Genomic Survey of Salt Acclimation-Related Genes in the Halophilic Cyanobacterium Euhalothece sp. Z-M001

Like other halophilic cyanobacterial genomes, the de novo-assembled genome of Euhalothece sp. Z-M001 lacks genes encoding keto-carotenoid biosynthesis enzymes, despite the presence of genes encoding carotenoid-binding proteins (CBPs). Consistent with this, HPLC analysis of carotenoids identified β-carotene and zeaxanthin as the dominant carotenoids. CBPs coexpressed with the zeaxanthin biosynthesis gene increased the survival rates of Escherichia coli strains by preventing antibiotic-induced accumulation of reactive oxygen species (ROS). RNA-seq analysis of Euhalothece revealed that among various salt resistance-related genes, those encoding the Na+ transporting multiple resistance and pH adaptation (Mrp) systems, glycine betaine biosynthesis enzymes, exopolysaccharide metabolic enzymes, and CBPs were highly upregulated, suggesting their importance in hypersaline habitats. During the early phase of salt deprivation, the amounts of β-carotene and zeaxanthin showed a negative correlation with ROS content. Overall, we propose that in some halophilic cyanobacteria, β-carotene and zeaxanthin, rather than keto-carotenoids, serve as the major chromophores for CBPs, which in turn act as effective antioxidants.


Results and Discussion
phylogenetic analysis and genome of Euhalothece. Phylogenetic analysis of 16S rRNA sequences derived from 59 cyanobacterial species, three eubacteria (used as outgroups) 21,22 , and three additional species (Euhalothece Z-M001, Euhalothece Z9404, and Nostoc punctiforme ATCC29133) revealed that Euhalothece is closely related to the stenohaline cyanobacterium Dactylococcopsis salina PCC8305 (supported by a perfect bootstrap value) (Supplementary Fig. S1; red color). This clade does not contain the Microcoleus and Geitlerinema taxa that form the basis of a monophyletic group comprising freshwater cyanobacteria subsections I and II. Current phylogeny reconstruction indicates that the halophilic cyanobacteria group is associated with diversification of freshwater cyanobacteria.
The complete circular genome of Euhalothece (GenBank ID: CP042326) is 3,317,848 bp in length (average GC content = 41.03%) and contains 3,354 coding sequences (CDSs), six rRNAs, 43 tRNAs, and three clustered regularly interspaced short palindromic repeat arrays ( Supplementary Fig. S2). Genes responsible for carotenoid biosynthesis, including crtR (ezr1976) and cruA (ezl1782), and CBPs such as ocp (ezr1969), hcp (ezr2074), and ctdh (ezr2075), were distributed on both strands, and no strong GC skew or single DNA replication origin inflection point was detected. In addition to the main chromosome, four scaffolds (GenBank IDs: CP042327-CP042330) were assembled into plasmids. All plasmids were circular, varying in size (14,665-67,514 bp), G + C content (40.04-40.85%), and gene number . In plasmids, genes encoding mobile element proteins, ParAB, primase, DNA invertase, DNA integrase, transposase, and phage tail protein, were identified, but those encoding CBPs and proteins involved in carotenoid biosynthesis and EPS metabolism were absent. Overall, the four plasmids were different but contained a 1.6 kb long block of similar sequence; the remaining plasmid sequence shared no similarity.
Growth, photosynthetic pigments, and epS content of Euhalothece acclimated to salt-depleted and -repleted conditions. In this study, salt deprivation retarded the growth of Euhalothece cells. Cells grown for 6 days on medium containing 3% NaCl grew faster (Fig. 1A), and contained more Chl a and phycocyanin (PC) than those grown under the salt-deprived conditions (Fig. 1B). By contrast, the total carotenoid and EPSs contents of cells under salt-depleted conditions increased significantly by several-fold (Fig. 1C) and 2.1-fold (Fig. 1D), respectively. Thus, Euhalothece seems to adopt primarily EPS-based salt-out and carotenoid-based ROS scavenging strategies under varying external salt concentrations.
By contrast, the KtrAB(E) system, which is present in all known cyanobacterial genomes and presumably plays a key role in K + import 19,23 , was identified in the genome of Euhalothece (Supplementary Table S2). Among various transporter-encoding genes, the mrp gene clusters seem to play a major role in Euhalothece, since genes encoding these multisubunit Na + /H + antiporters were constitutively expressed under salt-deficient (0% NaCl) and -sufficient (3% NaCl) conditions, as revealed by RNA-seq profiling (Supplementary Table S2) and qRT-PCR analysis (Fig. 2).
Among compatible solutes such as sucrose, trehalose, glucosylglycerol, and glycine betaine, glucosylglycerate, and proline, genes encoding proteins involved in biosynthesis of sucrose (spsA and spp), trehalose (treS and treY), glucosylglycerol (ggpP and ggpS), glycine betaine (gsmt and sdmt), and proline (proA/B/C) were annotated; however, genes encoding glucosylglycerate biosynthesis proteins were absent from the Euhalothece genome. Transcripts of genes encoding GSMT and SDMT were highly abundant compared with those of genes involved in biosynthesis of other solutes (Supplementary Table S2, Fig. 2). This suggests that glycine betaine acts a major compatible solute that protects against salt stress in Euhalothece, as in other hypersaline cyanobacteria 24 .
EPSs are anionic due to the presence of uronate and sulphate groups, which chelate cations such as Na + and heavy metal ions in various photosynthetic and non-photosynthetic bacteria. Over 500 putative genes associated with the cell surface structure have been identified in Synechocystis based on their homology to genes in E. coli Figure 1. Growth, photosynthesis pigments, and exopolysaccharide (EPS, D) content in Euhalothece sp. Z-M001 maintained for 6 days on S growth medium containing 0% or 3% NaCl. (A-D) Growth (A), chlorophyll a (Chl) and phycocyanin (PC) content (B), carotenoid (Car) content (C), and exopolysaccharide (EPS) content (D) of Euhalothece cells. Cells grown in 3% NaCl-containing liquid S-medium (+Na) were transferred to 0% NaCl (−Na) medium under growth light conditions for 6 days. Data represent mean ± S.E. (n = 3-5). Significant differences are indicated with an asterisk (*p < 0.05; Student's t-test).
and Salmonella 25 . To date, a few dozen putative genes have been tested experimentally. The gene cluster sll1722-sll1726, presumed to encode glycosyltransferases 26 , and the exoD (Slr1875), gumB (Sll1581), and gumC (Sll0923) genes 27 are involved in production of EPSs and in protection against various stresses caused by light, H 2 O 2 , NaCl, CdSO 4 , CoCl 2 , and Fe starvation. Furthermore, Synechocystis genes encoding permeases (slr0977 and sll0574) and ABC-transporters (slr0982 and sll0575) are involved in production of EPSs, which in turn play a crucial role in cell adherence 25 . Among the 151 Synechocystis gene homologs 25 , 118 genes potentially involved in polysaccharide biosynthesis were identified in the genome of Euhalothece (Supplementary Table S2). These 118 genes were expressed at different levels in the presence of 3% NaCl, but only six were responsive to salt deprivation. Expression of five genes, including spsA, epsB, rfbW, hlyB, and rfbP, increased by more than 2-fold, while that of only one gene encoding glucosyl transferase fell (Supplementary Table S2 and Fig. 2). Thus, Euhalothece utilizes EPS production as one salt acclimation strategy for survival in hypersaline natural habitats.
The K + transporter system Ktr 23 and the glucosylglycerol biosynthesis protein GgpS 28 in Synechocystis are regulated directly by the intracellular salt concentration. In addition to the post-translational regulatory mechanism, ion content directly regulates expression of salt-responsive genes at the transcriptional level via alternative sigma factor-dependent changes in the promoter selectivity of RNA polymerase 29,30 . In addition to sigma factor cascades, two component systems comprising a sensory histidine kinase (Hik) and the corresponding response regulator (Rre) mediate salt responses. For instance, Hik proteins such as Hik16, Hik33, Hik34, and Hik41 29 , along with Rre proteins 31,32 , act as salt sensors in Synechocystis. The genome of Euhalothece encodes two component systems comprising 41 Hik and 41 Rre proteins. Among the 41 Hik-encoding genes, expression of only three genes (Ezl3071, Ezr1858, and Ezr2811) was up-or down-regulated in response to a change in salt concentration (Supplementary Table S2). Thus, these Hik proteins could act as sensors of varying salt concentrations, although further investigation is needed. crts and cbps in the Euhalothece genome. In addition to the above-mentioned genes involved in salt acclimation via regulation of intracellular water potential, carotenoids play protective roles against salt stress by promoting protein and membrane stabilization, and by scavenging ROS; this occurs, for example, in the mesophilic cyanobacterium Synechocystis 30 . Thus, the carotenoid composition of Euhalothece was determined along with that of other halophilic cyanobacteria such as Euhalothece Z9404 and Microcoleus B353 (Fig. 3A). These three cyanobacterial species are halophilic and lack keto-carotenoid biosynthesis genes such as crtW and crtO; this is also true for other halophilic cyanobacteria such as Halothece PCC7418 and Dactylococcopsis salina PCC8305 ( Supplementary Fig. S1).
HPLC analysis detected trace amounts of Ech, but not Can, in these three stenohaline cyanobacteria, whereas Ech was identified as a dominant carotenoid in Synechocystis (Fig. 3A). In addition, Zea (5-6% of the total carotenoids) and β-Car (11-13% of the total carotenoids) were detected in Euhalothece and two other stenohaline cyanobacteria. A peak representing unknown compounds was detected in Euhalothece, the retention time of which was less than that of Can and Zea but more than that of Ech. This implies that the compound in the unknown peak is either more hydrophobic than both Can and Zea, but more hydrophilic than Ech, or a structurally different isoform (i.e., the cis-form) of the same carotenoid 14 .
A search for homologs in the complete genome sequence of Euhalothece using the functionally confirmed carotenogenesis genes of both Synechocystis and Nostoc sp. PCC 7120 33,34 as query sequences revealed the presence of crtE, crtB, crtP, crtQ, crtH, cruA, cruP, cruG, and crtR genes in Euhalothece. Among these genes, Ezl1782 and Ezr1976 were designated as genes encoding lycopene cyclase (cruA) and β-Car hydroxylase (crtR), respectively, because E. coli cells carrying these genes produced major peaks corresponding to β-Car and Zea, indicating that Ezl1782 and Ezr1976 gene products are involved in lycopene cyclization and β-Car hydroxylation in Euhalothece (Fig. 3B). Both genes were expressed constitutively in the presence of 3% NaCl and in the absence of NaCl (Supplementary Table S2, Fig. 2). However, as in several halotolerant and halophilic cyanobacterial genomes including Microcoleus sp. B353 22 and Halothece sp. PCC7418 20 , no keto-carotenoid biosynthesis genes such as crtW and crtO were detected. The failure of several Euhalothece crtW/O candidate genes to produce keto-carotenoids in β-Car-accumulating E. coli is consistent with this genomic survey. Based on this information, we propose the carotenoid biosynthesis pathway present in Euhalothece 35,36 (Fig. 3C). However, the absence of crtW and crtO does not necessarily equate to non-production of keto-carotenoids; indeed, it is consistent with previous studies showing that some halotolerant unicellular cyanobacteria, such as Dactylococcopsis salina PCC8305 and Halothece sp. PCC7418, harbor Ech as the predominant carotenoid 14 , despite the absence of crtO and crtW genes. eucBps bind to β-car, Zea, and keto-carotenoids. Genes with a predicted amino acid sequence similar to that of crtW/O were not identified in Euhalothece, but genes encoding OCP and its paralogs HCP and CTDH were present. The Euhcp gene was adjacent to the Euctdh gene. A phylogenetic tree of OCP, HCP, and CTDH constructed using the maximum likelihood method revealed that EuOCP belongs to the OCP1 clade ( Supplementary Fig. S3A), which includes the well-studied OCPs from Synechocystis (SyOCP) and Nostoc sp. PCC7120 (NoOCP) 37 . In general, the ocp1 gene co-exists with the fluorescence recovery protein (frp) gene in the genomes of most cyanobacteria 37 . However, the frp gene was absent from the genome of Euhalothece. The phylogenetic tree for HCP revealed that EuHCP belonged to the HCP5 clade ( Supplementary Fig. S3B). EuCTDH clustered with proteins in one of the major CTDH clades (CTDH1) and showed 91% identity to the amino acid sequences of these proteins ( Supplementary Fig. S3C).
To test whether soluble EuCBPs bind to Zea, the Euocp, Euhcp, and Euctdh genes were expressed in Zea-producing E. coli. Syocp was also expressed in E. coli as a positive control. Recombinant SyOCP proteins bound to Zea but did not exhibit photoconversion activity upon blue light exposure (Fig. 4A)  www.nature.com/scientificreports www.nature.com/scientificreports/ upon blue light illumination (Fig. 4B,C). Carotenoids were extracted from these holo-EuCBPs and analyzed by HPLC. Based on the retention times, carotenoids in the peaks were identified as Zea and β-Car, respectively ( Supplementary Fig. S4).
To test whether EuCBPs bind to keto-carotenoids, the Euocp, Euhcp, and Euctdh genes were expressed in Can-producing E. coli. To verify that the EuCBPs expressed in Can-accumulating E. coli are photoactive, we measured the absorption spectra of purified holo-EuCBPs (Fig. 4E,F) and holo-SyOCP (Fig. 4D). A shift in the spectrum of the isolated EuOCP toward red light (active OCP) was detected upon exposure to blue light but, like SyOCP, active OCP reverted back to the inactive form in darkness. The maxima of the absorption spectra of both active and inactive EuOCP were slightly different from that of SyOCP. However, the time involved in the conversion from the active to inactive state, and the efficiency of chromophorylation (A 472 /A 280 vs. A 467 /A 280 ), were almost identical between EuOCP and SyOCP. This implies that the function of EuOCP is similar to that of SyOCP in the presence of keto-carotenoids. Recombinant EuHCP and EuCTDH had the color of orchids and violets, respectively. The maxima of the absorption spectrum of EuHCP and EuCTDH were 528 and 564 nm, respectively (Fig. 4F), similar to that of CTDH from Thermosynecoccus elongatus 11 .
These Eucbps were expressed constitutively in Euhalothece cells containing 0% or 3% NaCl (Supplementary  Table S2). Interestingly, qRT-PCR analysis revealed that transcript levels of cruA, hcp, and ctdh genes increased upon salt deprivation (Fig. 2), suggesting that carotenoid biosynthesis and CBPs play important roles in Euhalothece under conditions of salt deprivation stress.  38 , and OCP and HCP exhibit antioxidant properties 9,39 . Thus, we hypothesized that E. coli producing both carotenoids and CBPs would show better survival under ROS stress than wild-type cells. In the present study, ROS production was induced by cefatoxime and gentamycin, which kill bacteria by inducing ROS production 40 . Inhibitory effects of antibiotics on E. coli growth were observed; however, the intensity of these effects was low in the presence of Zea alone or Zea-bound OCPs and HCPs. Under antibiotic-induced ROS stress, all three E. coli strains containing SyOCP, EuOCP, and EuHCP showed higher survival rates than E. coli cells containing only Zea (Fig. 5A). To identify whether the survival rate is related to the ROS scavenging effect of Zea and holo-EuCBPs, we compared the relative ROS content of E. coli cells using the fluorescence dye DCFH-DA. Higher ROS content was detected in all E. coli cells treated with antibiotics, although E. coli cells treated with gentamycin produced more ROS than those treated with cefatoxime. This result explains the lower survival rate of gentamycin-treated E. coli cells. After treatment with both antibiotics, ROS content in all three E. coli strains containing CBPs was lower than that in E. coli cells containing only Zea (Fig. 5B). We did not include CTDH for this in vitro assay because CTDH did not bind to Zea (Supplementary Fig. S4) 11 . These results strongly imply that EuCBPs are effective ROS scavengers. Indeed, the absence of ketolase and cbp genes is one of the main factors responsible for the sensitivity of Prochlorococcus species to high irradiance, and their inability to degrade ROS 41,42 . β-car and Zea content, RoS content, and levels of crt and cbp gene transcripts at the early phase of salt deprivation. The involvement of carotenoids β-Car and Zea, and the corresponding EuCBPs, in ROS scavenging can be inferred from the inverse relationship between carotenoid and ROS content. To verify this inference, Euhalothece cells grown in 3% NaCl were transferred to salt-deficient growth medium (0% NaCl) for 1 h. Cells acclimated to salt deficiency for 6 days showed a 4.5-fold increase in total carotenoid content (Fig. 1C); in addition, short-term exposure to salt deprivation increased the content of β-Car and Zea by 1.2-1.3-fold, but decreased that of ROS by 30% (Fig. 6A and B). This inverse relationship implies that carotenoid-dependent ROS scavenging is crucial for responses to salt deprivation. Transcript levels of hcp and ctdh increased several-fold (Fig. 6C) and remained higher even after acclimation to salt-deficient conditions (Fig. 2); by contrast, those of ocp remained unchanged in response to altered salt concentrations (Figs. 2 and 6C). However, expression of crtR and cruA increased transiently upon salt deprivation (Figs. 2 and 6D). The different responses of carotenoid biosynthesis and CBP genes to salt stress suggests that CBP-dependent ROS scavenging in Euhalothece is regulated at both the transcriptional and post-transcriptional level. www.nature.com/scientificreports www.nature.com/scientificreports/ In conclusion, among multiple salt acclimation strategies, maintenance of low internal water potential within Euhalothece cells is achieved by the MRP system pumping out intracellular Na + ions, a compatible solute glycine betaine and exogenous Na + -chelating EPSs. Additionally, Euhalothece cells are equipped with β-Carand  Genomic DnA isolation and purity assessment. Genomic DNA was isolated as described previously 43 .
To assess the purity of the extracted genomic DNA, a 16S rRNA gene was amplified by PCR in a 25 μl volume containing 2 mM of each universal 16S rRNA primer (27 F and 1492 R), 200 μM dNTPs, 1 × master mix, and 1 U DNA polymerase (Apex Taq; Genesee Scientific Corp.). PCR was performed using the following protocol: initial denaturation at 95 °C for 10 min, followed by 30 cycles of denaturation at 90 °C for 30 s, annealing at 50 °C for 30 s, and extension at 72 °C for 2 min, and a final extension at 72 °C for 5 min. The PCR products were precipitated with ethanol and resuspended in 15 μl nuclease-free water. The 16S rRNA gene amplicon (20-50 ng μl −1 ) was sequenced using commercially available universal primers 27F and 1492R (Supplementary Table S3).
Whole genome sequencing, assembly, and annotation. Genomic DNA was fragmented and sheared using a g-TUBE (Covaris Inc., Woburn, MA, USA) and then purified using AMPure PB magnetic beads (Beckman Coulter Inc. Bread, CA, USA). A 10 μL library was prepared using the PacBio DNA template Prep Kit 1.0 (for 3-10 kb). SMRTbell templates were annealed using PacBio DNA/Polymerase Binding Kit P6. Libraries were sequenced on PacBio SMRT Cell using the PacBio Sequel (Pacific Biosciences) sequencing platform (Macrogen; Seoul, Korea), yielding 490,426 reads (average read length, 9,180; N50 value, 16,358; total read length, 4,502,006,607 bp). Some reads that were fully contained in other reads provide no extra information with respect to constructing the genome; these were filtered out. After quality control, 81,788 reads were recovered (average length, 5,709; N50 value, 9,479; total read length, 466,969,904 bp). The filtered reads were used to assemble the genome using HGAP4 v4.0, with default options. The genome sequence was circularized using the Circlator algorithm and annotated using the RAST server or the NCBI Prokaryotic Genome Annotation Pipeline (PGAP). The Circos program was used to draw the chromosome map. Secondary metabolites were predicted using antiSMASH. CRISPR loci were detected using CRISPRFinder. The genome sequence was deposited in the NCBI BioProject under accession number PRJNA557809.

RNA isolation and quantitative real-time PCR (qRT-PCR).
Total RNA was isolated from Euhalothece, as described previously 44 , with some modifications. The extracted total RNA was treated with DNaseI (Takara) and then cleaned using the NucleoSpin RNA Cleanup Kit (Macherey-Nagel). First-strand cDNA was synthesized using the iScript cDNA Synthesis Kit (Bio-Rad), and qRT-PCR was carried out on the CFX96 Real Time System (Bio-Rad) using SYBR Green Supermix (Bio-Rad) and gene-specific primers (Supplementary Table S3). The 16S rRNA gene was used as the control in each trial of qRT-PCR.
RNA sequencing and data analysis. Construction of the cDNA libraries, RNA sequencing, and data analysis were carried out as described previously 45 . The Ribo-Zero rRNA Removal Kit (Epicentre, USA) was used to deplete rRNA from 2 μg total RNA. The cDNA libraries were quantified using Bioanalyzer 2100 (Agilent, USA) and sequenced on the Hi-seq 2500 (Illumina, USA) platform to generate 100 bp paired-end reads. The quality of the sequenced reads was evaluated using SolexaQA v1.13, and cleaned reads were extracted by removing low-quality (phred score <20) reads from the dataset 46 . To measure expression of each transcript, cleaned reads were mapped to transcripts using bowtie2 v2.1.0 (mismatch <= 2) and then normalized using the DESeq v1.22.1 library in R 46,47 . Statistical analysis was performed, and genes showing a 2-fold change between compared samples and a false discovery rate p-value < 0.01 in a binomial test were identified as differentially expressed. Dendrograms and heatmaps of differentially expressed genes were generated using amap v0.8-14 and gplots v3.0.3 library in R 48,49 . www.nature.com/scientificreports www.nature.com/scientificreports/ production of carotenoids with or without cBps in E. coli. E. coli cells transformed with PAC-ZEAX and PACtrcp-TPR6673 plasmids encoding various carotenoids were grown at 28 °C for 20 h, and Zea or Can production was induced by the addition of 0.2 mM IPTG. Then, Zea-or Can-binding holo-CBPs were produced by the transformation of pBAD-CBP plasmids into Zea-or Can-producing E. coli strains. The transformed cells were grown at 37 °C for 20 h. Then, 0.02% arabinose was added to the cultures, and the cells were grown at 28 °C for 20 h. The induced cells were harvested by centrifugation at 3,500 rpm at 4 °C for 10 min, and the cell pellets were stored at −70 °C until needed. LB and RM media were used for BL21 and LMG194 cell cultures, respectively. Chloramphenicol (35 μg/ml) and ampicillin (100 μg/ml) were used for selecting and maintaining the transformed cells. pigment levels. Chlorophyll (Chl) a, phycocyanin (PC), and carotenoid (Car) levels in cells grown for 3 days on S-liquid media were estimated using whole-cell spectra 52 or 100% N,N,-dimethylformamide extracts 53 .

Construction of plasmids expressing carotenoid biosynthesis and cbp genes. The PAC-ZEAX
Carotenoid extraction and HPLC analysis. Carotenoids from lyophilized cyanobacterial cells (15 mg dry cell weight) were extracted using HPLC-grade methanol by sonication for 1 h, followed by centrifugation at 12,000 rpm at 4 °C for 20 min. The supernatants were collected, and the centrifugation was repeated two more times. The supernatants from each centrifugation were pooled, evaporated, concentrated under nitrogen gas, and dissolved in HPLC-grade acetone. The dissolved residues were filtered before HPLC analysis. To extract carotenoids from E. coli cell cultures, cold hexane:acetone (1:1, v/v) solution was used 54 . HPLC was performed using an Agilent 1100 HPLC system (Agilent Technologies, Palo Alto, CA, USA) equipped with a PDA detector and YMC carotenoid C 30 column (5 μm, 250 mm × 4.6 mm; Waters Corp., Milford, MA, USA). The mobile phase consisted of solvent A (92% methanol in 10 mM ammonium acetate) and solvent B (100% methyl tert-butyl ether) at varying ratios 55 . The flow rate and column temperature were maintained at 1 ml/min and 40 °C, respectively. Twenty microliters of each sample was injected into the HPLC machine. To detect carotenoids, the absorbance of samples was measured at 450 nm. Carotenoid species were determined by comparing the peaks of samples with those of standards (CaroteNature Lupsingen, Switzerland). The concentration of each carotenoid was determined as described previously 56 , and all procedures were conducted in darkness.
isolation of holo-cBps. Holo-CBPs were isolated from E. coli cells in the dark, as described previously 39 , with some modifications. Frozen pellets of E. coli cells were resuspended in the lysis buffer containing 50 mM Tris-HCl (pH 8), 100 mM NaCl, 0.1% IGEPAL, and 10% glycerol, and the mixtures were sonicated for 1-1.5 h. After sonication, the samples were centrifuged at 13,000 rpm for 10 min at 4 °C. The supernatant was transferred to a tube containing Ni-NTA resin, previously resuspended with 300 mM NaCl and 20 mM imidazole at 4 °C for 2 h. After incubation, the Ni-NTA resin was transferred to a column and then washed with 50 mM Tris-HCl buffer (pH 8.0) containing 300 mM NaCl, 0.1% IGEPAL, 10% glycerol, and 20 mM imidazole. CBPs were then eluted using 50 mM Tris-HCl buffer (pH 8.0) containing 300 mM NaCl and 250 mM imidazole. The eluted proteins were stored at −70 °C until needed for further analysis.
Antibiotic treatment of E. coli. E. coli BL21 cells carrying the PAC-ZEAX plasmid alone or together with pBAD-CBPs were grown at 37 °C in LB media containing chloramphenicol (35 μg/ml) or both chloramphenicol (35 μg/ml) and ampicillin (200 μg/ml), respectively. After 6 h, the cultures were transferred to 28 °C and then grown for another 12 h. Six hours after the transfer to 28 °C, 0.02% arabinose was added to BL21 cells containing both PAC-ZEAX and pBAD-CBP plasmids, and the cultures were grown for another 6 h. Then, fresh LB media was added to the overnight cultures to dilute them to an O.D. 600 of 0.1. The cultures were grown further to an O.D. 600 of 0.4-0.6; BL21 cells without plasmids were grown at 37 °C, while BL21 cells containing PAC-ZEAX alone or both PAC-ZEAX and pBAD-CBP were grown at 28 °C. The cells were diluted again using 5 mL LB in a 6-well plate (O.D. 600 = 0.1), and cefatoxime (50 μg/ml) or gentamycin (60 μg/ml) was added to the culture. The culture was grown at 37 °C and 130 rpm, and O.D. 600 was measured after 2 h. All cultures were grown in the dark.

RoS detection.
Overnight cultures of E. coli were diluted to an O.D. 600 of 0.1 using fresh LB media containing 5 μM 2′,7′-dichlorodihydrofluorescein diacetate (DCFH-DA). Euhalothece cells grown in the presence of 3% NaCl or 1.5 h after transfer to salt-deficient S-media were suspended in 1 mL 1X phosphate-buffered saline containing 5 μM DCFH-DA (O.D. 580 = 0.2). Emission spectra were measured using a fluorescence spectrometer (LS-55, Perkin Elmer, Waltham, USA) at an excitation wavelength of 480 nm. Fluorescence emission at 520 nm was used to determine the ROS production.
Extraction and quantification of exopolysaccharides. Cells grown for 7 days in S-solid medium were inoculated in S-liquid medium containing 3% NaCl or 0% NaCl (O.D 750 = 0.7). The cultures were grown for 9 days at 28 °C under continuous white light at an intensity of 10 µmol m −2 s −1 . Samples (2 ml) were taken from the cultures and centrifuged at 12,000 rpm for 10 min. After washing twice with 1 mL ultra-pure water, the pellets were resuspended in 1 mL ultra-pure water and boiled for 15 min at 100 °C. After cooling and centrifugation at 13,000 rpm for 15 min, supernatants were taken to measure the concentration of total carbohydrate contents of EPSs using the phenol-sulfuric acid method 57 . phylogenetic analysis. Sequences of the 16S rRNA genes of 58 cyanobacteria, whose taxa were already verified 22 , were retrieved from GenBank. The 16S rRNA gene sequences of four additional species (Euhalothece Z-M001, Euhalothece Z9404, Microcoleus IPPAS B-353, and Nostoc punctiforme ATCC29133) were confirmed by sequencing and added to the retrieved sequences. Multiple alignments were made using MAFFT ver.7 with the L-INS-i parameter 58 . "The maximum likelihood (ML)" tree was constructed using PhyML and the best-fitting substitution model was selected as GTR + G + I using Akaike information Criterion (AIC) [59][60][61] .