield, Southampton SO17 1BJ, United Kingdom j Departamento de Biologia, Faculdade de Ciˆ encias, Universidade do Porto, Rua do Campo Alegre s/n, 4169-007 Porto, Portugal HIGHLIGHTS GRAPHICAL ABSTRACT •Extracellular vesicles (EVs): overlooked metal resistance mechanism in bacteria. •Vesiculation is a common response mechanism to copper-induced stressful conditions. •Cyanobacterial EVs are packaged with metal binding proteins. •Cyanobacterial EVs work as coppersecreting nanocapsules. ARTICLE INFO Keywords: Metal homeostasis Copper detoxification mechanisms Copper secretion Cyanobacteria Bacterial extracellular vesicles ABSTRACT Metal homeostasis is fundamental for optimal performance of cell metabolic pathways. Over the course of evolution, several systems emerged to warrant an intracellular metal equilibrium. When exposed to growthchallenging copper concentrations, Gram-negative bacteria quickly activate copper-detoxification mechanisms, dependent on transmembrane-protein complexes and metallochaperones that mediate metal efflux. Here, we show that vesiculation is also a common bacterial response mechanism to high copper concentrations, and that extracellular vesicles (EVs) play a role in transporting copper. We present evidence that bacteria from different ecological niches release copious amounts of EVs when exposed to copper. Along with the activation of the classical detoxification systems, we demonstrate that copper-stressed cells of the cyanobacterium Synechocystis sp. PCC6803 release EVs loaded with the copper-binding metallochaperone CopM. Under standard growth conditions, CopM-loaded EVs could also be isolated from a Synechocystis strain lacking a functional TolC-protein, which we characterize here as exhibiting a copper-sensitive phenotype. Analyses of Synechocystis tolC-mutant’s * Corresponding author at: i3S – Instituto de Investigaç˜ ao e Inovaç˜ ao em Saúde, Universidade do Porto, Rua Alfredo Allen, 208, 4200-135 Porto, Portugal. E-mail address: [email protected] (P. Oliveira). Contents lists available at ScienceDirect Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat https://doi.org/10.1016/j.jhazmat.2022.128594 Received 16 September 2021; Received in revised form 18 February 2022; Accepted 24 February 2022 Journal of Hazardous Materials 431 (2022) 128594 2 EVs isolated from cells cultivated under standard conditions indicated the presence of copper therein, in significantly higher levels as compared to those from the wild-type. Altogether, these results suggest that release of EVs in bacteria represent a novel copper-secretion mechanism, shedding light into alternative mechanisms of bacterial metal resistance. 1. Introduction Copper is an essential micronutrient. Many proteins and enzymes, such as cytochrome oxidases, laccases, superoxide dismutases, plastocyanin, or nitrite reductase (Argüello et al., 2013), which are involved in a variety of key biological functions in most living organisms require copper as a redox cofactor (Xu et al., 2021). Nevertheless, the ability to undergo redox reactions makes free copper a potential hazard (Argüello et al., 2013), catalyzing the formation of highly reactive hydroxyl radicals (Gaetke and Chow, 2003). This is particularly relevant when organisms are exposed to high copper concentrations. In such conditions, copper may also attack and destroy iron-sulfur clusters, thereby releasing iron, which can in turn cause further oxidative stress (Rensing and McDevitt, 2013), leading to severe cell damage. Due to its toxicity in high concentrations, copper can be used to control unwanted microbial growth. This feature has been streamlined in the course of evolution: the innate immune system in mammals utilizes copper intoxication to reduce intracellular survival of bacterial pathogens (Djoko et al., 2015). Besides, several anthropogenic activities also employ copper-based solutions to control microbial biofilm formation, including in medical settings, industrial equipment and plumbing systems (Gomes et al., 2020). Even certain sectors of the agri-food industry, such as grape-wine production (EFSA et al., 2018) and aquaculture (Yanong, 2010) resort to copper-based compounds to tackle fungal and bacterial growth. Despite the toxicity and risk of contamination, copper-based compounds are easy and cheap to obtain, do not discolor the water, and are one of the few effective options for organic growers, which still motivates their use (Romanazzi et al., 2020; Yanong, 2010). Throughout evolution, bacteria have developed copper-transport systems tightly controlling uptake and secretion, but also trafficking through different compartments, promoting a highly regulated balance of copper metabolism (Argüello et al., 2013; Rensing and McDevitt, 2013). Particularly, when exposed to high copper levels, bacteria have shown to activate several systems that effectively help in copper efflux, reducing its noxious effects intracellularly. In spite of different bacteria exhibiting different combinations of copper-efflux components (Argüello et al., 2013), the two essential mechanisms for copper export rely on: (i), transmembrane-protein complexes that cross the cell wall and pump copper from the cytoplasm, or periplasm in Gram-negative bacteria, directly into the extracellular space; and (ii), export of metallochaperones with copper binding capacity, which can work as chelating agents (Nies, 2003; Argüello et al., 2013; Rensing and McDevitt, 2013). Bacterial cells regulate these processes mainly by modifying transporter abundance through transcriptional control, which also facilitates co-regulation of chaperone and chelating metalloproteins (Argüello et al., 2013). Similar to higher organisms, bacteria release extracellular vesicles (EVs), which have been shown to play fundamental roles in bacterial cell survival, including secretion, communication, and nutrient uptake (Lima et al., 2020). Many studies on vesicles from Gram-negative bacteria predominantly discuss the classic concept of outer membrane vesicles (OMVs). However, there has been a trend in the past few years to refer generally to vesicles from Gram-negative bacteria as EVs (e.g. Biller et al., 2021; McMillan and Kuehn, 2021; Biller et al., 2022). This is mainly due to the fact that recent studies have shown additional routes for vesicle formation, which are not limited to OMVs (Toyofuku et al., 2019). Thus, in this study, we collectively refer to the vesicles produced by bacteria as EVs. The Cyanobacteria phylum is a large phylogenetic group of the domain bacteria, with the particular feature of being capable of performing oxygenic photosynthesis with water as an electron donor (Garcia-Pichel et al., 2020), similarly to green algae and plants. These photoautotrophic microorganisms play a key role in the biogeochemical cycle of carbon on a global scale, as estimates predict a global net primary production of 12 gigatonnes each year for marine cyanobacteria only, representing approximately 25% of ocean net primary production (Flombaum et al., 2013). The capacity of cyanobacteria to release EVs was documented in detail for the first time in 2014, and cyanobacterial EVs were proposed to function as vehicles for the movement of carbon through marine food webs, as vectors for horizontal gene transfer, and as decoys for predators and phages (Biller et al., 2014). More recently, the model cyanobacterium Synechocystis sp. PCC 6803 (hereafter Synechocystis) was also shown to vesiculate (Pardo et al., 2015), and a Synechocystis mutant strain, lacking a functional TolC-protein, was described to release substantially more EVs than the parental strain (Oliveira et al., 2016). TolC is an outer membrane protein, essential for the export of proteins (Oliveira et al., 2016; Agarwal et al., 2018; Gonçalves et al., 2018), fatty acids (Bellefleur et al., 2019), and exogenous compounds that reach the cytoplasm (Oliveira et al., 2016; Gonçalves et al., 2018). It was proposed that hypervesiculation in the Synechocystis tolC-mutant was a response to the impaired classical secretion (Oliveira et al., 2016). Nevertheless, deletion of inner membrane and periplasmic adaptor components of TolC-dependent efflux systems showed that none of those mutants reached the high vesiculation capacity presented by the tolC-deletion mutant (Gonçalves et al., 2018). Thus, it remains to be fully determined why the Synechocystis tolC-deletion mutant hypervesiculates. In this work, we carried out a comprehensive study of the Synechocystis tolC-mutant phenotype, in order to clarify why this strain releases more EVs than the wild-type. Combining a quantitative proteomic study with metal resistance tests, intracellular metal quantifications, and biochemical analyses, we show that the Synechocystis tolC-mutant accumulates copper intracellularly, uncovering the role of TolC in copper efflux in Synechocystis. More importantly, we show that EVs release is a common bacterial response upon exposure to high copper levels, and we detect the presence of a copper-binding chaperone in isolated cyanobacterial EVs. Moreover, detection and quantitation of copper in EVs isolated from the copper-sensitive Synechocystis tolC-mutant, cultivated under standard growth conditions, highlights that EVs in bacteria represent an overlooked copper-secretion mechanism. These results contribute to a deeper understanding of the intricate mechanisms warranting copper homeostasis in bacteria, further expanding the role of EVs in bacterial metal metabolism. 2. Experimental 2.1. Strains and growth conditions The unicellular cyanobacterium Synechocystis sp. PCC 6803 (hereafter Synechocystis) wild-type strain and respective mutants used in this work, namely ΔtolC, ΔcopB, Δsll1180 (or ΔhlyB), Δslr1053 (or ΔacrA), and EV-trc-GFP (see below) were routinely maintained in BG11 medium (Stanier et al., 1971) supplemented, when necessary, with appropriate antibiotics, in 100 mL Erlenmeyer flasks, with orbital shaking, under a regimen of 12 h light (25 µmol photons m −2 s −1 )/ 12 h dark at 30 ºC. Growth was assessed by measuring optical density at 730 nm (OD 730 ). Cyanobacterial strains were routinely streaked on solid BG11 medium S. Lima et al. Journal of Hazardous Materials 431 (2022) 128594 3 plates to assess the axenic state of the cultures. Unless stated otherwise, Synechocystis cells were inoculated at an initial OD 730 of 0.05 in BG11 and grown in glass gas washing bottles with aeration (approx. 1 L air min −1 ), under a regimen of 16 h light (45 µmol photons m −2 s −1 )/ 8 h dark at 28 ºC, for approximately 3.5 days (unless stated otherwise), reaching a final OD 730 of approximately 0.8–1.0. Escherichia coli [strain BL21 (DE3)] and Pseudomonas aeruginosa [PAO-1 (ATCC 47085)] were routinely maintained/streaked in LB medium agar plates. Both strains were grown over-night (ON) in 50 mL sterile conical centrifuge tubes, containing 5 mL of liquid LB medium, with orbital shaking (150 rpm for E. coli, or 200 rpm for P. aeruginosa) at either 37 ºC (E. coli) or 30 ºC (P. aeruginosa). ON-grown cultures were used to inoculate 200 mL of LB medium, LB supplemented with 0.5, 1 and 2 mM CuSO 4 for E. coli, or LB supplemented with 1 and 3 mM CuSO 4 for P. aeruginosa, in 1000 mL Erlenmeyer flasks with an initial OD 600 of 0.05, at the above-mentioned conditions for 15–16 h, reaching a final OD 600 of approximately 3.0–5.0. 2.2. Proteomics analysis by iTRAQ The proteomes of Synechocystis wild type and tolC-mutant were analyzed by 8-plex isobaric tags for relative and absolute quantification (iTRAQ), using four biological replicates. Cultures were grown in BG11 in 500 mL Erlenmeyer flasks, with orbital shaking, under a regimen of continuous light (35 µmol photons m −2 s −1 ) at 30 ºC. Upon reaching an OD 730 of approximately 1.0–1.5, cells were collected by centrifugation (4400 g, at room temperature), washed with potassium phosphate buffer (50 mM K 2 HPO 4 , 50 mM KH 2 PO 4 , pH 6.9) and stored at −80 ºC until further analysis. A detailed description of protein extraction and analysis can be found in the Supporting information file (Text S1). 2.3. Synechocystis metal resistance analysis To evaluate the metal resistance capacity of Synechocystis wild-type and mutant strains, cells were grown in solid BG11 (lacking sodium thiosulfate) supplemented with different metals, namely boron (463 µM H 3 BO 3 ), cadmium (5 µM CdN 2 O 6 .4 H 2 O), cobalt (5 µM Co (NO 3 ) 2 .6 H 2 O), copper (4 µM CuSO 4 ), iron (229 µM Ferric ammonium citrate), manganese (91.5 µM MnCl 2 .4 H 2 O), molybdenum (16.1 µM Na 2 MoO 4 .2 H 2 O), nickel (7 µM NiSO 4 .6 H 2 O), or zinc (14 µM ZnSO 4 ). Synechocystis cultures were grown as indicated in “Strains and growth conditions”. Later, cells were harvested by centrifugation (4400 g, 10 min at room temperature) and the cell pellet was suspended in fresh liquid BG11 medium, to a final OD 730 of 10. Serial dilutions were then prepared, and 15 µL of each cell suspension were spotted onto solid BG11 supplemented with each metal. Petri dishes were incubated for 7 days, under a regimen of 16 h light (25 µmol photons m −2 s −1 )/8 h dark at 28 ºC, and cell growth was registered by photographing the dishes. To quantify growth differences, a densitometry analysis was carried out using ImageJ (Schneider et al., 2012). 2.4. Intracellular metal quantification by atomic absorption spectrometry The intracellular levels of cobalt, copper, iron and zinc in Synechocystis wild-type and tolC-mutant strain were quantified by atomic absorption spectrometry (AAS). Briefly, cells were cultivated as described above and collected at an OD 730 of 0.8–1.0. Then, cells were pelleted by centrifugation at 4400 g for 10 min, and thoroughly washed using extraction buffer (20 mM Tris-HCl, 5 mM EDTA, pH 8.0). Later, cells were transferred to a screw-capped, 2 mL microcentrifuge tube, collected by centrifugation at 16,000 g for 1 min at 4 ºC, and washed twice again with extraction buffer for complete removal of BG11 traces. Finally, cell pellets were mixed with 100 µL of 65% HNO 3 , and incubated at 75 ºC for 12 min, vigorously vortexed every 2 min. Following this procedure, 1 mL of Milli-Q® Type 1 Ultrapure water was added, samples mildly vortexed, and centrifuged at 16,000 g for 5 min at room temperature. The resulting supernatant was collected for AAS analyses. Copper, iron and zinc levels were determined with an atomic absorption spectrometer coupled to a flame atomization detector (Perkin Elmer), while cobalt levels were determined with a spectrometer with an electrothermal detector (Perkin Elmer). For the various metals, quantifications were performed against calibration curves using specific aqueous pattern solutions. Results are presented in µg of metal L −1 culture OD 730−1 . 2.5. Total RNA extraction and analysis of transcript levels by RT-qPCR Sample collection, RNA extraction, all the control measurements and PCRs were performed as previously described (Gonçalves et al., 2018). For cDNA synthesis, 5 µg of total RNA was transcribed with the iScriptTM Advanced cDNA Synthesis Kit for RT-qPCR (Bio-Rad) in a final volume of 20 µL, following the manufacturer’s instructions. A control PCR was performed using 1 µL of cDNA as template in reaction mixtures containing: 0.5 U of GoTaq® G2 Flexi DNA Polymerase (Promega), 1x Green GoTaq Flexi buffer, 200 µM of each dNTP, 1.5 mM MgCl 2 , 0.25 µM of each rnpB primer (Table S1). For the determination of the primers’ efficiency, three-fold standard dilutions of the cDNAs were made (1/3, 1/9, 1/27, 1/81 and 1/243) and stored at −20 ◦C. RT-qPCRs were performed on Hard-Shell 384-Well PCR Plates (thin wall, skirted, clear/white) covered with Microseal® B adhesive seal (Bio-Rad). Ten microliter reactions were manually assembled and contained: 0.25 μ M of each primer (Table S1), 5 µL of iTaq™ Universal SYBR® Green Supermix (Bio-Rad) and 1 µL of template cDNA (dilution 1/9). The PCR profile was: 3 min at 95 ◦C followed by 45 cycles of 30 s at 95 ◦C and 30 s at 60 ◦C; and in the end a melting curve analysis (10 s cycles between 55 and 95 ◦C with a 0.5 ◦C increment per cycle). Primer efficiency standard curves and negative controls (no template cDNA) were included in all assays. RT-qPCRs were performed with three biological replicates and technical triplicates of each cDNA sample in the CFX384 Touch™ Real-Time PCR Detection System (Bio-Rad). The normalized expression of the target genes was calculated using the Bio-Rad CFX Maestro™ 3.1 software (Bio-Rad), implementing an efficiency-corrected delta-delta Cq method (ΔΔCq). Statistical analysis was performed by means of a one-way ANOVA using the same software, and tests were considered significant if P <0.05. Genes bfrA, sll0083 and slr0769 were validated as reference genes for data normalization (M value <0.5), using the reference gene selection tool available in the Maestro™ software. The size of the PCR products for each gene was checked by agarose gel electrophoresis performed by standard protocols. These products were also cloned into the pGEM-T® Easy (Promega) vector according to the manufacturer’s instructions. Identity of the amplicons was confirmed by Sanger sequencing (StabVida). These experiments were compliant with the MIQE guidelines (Bustin et al., 2009) to promote the effort for experimental consistency and transparency, and to increase the reliability and integrity of the results obtained. 2.6. Generation of the Synechocystis strain producing sfGFP-loaded EVs To construct the Synechocystis strain expressing the heterologous protein sfGFP and further targeting it to EVs, the synthetic promoter P trc. x.lacO (Ferreira et al., 2018), the signal peptide sequence of the Synechocystis native protein FutA2 (Polyviou et al., 2018), which determines its translocation to the periplasm, and the gene encoding the reporter protein super-folder Green Fluorescent Protein (sfGFP) were amplified by PCR using specific oligonucleotides (Table S1). Assembly of the different DNA fragments was performed by sequential overlap-extension PCR. The 953 bp fragment was digested with PstI and XbaI, and ligated to plasmid pSN15KPO previously digested with the same restriction enzymes, rendering plasmid pEV-trc-GFP. Identity of the fragment was determined by Sanger sequencing. Plasmid pSN15KPO is a variant of plasmid pSN15K (Pinto et al., 2015), in which the kanamycin resistance cassette was replaced by that of plasmid pUC4K (GE S. Lima et al. Journal of Hazardous Materials 431 (2022) 128594 4 Healthcare). Synechocystis wild-type cells were naturally transformed with pEV-trc-GFP, and fully segregated mutants were obtained, rendering strain EV-trc-GFP. 2.7. Total protein extraction, concentration of cell-free extracellular medium and protein analyses Cells from Synechocystis wild-type and mutant strains were grown as described above, and harvested by centrifugation at 4400 g for 10 min at room temperature. Total protein extracts were obtained as described (Lopes Pinto et al., 2011). In brief, cell pellets were suspended in protein extraction buffer [10 mM HEPES, 0.5% (v/v) Triton X-100, 10 mM EDTA, 2 mM DTT, 10% (v/v) glycerol, pH 8.0, supplemented with protease inhibitor cocktail (Roche)] and disrupted by sonication (Branson Sonifier 250). Subsequently, samples were centrifuged at 10, 000 g for 10 min at 4 ◦C, and the resulting supernatant transferred into a new 1.5 mL microcentrifuge tube and stored at −20 ºC until further analysis. Protein concentration was determined using the BCA method (Thermo Fisher Scientific), with bovine serum albumin as standard. For cell-free, extracellular medium concentrated samples, the method described by Oliveira et al. (2015) was followed. Briefly, cells were cultivated as described up to an OD 730 of approximately 1.0–1.5, and later separated from the extracellular medium by centrifugation (4400 g, 10 min at room temperature). Spent medium was filtered through 0.2 µm pore size filters for cell removal, and further concentrated approximately 1000-fold by ultrafiltration, using Amicon® Ultra-15 Centrifugal Filter Units (Merck) with molecular weight cut-off of 3 kDa. Concentrated samples were either used immediately or stored at −20 ºC. The amount of concentrated spent medium sample analyzed was normalized by cell culture density (OD 730 ), volume of cell-free culture medium concentrated, and concentration factor. Total protein and cell-free, concentrated extracellular medium samples were electrophoresed on homogeneous 12 or 16% (w/v) SDS polyacrylamide gels, stained with colloidal Coomassie Brilliant Blue (Sigma), and gels were scanned with a GS-800 Calibrated Densitometer (BioRad). For Western blotting, after SDS-polyacrylamide gel electrophoresis, protein samples were transferred onto a nitrocellulose membrane (GE Healthcare). Membranes were incubated in Blocking buffer [5% (w/v) fat-free milk powder, and 0.05% (v/v) Tween-20 in TBS (20 mM Trisbase, 0.8% (w/v) NaCl, pH 7.6)] for 2 h before being incubated overnight at 4 ºC with rabbit anti-CopM antibody (Giner-Lamia et al., 2015) (1:3000) in fresh blocking buffer. After incubation with the primary antibody, membranes were washed twice in TTBS [0.05% (v/v) Tween-20 in TBS], incubated for 1 h with an anti-rabbit secondary antibody conjugated with horseradish peroxidase (Sigma) (1:10000), washed again twice with TTBS and twice with TBS. For sfGFP detection, the commercial mouse anti-GFP monoclonal antibody (Roche) (1:5000), and the goat-anti-mouse secondary antibody (Invitrogen) (1:5000) were used. Detection was carried out on a ChemiDoc MP System (BioRad), using the Clarity™ Western ECL Substrate (BioRad) reagent according to the manufacturer’s instructions. 2.8. Extracellular vesicles isolation and analysis Synechocystis cultures were grown as presented above (“Strains and growth conditions”) and extracellular vesicles (EVs) were isolated as follows: cyanobacterial cells were separated from the extracellular medium by centrifugation at 4400 g for 10 min at room temperature, and further removed from the extracellular medium by filtration (0.2 µm filter pore size). Cell-free extracellular medium was then concentrated by ultrafiltration (4400 g, 16 ◦C) using Amicon® Ultra-15 Centrifugal Filter Units (Merck) with a molecular weight cut-off of 100 kDa. Later, concentrated samples were diluted with 15 mL of fresh BG11 medium and ultracentrifuged (70 Ti fixed angle rotor, Beckman Coulter Optima L80 XP Ultracentrifuge) using polycarbonate bottles with cap assembly (#355618) for 3 h at 100,000 g at 4 ◦C. Finally, the resulting pellet was suspended in sterile BG11 growth medium, and stored at −80 ºC until further analysis. E. coli cultures were grown as described, and EVs isolated similarly as performed for Synechocystis EVs. Briefly, cells were collected by centrifugation at 4400 g for 15 min at room temperature, and further removed from the extracellular medium by filtration (0.45 µm filter pore size). Then, cell-free extracellular medium was concentrated by ultrafiltration (4400 g, 4 ◦C) with centrifugal filters with a molecular weight cut-off of 100 kDa (Pall). Prior to ultracentrifugation, samples were mixed with 15 mL of PBS (Phosphate Buffered Saline, 1X, pH 7.5) and filtered once again (0.45 µm) to ensure complete removal of any biological contaminant. Ultracentrifugation was performed for 2 h at 174,900 g (70 Ti fixed angle rotor, Beckman Coulter Optima L80 XP Ultracentrifuge) at 4 ◦C. The resulting pellet was suspended in sterile PBS, and stored at −80 ºC until further analysis. A similar procedure was performed to isolate EVs from P. aeruginosa. Cells were collected by centrifugation at 10,000 g for 30 min at room temperature, and further removed from the extracellular medium by filtration (0.45 µm filter pore size). Then, cell-free extracellular medium was also concentrated by ultrafiltration (4400 g, 4 ◦C). Prior to ultracentrifugation, samples were treated as mentioned: washed with 15 mL of PBS and filtered once again (0.45 µm). Ultracentrifugation was performed for 3 h at 150,000 g at 4 ◦C. The resulting pellet was suspended in sterile PBS, and stored at −80 ºC until further analysis. Various methodologies were performed to evaluate EVs physical features and biochemical composition. Briefly, EVs number and size were analyzed by nanoparticle tracking analysis and their morphological aspect by electron microscopy. Regarding EVs composition, proteins and lipopolysaccharides (LPS) were routinely analyzed by SDS- polyacrylamide gel electrophoresis. Detection of CopM in isolated Synechocystis EVs was carried out by Western blot, and vesicular copper levels were quantified by atomic absorption spectrometry (AAS) (see further details about these techniques in Supporting information Text S1). 3. Results 3.1. Inactivation of the TolC-like protein in Synechocystis results in several adaptations at the protein level The TolC-like protein (Slr1270) has been shown in Synechocystis to play key roles in protein export and multidrug efflux (Oliveira et al., 2016) as part of the type I secretion system (T1SS) and of the resistance-nodulation-division (RND) efflux pumps (Gonçalves et al., 2018). In addition, a Synechocystis tolC-mutant strain was shown to release more EVs than the respective wild-type (Oliveira et al., 2016; Gonçalves et al., 2018). In order to get closer insights into the adaptations occurring in Synechocystis cells lacking a functional TolC-like protein, iTRAQ-based quantitative shotgun proteomics was carried out. In this study, a total of 1135 proteins were identified with quantification information [with at least two peptides, and at a false discovery rate (FDR) of less than 1%], corresponding to approximately 32.4% of the total predicted Synechocystis proteome. 212 proteins were found to be differentially expressed between the strains analyzed (Table S2). Among the statistically significant differentially expressed proteins, 46 show at least 1.35-fold up- or down-regulation in respect to wild-type protein levels (Table 1), with more than 35% of the hits falling in the “hypothetical” or “functionally uncharacterized” protein categories. Some of the proteins differentially expressed have been previously reported to accumulate differently in the tolC-mutant strain as compared to the wild-type, namely proteins Sll1951 (−3.25-fold), Slr1841 (1.38-fold), and ApcE (1.41-fold) (Oliveira et al., 2016). In addition, proteins related to metal uptake and homeostasis, particularly CopM-Slr6039/Sll0788 (1.75-fold), FutA1 (1.43-fold), FrpC (−2.06-fold), or whose function is dependent on metal cofactors, namely IsiB (3.18-fold) and cytochrome c 6 (−2.55-fold), were found amongst S. Lima et al. Journal of Hazardous Materials 431 (2022) 128594 5 the most differentially expressed proteins. 3.2. TolC is essential for metal homeostasis in Synechocystis Based on these proteomic results, we wondered whether the tolC- mutant could be under some sort of metal imbalance. To address this question, the growth of the cyanobacterial strains was tested in the presence of selected metals, namely boron, cadmium, cobalt, copper, iron, manganese, molybdenum, nickel, and zinc (Fig. 1A). No differences in growth could be observed between Synechocystis wild-type and tolC-mutant in medium supplemented with boron, cadmium, manganese, molybdenum, or nickel. In contrast, significant growth impairment was detected in the mutant strain cultivated in medium supplemented with copper and zinc, and, to a lesser extent, cobalt and iron, when compared to the wild-type. Therefore, we proceeded to determine the intracellular levels of cobalt, copper, iron and zinc by atomic absorption spectrometry in both strains cultivated under standard growth conditions (Figs. 1B and S1). While similar intracellular levels for cobalt, iron and zinc could be detected in both strains (Fig. S1), tolC-mutant cells were found to accumulate approximately 3-fold more copper than the wild-type (Fig. 1B). At this point, we hypothesized that the role of TolC in copper homeostasis could be related to: (i) the export of an unknown protein with copper-binding ability, via the T1SS; (ii) the efflux of copper through an RND pump; or (iii) the efflux of copper through a heavy metal efflux (HME) pump. To address these possibilities, we cultivated Synechocystis wild-type, and mutant strains ΔtolC, ΔhlyB (lacking the inner membrane component of the T1SS (Gonçalves et al., 2018)), ΔacrA (lacking the periplasmic adaptor protein of the main RND pump in Synechocystis (Gonçalves et al., 2018)), and ΔcopB (mutant lacking the periplasmic adaptor protein of the copper-specific HME-pump in Synechocystis (Giner-Lamia et al., 2012)) in medium supplemented with copper, and the growth of the various strains was analyzed (Fig. 2). All strains grew similarly under standard growth conditions (0.5 µM Cu 2+ ), and cells of the acrA-mutant strain presented the same growth capacity as the wild-type strain when cultivated in medium supplemented with copper (4.0 µM Cu 2+ ). However, while cells of the hlyB-mutant could grow slightly better than those of the wild-type strain under copper-challenging conditions, both the copB- and tolC-mutant strains exhibited growth impairment when compared to the wild-type. Remarkably, the tolC-mutant showed a more severe growth impairment as compared to the copB-mutant (Fig. 2). Altogether, these observations suggest that TolC plays a key role in metal homeostasis in Synechocystis, particularly in keeping an intracellular balance of copper levels, likely working as the outer membrane component of the HME-pump dedicated to copper export. 3.3. Absence of TolC in Synechocystis triggers upregulation of copper detoxification mechanisms Two main mechanisms have been described in Synechocystis to warrant copper detoxification, both being involved in metal export when cells are challenged with stressfully high copper concentrations: the HME-pump (dependent on the CopBA system), and the periplasmic metallochaperone system (dependent on CopM). In order to evaluate how these systems responded in the Synechocystis tolC-mutant strain, Western blotting experiments and/or transcriptional analyses were carried out. On one hand, RT-qPCR experiments revealed that the tolC- mutant has significantly higher copB transcript levels than the wild-type (Fig. 3A). On the other hand, transcript analysis by Northern blotting showed that the tolC-mutant accumulates higher levels of copM transcripts than the wild-type (data not shown). In agreement, iTRAQ results (Table 1) also indicated that CopM (Slr6039/Sll0788) is significantly overexpressed in tolC-mutant cells. Using a CopM-specific antibody (Giner-Lamia et al., 2015), Western blotting experiments determined that the intracellular CopM levels in the tolC-mutant strain are Table 1 List of proteins identified in the iTRAQ (isobaric tags for relative and absolute quantification) proteomics analysis with differential accumulation (>1.35-fold; <−1.35-fold) in the Synechocystis sp. PCC 6803 tolC-mutant strain in respect to the wild-type strain cultivated under standard growth conditions. Uniprot # peptides Protein names Gene names Fold Changes P27319 22 Flavodoxin isiB sll0248 3.18 P73742 100 Sll0858 protein sll0858 2.95 P73238 31 Slr2018 protein slr2018 2.18 P73354 28 Serine protease HtrA htrA slr1204 2.08 Q6YRW5 2 Slr6039 protein slr6039 1.75 P72656 2 Ribonuclease E/G-like protein rne slr1129 1.75 Q55604 3 Slr0769 protein slr0769 1.69 P73066 29 Ycf23 protein ycf23 slr2032 1.46 Q55665 58 Glutamate-1-semialdehyde 2,1-aminomutase (GSA) hemL gsa sll0017 1.44 P72827 21 Iron uptake protein A1 futA1 slr1295 1.43 P73315 20 50 S ribosomal protein L22 rplV rpl22 sll1803 1.43 P74791 6 PirA (PII interacting protein A) ssr0692 1.42 P37101 163 Phosphoribulokinase (PRK) prk ptk sll1525 1.41 Q55544 60 Phycobiliprotein ApcE apcE slr0335 1.41 P74786 6 UPF0426 protein ssl0294 ssl0294 1.39 P72851 9 50 S ribosomal protein L28 rpmB rpl28 ssr1604 1.38 Q55797 14 Phosphoheptose isomerase (EC 5.3.1.28) (Sedoheptulose 7- phosphate isomerase) gmhA sll0083 1.38 P73409 9 Slr1841 protein slr1841 1.38 P74478 16 Slr1926 protein slr1926 1.38 P73288 40 Aklaviketone reductase sll1825 1.35 P73644 6 S-adenosyl-methionine: precorrin-2 methyltransferase cbiL -1.35 Q55770 40 Sll0185 protein sll0185 -1.36 P52981 22 1,4-alpha-glucan branching enzyme GlgB glgB sll0158 -1.36 P74722 3 Slr0586 protein slr0586 -1.36 P37277 6 Photosystem I reaction center subunit XI (PSI subunit V) (PSI- L) psaL slr1655 -1.36 Q55146 9 Sll0064 protein sll0064 -1.37 Q55162 10 Sll0051 protein sll0051 -1.38 Q55961 8 Slr0699 protein slr0699 -1.39 P73111 19 Sll1835 protein sll1835 -1.39 P73477 7 Slr1232 protein slr1232 -1.42 P74748 10 Slr0602 protein slr0602 -1.42 P73595 59 Uncharacterized WD repeatcontaining protein slr1410 slr1410 -1.42 Q55735 3 Sll0394 protein sll0394 -1.44 P73559 7 Slr0878 protein slr0878 -1.45 P73594 62 Uncharacterized WD repeatcontaining protein slr1409 slr1409 -1.46 Q55561 10 Sll0167 protein sll0167 -1.46 P73731 5 Photosystem II reaction center Psb28 protein slr1739 -1.46 P74485 15 Sll1863 protein sll1863 -1.52 P74486 2 Sll1862 protein sll1862 -1.54 Q55549 2 Slr0168 protein slr0168 -1.54 P74162 24 Sll1380 protein sll1380 -1.70 Q55776 9 Sll0180 protein sll0180 -1.79 P73019 5 Iron-regulated protein frpC sll1009 -2.06 P46445 14 Cytochrome c 6 petJ sll1796 -2.55 Q55747 42 Slr0420 protein slr0420 -2.64 P73817 13 Hemolysin sll1951 -3.25 S. Lima et al. Journal of Hazardous Materials 431 (2022) 128594 6 Fig. 1. The Synechocystis sp. PCC 6803 tolC-mutant shows a growth-sensitive phenotype when cultivated in the presence of elevated concentrations of specific metals. A. Cells of Synechocystis wild-type (WT) and tolC-mutant (Δ) strains were cultivated in standard liquid conditions up to an OD 730 of approximately 0.8–1.0, concentrated to an OD 730 of 10, and later diluted (OD 730 of 5, 2, 1, 0.5). Cells of each dilution were spotted onto standard BG11 agar plates (BG11), or BG11 agar plates supplemented with boron (B - 463 µM), cadmium (Cd - 5 µM), cobalt (Co - 5 µM), copper (Cu - 4 µM), iron (Fe - 229 µM), manganese (Mn – 91.5 µM), molybdenum (Mo - 16.1 µM), nickel (Ni - 7 µM), or zinc (Zn - 14 µM), and cultivated for 7 days. Presented photographs are representative of the results obtained with 3 independent biological replicates. Bold and underlined metal symbols indicate conditions in which the Synechocystis tolC-mutant showed growth-sensitive phenotype when compared to the wild-type strain, and further selected for intracellular quantification of the respective metal’s concentration. B. Intracellular copper concentration determined for Synechocystis wild-type (WT) and tolC-mutant (ΔtolC) strains cultivated under standard liquid BG11 conditions. Copper levels were determined by atomic absorption spectrometry, and are shown in µg Cu 2+ .L culture −1 .OD 730−1 . Error bars represent the standard deviations of at least three independent biological replicates. ****, P≤0.0001. Fig. 2. The Synechocystis sp. PCC 6803 tolC-mutant coppersensitive phenotype is not associated to its T1SS- or to its RND-dependent function. Cells of Synechocystis wild-type (WT), and mutant strains ΔtolC, ΔhlyB (T1SS), ΔacrA (RND), ΔcopB (HME-pump) were cultivated in standard liquid conditions up to an OD 730 of approximately 0.8–1.0, concentrated to an OD 730 of 5, and later diluted (OD 730 of 1 and 0.5). Cells of each dilution were spotted onto standard BG11 agar plates (0.5 µM Cu 2+ ), or BG11 agar plates supplemented with copper (4 µM Cu 2+ ), and cultivated for 7 d. Presented photographs are representative of the results obtained with 3 independent biological replicates. The panel to the right shows cell density quantification by densitometry analysis (Image J) determined for each cyanobacterial strain, spotted at initial OD 730 of 0.5, and cultivated in BG11 agar plates supplemented up to a final concentration of 4 µM copper sulfate. Error bars represent the standard deviations of at least three independent biological replicates. *, P≤0.05; ***, P≤0.001; ****, P≤0.0001, each determined against the WT. Fig. 3. Synechocystis sp. PCC 6803 tolC-mutant shows high upregulation of the copper detoxification mechanisms under standard growth conditions. A. RT-qPCR analysis of the copB transcript levels in Synechocystis wild-type (WT) and tolC-mutant (ΔtolC) strains cultivated under standard growth conditions. B. Western blot analysis of the CopM-protein levels in: 10 µg of total protein extracts (Intracellular - IC); cell-free, concentrated extracellular medium (Extracellular medium - EM); and isolated extracellular vesicles (EVs), of Synechocystis wild-type and tolC-mutant strains cultivated under standard growth conditions. Concentrated extracellular medium and EVs samples were normalized before loading for initial cell-free culture volume concentrated, final volume of concentrated sample, and culture’s OD 730 . upCopM: unprocessed CopM-protein; ppCopM: processed CopM-protein. C. Quantification of CopM protein levels detected by Western blotting in total protein extracts (Intracellular - IC), and in the cell-free extracellular growth medium (Extracellular medium - EM), as determined by densitometry analysis (ImageJ). Contribution of EVs for CopM levels detected in the complete cell-free extracellular medium is highlighted in the graph. Error bars represent the standard deviations of at least three independent biological replicates. *, P≤0.05; ****, P≤0.0001. S. Lima et al. Journal of Hazardous Materials 431 (2022) 128594 7 approximately 6-fold higher than those detected in the wild-type (Fig. 3B and C). Moreover, the two previously described CopM forms could be detected (Fig. 3B): the unprocessed, full-length protein, and the processed CopM, resulting from the signal peptide removal from its N-terminus upon translocation from the cytoplasm to the periplasm (Giner-Lamia et al., 2015). CopM has been described as a periplasmic copper-binding chaperone, but it has also been found to accumulate in the extracellular medium (Giner-Lamia et al., 2015). Accordingly, CopM could also be detected here in cell-free, concentrated extracellular medium samples obtained from both strains, even though in significantly higher amounts in the mutant (Fig. 3B and C). These results suggest that both copper-detoxification mechanisms are activated, and are highly expressed in a Synechocystis strain lacking a functional TolC-protein. 3.4. The copper-binding chaperone CopM can be found in EVs isolated from Synechocystis After finding high levels of CopM in extracellular medium samples, it became important to determine whether CopM could be detected in Synechocystis EVs. To that end, extracellular medium of both Synechocystis wild-type and tolC-mutant strain were concentrated by ultrafiltration, and their respective EVs isolated by ultracentrifugation. Western blotting confirmed the presence of CopM protein in EVs, particularly in those isolated from the tolC-mutant (Fig. 3B). When normalizing the amount of CopM detected in EVs isolated from wild-type and mutant strain, by the total amount of vesicles recovered in each respective condition as determined by the amount of LPS (molecular marker of EVs released by Gram-negative bacteria), we could estimate that the CopM protein levels are approximately 4-fold higher in tolC-mutant EVs as compared to wild-type. Yet, the amount of CopM identified therein corresponded just to a fraction of the total extracellular CopM pool (Fig. 3C). To investigate if proteins reaching the periplasm of Cyanobacteria could be detected freely available in the growth medium as well as in EVs, we generated a Synechocystis mutant expressing the heterologous reporter protein sfGFP (Fig. 4). The gene encoding sfGFP was fused with the Tat-dependent, signal peptide described in the Synechocystis periplasmic protein FutA2 (Polyviou et al., 2018), and its expression was controlled by the P trc.x.lacO promoter (Ferreira et al., 2018). Confocal microscopy analysis of transformed cells clearly indicated accumulation of sfGFP-protein in the outermost region of the cell, indicating successful targeting of the reporter to the periplasm (Fig. 4A). Moreover, confocal microscopy also showed that sfGFP is packaged in isolated EVs (Fig. 4B). In agreement, Western blotting analyses of isolated EVs confirmed the presence of sfGFP therein, as well as freely available in EVs-free, concentrated extracellular medium (data not shown). The amount of sfGFP freely available in the medium was close to 8-fold higher than that found in EVs (Fig. 4C). This experiment has now elucidated that soluble proteins reaching the Synechocystis periplasm may leave the cell by translocation across the outer membrane or by packaging into EVs, supporting the possibility of CopM being exported by the Synechocystis cell by both systems. Moreover, it is also possible that EV-protein cargo is released into the extracellular space by vesicle leakage or breakdown. 3.5. Synechocystis wild-type strain produces more CopM-loaded EVs when cultivated in high copper concentrations In light of these results, the hypervesiculating phenotype observed in the Synechocystis tolC-mutant strain could be partly interpreted as a mechanism responding to an intracellular copper accumulation. To evaluate how the Synechocystis wild-type strain responded in terms of EVs’ production when challenged to grow in the presence of high amounts of copper, the cyanobacterium was cultivated under increasing concentrations of the metal, and the total amount of released EVs was monitored (Figs. 5 and S2). All evidence combined suggests that Synechocystis wild-type cells indeed produce more EVs when exposed to high copper concentrations, a response that seems to follow a copper concentration-dependent manner. Interestingly, while regular increments in Cu 2+ concentration up to 3 µM resulted in modest, but steady increase of EVs’ production [as determined by the amount of LPS (Fig. 5A and B) and nanoparticles (Fig. S2)], presence of 4 µM of Cu 2+ in the medium resulted in a strong boost in vesicle production. In addition, presence of CopM in isolated EVs was also investigated by Western blot: in agreement with the increase in EVs production, the results indicate that the amount of CopM also increases (Fig. 5A). Moreover, detailed analysis of the LPS band profile indicates that EVs released by Synechocystis cells challenged with 4 µM of Cu 2+ present a 5-fold decrease in the amount of high molecular weight, smooth-type LPS relative to the low molecular weight, rough-type LPS (Fig. 5C). This significant difference was also detected in EVs isolated from the Synechocystis tolC- mutant strain, though to a lesser extent (Fig. 5C). 3.6. Other model Gram-negative bacteria also release EVs when cultivated in the presence of high copper concentrations Next, we decided to analyze whether this EVs production phenotype in response to growth-challenging copper concentrations was exclusive of Cyanobacteria or if it could also be observed for other Gram-negative bacteria. Thus, similar experiments as described for Synechocystis wildtype were carried out for the enteric Escherichia coli [strain BL21 (DE3)], and for the opportunistic pathogen Pseudomonas aeruginosa (strain PAO-1). EVs production followed a similar pattern as that observed for Synechocystis wild-type, i.e. as copper-concentration in the medium increased, the amount of produced vesicles increased as well Fig. 4. sfGFP translocated to the Synechocystis sp. PCC 6803 periplasm can also be detected in the extracellular medium, both as free protein and in extracellular vesicles. A. Confocal micrographs of EV-trc-GFP cells (Synechocystis cells expressing sfGFP fused with the signal peptide of the protein FutA2, and under the control of the P trc.x.lacO synthetic promoter) depicting the red autofluorescence signal from the cells’ photosynthetic pigments, and the green signal from the sfGFP reporter. A panel showing the result of merging the two micrographs is also presented. Scale bar: 5 µm. B. Confocal micrograph showing green fluorescent foci obtained from EVs isolated from EV-trc-GFP cells. Scale bar: 3 µm. C. Western blot analysis of sfGFP protein in 10 µg of total protein (IC - intracellular), in cell-free concentrated extracellular media (EM) and isolated extracellular vesicles (EVs) samples of EV-trc-GFP cells. EM and EVs samples were normalized for initial cell-free culture volume concentrated, final volume of concentrated sample, and culture’s OD 730 . D. Quantification of sfGFP protein levels detected by Western blotting in the complete cell-free extracellular growth medium (EM), and in EVs only, as determined by densitometry analysis (ImageJ). Error bar represents the standard deviation of at least three independent biological replicates. S. Lima et al. Journal of Hazardous Materials 431 (2022) 128594 8 (Figs. 6A and B, S3). Furthermore, a significant decrease in the relative amount of smooth- to rough-type LPS could be observed in EVs isolated from cells of both strains exposed to high copper concentrations as compared to EVs isolated from cells cultivated under standard growth conditions (Fig. 6C). This result is similar to the response detected in Synechocystis wild-type (Fig. 5C). Altogether, this suggests that the selected Gram-negative bacteria, which include members that occupy different ecological niches respond in a similar manner when exposed to high copper concentrations. 3.7. Copper can be detected in isolated cyanobacterial EVs Finally, to address the possibility of bacterial EVs to work as copperreleasing capsules, the copper amount in vesicles isolated from Synechocystis wild-type and tolC-mutant strains was analyzed. The choice of strains lies on the fact that the mutant accumulates copper intracellularly already in standard growth conditions (Fig. 1). This means that there is no need to supplement the medium with extra amounts of copper, which could represent a technical challenge when quantifying copper in EVs. Thus, Synechocystis wild-type and tolC-mutant strains were cultivated in standard growth medium (0.5 µM Cu 2+ ), their EVs isolated, thoroughly washed to remove any possible metal binding to the vesicles’ surfaces, and later analyzed by atomic absorption spectrometry. Copper could be detected and quantified in EVs samples obtained from both strains. However, the amount of copper quantified in tolC- mutant-derived EVs was significantly higher than that of wild-type vesicle fractions, in approximately 3-fold (Fig. 7). 4. Discussion 4.1. TolC is the outer membrane component of the Synechocystis copperefflux pump A Synechocystis strain lacking a functional TolC protein has been shown to be severely impaired in secreting several types of molecules, including proteins and antibiotics (Oliveira et al., 2016; Agarwal et al., 2018), drugs (Gonçalves et al., 2018), and even fatty acids (Bellefleur et al., 2019). This phenotype is related to the fact that TolC works as the outer membrane component in multiple tripartite systems (Gonçalves et al., 2019), in which substrate selectivity is defined by the inner membrane component and the periplasmic adaptor protein (Gonçalves et al., 2021; Russo and Zedler, 2021). Thus, TolC can promiscuously (caption on next column) Fig. 5. Synechocystis sp. PCC 6803 wild-type produces extracellular vesicles in response to copper supplementation in the medium in a concentration dependent manner. Synechocystis wild-type (WT) cells were grown under standard growth conditions (liquid BG11 medium with 0.5 µM Cu 2+ ), and later distributed into BG11 standard medium (0.5 µM Cu 2+ ) or BG11 medium supplemented with copper to a final concentration of 1, 2, 3 and 4 µM Cu 2+ . In addition, Synechocystis tolC-mutant cells (ΔtolC) were cultivated in standard conditions, and then further grown in fresh BG11 medium. For all conditions, cyanobacterial cells were cultivated for approximately 7 d, and their respective extracellular vesicles (EVs) isolated. A. Isolated EVs were analyzed by Western blotting for the detection of CopM-protein levels (upper panel). M, NZYColour Protein Marker II ladder (NZYTech). Molecular weight of selected bands is shown to the left in kDa. As lipopolysaccharides (LPS) are regarded as a marker for the presence and abundance of Gram-negative EVs, EVs samples were separated by electrophoresis on 16% (w/v) SDS-polyacrylamide gels and LPS were stained using the Pro-Q® Emerald 300 Lipopolysaccharide Gel Stain Kit (Life Technologies) (bottom panel). LPS with (Smooth) or without (Rough) O- antigen moieties are indicated. B. Total LPS content of Synechocystis WT EVs was quantified by densitometry analysis (ImageJ). C. Quantification of smooth- to rough-LPS ratio for EVs isolated from Synechocystis WT, Synechocystis WT cultivated with 4 µM Cu 2+ , and Synechocystis tolC-mutant cells. Error bars represent the standard deviations of at least three independent biological replicates. **, P≤0.01; ***, P≤0.001, each determined against the WT cultivated under standard growth conditions (0.5 µM Cu 2+ ). S. Lima et al. Journal of Hazardous Materials 431 (2022) 128594 9 assemble with either T1SS-components or RND-efflux pumps to promote the export of molecules directly from the cytoplasm to the extracellular medium. The observation of an intracellular copper accumulation in Synechocystis tolC-mutant, accompanied by adaptations related to such copper-imbalance, including upregulation of copB transcripts (Fig. 3), high overexpression of CopM (Table 1 and Fig. 3), moderate overexpression of plastocyanin (Table S2), strong and moderate downregulation of cytochrome c 6 (PetJ) and of Slr0602, respectively (Table 1), with the latter three proteins part of the PetRP regulon, recently described to be copper-responsive (García-Ca˜ nas et al., 2021), highlight the need for the mutant cells to secrete copper. Taken together, our results support the possibility of TolC working as the outer membrane component of the Synechocystis copper-efflux pump. One of the best described copper efflux systems is the E. coli CusCFBA transport system (Franke et al., 2003). In this microorganism, CusA is the inner membrane component, CusB the periplasmic adaptor protein, CusC the outer membrane component, and CusF is a soluble periplasmic protein that transports copper in the periplasm to the pump (Nies et al., 2003). Remarkably, despite low sequence similarity, CusC is structurally very similar to TolC (Kulathila et al., 2011), but TolC is not able to physiologically replace CusC in E. coli, in respect to conferring resistance to copper ions in vivo (Franke et al., 2003). The most likely reason for this observation would be an inability of TolC to form a functional pump with CusAB (Kulathila et al., 2011). It has been suggested that the linkage among components maintaining copper homeostasis seems evolutionarily driven by protein-protein interactions Fig. 6. Model bacterial strains produce extracellular vesicles in response to copper supplementation in the medium in a concentration dependent manner. Bacterial strains Escherichia coli [strain BL21 (DE3) left-hand panels] and Pseudomonas aeruginosa [strain PAO-1 righthand panels] were cultivated either in LB liquid medium (LB) or in LB liquid medium supplemented with 0.5, 1 and 2 mM CuSO 4 for E. coli, or 1 and 3 mM CuSO 4 for P. aeruginosa. After overnight cultivation, EVs were isolated. A. EVs samples were separated by electrophoresis on 16% (w/v) SDS-polyacrylamide gels and lipopolysaccharides (LPS) were stained using the Pro-Q® Emerald 300 Lipopolysaccharide Gel Stain Kit (Life Technologies). LPS with (Smooth) or without (Rough) O-antigen moieties are indicated. B. Total LPS content of E. coli and P. aeruginosa EVs were quantified by densitometry analysis (ImageJ). C. Quantification of smooth- to rough-LPS ratio for EVs isolated from E. coli cultivated in LB medium or medium supplemented with 2 mM Cu 2+ , and P. aeruginosa cultivated in LB medium or medium supplemented with 3 mM Cu 2+ . Error bars represent the standard deviations of at least three independent biological replicates. *, P≤0.05; **, P≤0.01, each determined against the respective strain cultivated in standard LB medium (LB). S. Lima et al.