Introduction

Microbial exopolysaccharides (EPS) are high-molecular-weight polymers produced by microorganisms that inhabit diverse ecological niches1. Interest in bacterial and archaeal EPS has grown recently because of their environmentally friendly properties, including non-toxicity, biocompatibility, biodegradability and distinctive physicochemical characteristics2. EPSs are composed of sugar residues secreted by microorganisms into their surrounding environment3. Their composition and structure vary widely: they may be homopolysaccharides or heteropolysaccharides, and can include a variety of organic and inorganic substituents4.

Microbial EPSs have numerous application in pharmaceuticals, nutraceuticals, functional foods, cosmetics, and insecticides1,3,4,5. They are used as thickeners, stabilizers, prebiotics, gelling agents, emulsifiers, viscosifiers, biosurfactants, bioabsorbents, and drug-delivery vehicles1,6. Additionally, EPSs display antioxidant, anticancer, and cytoprotective activities, and can serve in bioremediation as bioflocculants and heavy-metal binding agents3,7. These properties support pollutant removal, wastewater treatment, bioremediation, and soil stabilization, thereby improving soil quality6. Despite their potential, commercial EPS-derived products remain under development and require substantial research to optimize production processes. A major barrier is the high cost of EPS production. The circular bioeconomy offers a promising approach to improve cost-effectiveness by using waste substrates, such as molasses and whey, for microbial EPS production, particularly by extremophilic microbes7,8,9,10.

Microorganisms isolated from extreme environments often produce EPS with distinctive chemical and physical properties that set them apart from biopolymers found elsewhere in the biosphere11. Such EPSs form a substantial part of the extracellular polymeric matrix surrounding microbial cells in extreme habitats like Antarctic ecosystems, marine environments, saline lakes, geothermal springs, and deep-sea hydrothermal vents1,7,12,13,14,15,16. Extremophiles employ multiply adaptation strategies to withstand stressors such as high temperatures, elevated salinity, low pH, and intense radiation. EPS biosynthesis is among the most widespread protective mechanisms17.

Hypersaline environments, characterized by very high salt concentrations, pose particular challenges to life. Many halophiles synthesize EPS as an adaptive and protective strategy3. Halophilic EPS producers are of growing industrial interest because their biopolymers often possess unique chemical composition and properties16,18. Moreover, EPSs from halophiles are non-pathogenic and thus suitable for applications in the food, pharmaceutical and cosmetic industries19,20,21,22.

EPS production occurs in both Bacteria and Archaea. Among Archaea, genera such as Haloferax, Haloarcula, Halococcus, Natronococcus, Halorubrum, Haloterrigena, and Halobacterium are known EPS producers23,24,25,26,27,28,29. Notable strains include Haloferax mediterranei (ATCC 33500), Haloferax gibbonsii (ATCC 33959), Haloferax denitrificans (ATCC 35960)30, and Haloarcula japonica T5 (DSM 12772)31, which produce substantial amounts of EPS. For example, H. japonica T5 synthesizes 370 mg L⁻1 of EPS from glucose, and its polymer is composed of mannose, galactose, and glucuronic acid in a relative ratio of 2:1:317.

Despite their significance, the genetic basis of EPS biosynthesis in Archaea remains poorly understood32. EPS production typically involves a complex regulatory network with multiple interacting components33. Genome sequencing and annotation are therefore valuable tools for elucidating the genetic determinants of biopolymer synthesis.

This study investigate EPS production and the chemical composition of H. japonica strain SST1, isolated from a subterranean salt deposit in Avan, Armenia. In addition, genomic analyses were performed to identify putative genes involved in EPS biosynthesis and assembly. The feasibility of using molasses as a cost-effective substrate for EPS production by H. japonica SST1 was also assessed in the context of circular bioeconomy principles. Finally, the potential biomedical applications for EPS were analyzed with respect to modulation of the innate immune system. Recognition of carbohydrate patterns is a central immunological mechanism mediating molecular communication between microbes, their hosts, and the environment34. Hosts possess a diverse repertoire of pattern-recognition receptors (PRRs) that detect damage- and microbe-associated molecular patterns. Among these, lectins, a conserved family of carbohydrate-binding proteins, play key roles in modulating innate immune responses. C-type lectins are characterized C-type lectin-like domains that typically mediate Ca2+-dependent carbohydrate binding. To explore how the EPS from H. japonica SST1 interacts with the host immune system, we evaluated the binding of several human C-type lectins – dendritic cell-specific intercellular adhesion molecule-3-grabbing non-integrin (DC-SIGN), Langerin, the mannose receptor (MR) and macrophage galactose lectin (MGL) to the EPS using an ELISA-based solid-phase assay.

Materials and methods

Isolation and phenotypic characterization of archaeal strain SST1

Salt stones from a subterranean salt deposit in Avan, Armenia (40°13′30.15″ N; 44°33′50.32″ E) were sampled and stored at 4 °C until further use. Sampling was conducted at a depth of 300 m in the salt mine (Figure S1). The Avan subterranean salt deposit represents remnants of ancient hypersaline waters from the Neo-Tethys Ocean during the Pyrenean epoch35. The chemical composition of the deposit includes NaCl (93%), Ca2+ (0.8%), Mg2+ (0.8%), SO42− (2.0%), and insoluble residues (5%)36.

To enrich aerobic halophilic archaea, the salt stone samples were milled under sterile conditions, and 1 g of the resulting salt sediment was added to medium A, which contained (g L−1) NaCl, 250; yeast extract, 5; casamino acids, 5; Na-glutamate, 1; KCl, 2; Na-(tri)citrate, 3; MgSO4·7H2O, 20; FeCl3·4H2O, 0.036; MnCl2·4H2O, 0.036; with a final pH of 7.237. The medium was sterilized by autoclaving at 121 °C for 15 min. Enrichment cultures were incubated in aerated, sterile flasks at 37 °C under shaking at 150 rpm for 14 days. Subsequently, 0.5 mL aliquots of appropriately diluted enrichment cultures were spread onto plates containing the same medium supplemented with 1.5% (w v−1) agar and incubated at 37 °C until visible growth appeared. Colonies with distinct morphologies were selected and further purified by streaking onto fresh solid medium at least three times. Cell morphology and shape were examined using both light (OMAX A3590U) and phase-contrast microscopy (Nikon; Eclipse E400) of freshly prepared wet mounts after fixation with 2% (w v−1) acetic acid38. A modified fixation method for halophilic archaea was used to prepare samples for scanning electron microscopy (SEM)39. The procedure involved double fixation. First, cells were prefixed in a mixture of 6% glutaraldehyde and 1% OsO4, in 0.1 M cacodylate buffer for 2 h, followed by post fixation with 1% OsO4 in 0.1 M cacodylate buffer for an additional 2 h. To maintain osmotic stability, 20% sodium chloride was added to the cacodylate buffer. Following fixation, specimens were dehydrated through a graded ethanol series, processed using a Polaron critical point drying apparatus, and dried with liquid CO2. The samples were then coated with gold using a sputter coater and visualized with a Zeiss Supra 55Vp scanning electron microscope at the Molecular Imaging Center Platform, University of Bergen (https://www.uib.no/en/rg/mic).

Tests for catalase and oxidase activity, as well as the hydrolysis of casein, gelatin and starch, and nitrate reduction, were performed following the methods as described37. The utilization of d-glucose, l-propionate, l-arginine, d-cellobiose, casamino acid, and Na-glutamate was also examined individually by incorporating each into the basal medium as a sole carbon and energy source. The growth of the archaeal isolate was evaluated across a pH range of 4.0–12.0 at 37 °C. The NaCl requirement for growth was assessed in the same medium supplemented with 0, 5, 10, 15, 20, 25 or 30% (w v−1) NaCl. Similarly, the Mg2+ requirement was studied by supplementing the medium with 0, 20, 50, 70, 100, or 200 mM Mg2+. Growth was also tested at different temperatures (20–45 °C) to determine the optimal growth conditions. Growth was monitored by measuring optical density (OD) at 540 nm. Each experiment was performed in triplicates. Two-dimensional (2D) thin-layer chromatography (TLC) of the polar lipids of the strain was performed at DSMZ. For this analysis, 1 g of wet biomass was collected from an actively growing liquid culture. Polar lipids were extracted from freeze-dried cell material using chloroform/methanol/aqueous NaCl (0.3%) solution as described40. Lipids were recovered in the chloroform phase and separated by 2D silica gel TLC. The first dimension was developed using chloroform/methanol/water, while the second dimension used chloroform/methanol/acetic acid/water. Total lipid material was visualized with molybdatophosphoric acid, while specific functional groups were detected using anisaldehyde staining. TLC plates were heated to 150 °C for 10 min to reveal glycolipids and phospholipids as described41. The strain is maintained in the Microbial Culture Collection at the Department of Biochemistry, Microbiology and Biotechnology, YSU.

DNA isolation and genome sequencing

DNA extraction was carried out using a modified cetyltrimethylammonium bromide (CTAB) protocol42. The strain was grown in the specified liquid medium, harvested by centrifugation at 10,000 rpm for 5 min, and resuspended in TE buffer (10 mM Tris–HCl, pH 8.0, and 1 mM EDTA, pH 8.0). To the lysate, 10% SDS, RNase A, and Proteinase K (20 mg mL−1) were added, followed by incubation at 55 °C for 1 h. Subsequently, 5 M NaCl and CTAB/NaCl were added, and the mixture was incubated at 65 °C for 20 min. DNA purification was performed using chloroform:isoamyl alcohol (24:1), followed by centrifugation at 15,000 rpm for 10 min. The clear supernatant was transferred to a fresh tube, mixed with 0.6 volumes of isopropanol, and stored at -20 °C overnight. The mixture was centrifuged at maximum speed, and the resulting DNA pellet was washed with 70% ethanol and centrifuged for 5 min. The purified DNA pellet was air-dried and stored at -20 °C for further use. Whole-genome shotgun sequencing of the strain was performed using Illumina paired-end technology at Eurofins Genomics, Germany (www.gatc-biotech.com).

Whole genome assembly, gene prediction and annotation

The sequence data were assembled using CLC Genome Workbench software with the default parameters. Genome completeness was assessed using the CheckM (v1.0.18) (https://github.com/Ecogenomics/CheckM)43. The genome of strain SST1 was annotated using multiple platforms including the National Center for Biotechnology Information (NCBI) Prokaryotic Genome Annotation Pipeline (PGAP)44 (www.ncbi.Nlm.nih.gov/genome/annotation_prok/), Rapid Annotation using Subsystem Technology (RAST)45 (http://rast.nmpdr.org/), the Department of Energy Systems Biology Knowledgebase (KBase) (https://www.kbase.us/), UniProt (v.4.0)46 (https://www.uniprot.org/) and the Kyoto Encyclopedia of Genes and Genomes (KEGG) (v.114.0)47 (https://www.kegg.jp/kegg/). Digital DNA-DNA hybridization (dDDH) was performed using the Genome-to-Genome Distance Calculator (GGDC 2.1) provided by DSMZ (https://tygs.dsmz.de/) with recommended parameters48. Haloarcula japonica DSM6131, Haloarcula salaria JCM15759, Haloarcula tradensis JCM15760, Haloarcula argentinensis DSM12282, Haloarcula sebkae JCM19018T, Haloarcula amylolytica JCM13557, Haloarcula hispanica ATCC33960, Haloarcula marismortui ATCC43049, Haloarcula quadrata DSM11927, Haloarcula sinaiiensis ATCC33800, Haloarcula salina YGHS18, Haloarcula rubra F13, Haloarcula halobia XH51, Haloarcula ordinaria ZS-22-S1, Haloarcula marina DT1T, H․ halophila DFY41 were used as reference strains for dDDH comparisons. Average nucleotide identity (ANI) between two genomic datasets (H. japonica SST1 and H․ japonica DSM6131) was calculated using the OrthoANIu algorithm on the EzBioCloud server (https://www.ezbiocloud.net/tools/ani)49,50. The phylogenomic tree of SST1 and other Haloarcula strains was constructed using the TYGS platform (v400) (https://tygs.dsmz.de/). Carbohydrate-active enzymes (CAZy) families were analyzed using Hidden Markov Models (HMMs) from the dbCAN2 database (v10) (https://www.cazy.org/). EPS biosynthetic gene clusters were annotated with the antiSMASH server (v7.0) (https://antismash.secondarymetabolites.org/)51. The circular genome of SST1 and EPS gene cluster structure were visualized using Proksee including Alien Hunter (v1.3.0), features (v2.0.3), GC Content and GC Skew (v1.0.2) (https://proksee.ca/)52. Pairwise comparison of EPS gene clusters in Haloarcula strains was carried out using BLASTp within pyGenomeViz (v1.0.0) (https://moshi4.github.io/pyGenomeViz/)53. The KAAS–KEGG (https://www.genome.jp/kegg/kaas/) Automatic Annotation Server was used to reconstruct sugar metabolic pathways54,55,56. Gene features of essential biosystems were manually verified using BLASTp against the NCBI non-redundant database (https://sky-blast.com/blast/p).

EPS production and recovery

The screening for EPS production was carried out by incubating the strain at 37 °C and pH 7.5 on solidified medium B containing (g L−1) single carbon source, 20; yeast extract, 2; casamino acid, 1; MgSO4·5H2O, 21; MgCl2, 18; KCl, 4.2; NaCl, 200; CaCl2, 0.44 and 12 mM Tris HCl (pH 7.5). Sterile stock solutions of glucose, fructose, sucrose, arabinose, ribose, xylose, maltose, or galactose were prepared by filtration (0.22 µm size-pore) and added to medium B at a final concentration of 20 g L−1. Colony mucosity on the solid medium was used as an indicator of potential EPS production.

EPS production was further assessed in liquid medium B supplemented with 20 g L−1 glucose, maltose or sucrose as the sole carbon source. In addition, molasses-based medium was used as a cost-effective alternative. Molasses (supplied by Co.Pro.B, Italy (http://www.coprob.com)), containing 50% (w v−1) sucrose7, was added to medium B at a final concentration of 40 mL L−1. Batch cultures were incubated at 150 rpm, 37 °C, and pH 7.0 for up to 120 h to monitor microbial growth and EPS production kinetics. Growth was quantified by measuring OD at 540 nm wavelength. After incubation, cells were harvested by centrifugation at 7,000 rpm for 20 min, freeze-dried, and weighed using by analytical balance to determine the dry cell weight (DCW). For EPS recovery, the supernatant was mixed with an equal volume of cold absolute ethanol under stirring in an ice bath and stored at -20 °C overnight. The precipitated EPS was collected by centrifugation (at 10,000 rpm for 45 min at 4 °C), dissolved in hot distilled water (70 °C) and dialyzed against tap water for five days using dialysis tubes (Spectra/Por MWCO; molecular mass cut-off 8 kDa). Finally, the raw extracellular product (EP) was lyophilized using a Heto drywinner and weighed.

EP chemical composition and EPS molecular mass determination

The carbohydrate content of the raw EP was determined using the phenol–sulfuric acid method with glucose as the standard57. A calibration curve was generated using glucose solution at a concentration of 1 mg mL−1. Protein content was quantified by the Lowry method58, with bovine serum albumin (BSA) as the standard in the concentration range of 0–50 µg mL−1. Uronic acids were measured colorimetrically with the m-hydroxybiphenyl reagent, using glucuronic acid as the standard in the concentration range of 0–50 µg mL−159. Nucleic acid content was assessed spectrophotometrically at 260 nm, using standard conversion factors (for dsDNA, a factor of 50 was used). All assays were performed in duplicate.

The monosaccharide composition of EPS was analyzed after hydrolysis with 0.5 M trifluoroacetic acid at 120 °C for 2 h. Separation and identification of sugars was carried out as previously described60 using both TLC and high-pressure anion-exchange chromatography with pulsed amperometric detection (HPAE-PAD) on a DIONEX ICS-5000+ DC system. For TLC, silica plates were developed in a butanol/acetic acid/water (6/2/2, by volume) solvent system and visualized using α-naphthol solution. The relative molar ratios of monomer sugars were determined by HPAE-PAD using a CarboPAC PA1 column. Elution was carried out isocratically with 16 mM NaOH at flow rate of 0.25 mL min−161. Monomer quantification was based on external calibration curves, with standard retention times of 12.97, 14.03, and 15.70 min for galactose, glucose, and mannose, respectively. Monosaccharide analysis was performed by GC–MS of acetylated methyl glycosides62. Briefly, EPS samples (0.5 mg) were subjected to methanolysis with HCl/CH3OH (1.25 M, 1 mL) at 80 °C for 16 h. After methanol evaporation, the resulting methyl glycosides were acetylated with acetic anhydride in pyridine at 100 °C for 30 min, dried, dissolved in 400 µL of acetone, and injected into a Thermo Scientific Focus GC Series instrument equipped with a TG-SQC capillary column (30 m × 0.25 mm × 0.25 µm, He as carrier gas, flow rate: 1 mL min−1). The acetylated methyl glycosides were analyzed using the following temperature program: 150 °C for 3 min, followed by a ramp from 150 to 330 °C at 3 °C min− 1.

The molecular mass of EPS was estimated by gel filtration chromatography on a Sephacryl 300 HR column (Sigma) eluted with 0.2 M acetate buffer (pH 5.1) at a flow rate of 0.3 mL min−163,64,65. The column was calibrated with standard dextrans (100, 500, and 1000 kDa, Fluka). Fractions of 3.5 mL were collected and analyzed using the α-naphthol spot test and the Dubois assay as previously described.

Nuclear magnetic resonance (NMR) analysis of EPS

1H and 13C one-dimensional NMR spectra of EPS released by strain SST1 were recorded using a Bruker 400 MHz spectrometer at 50 °C. For 1H NMR analysis, the sample was exchanged twice with D2O, including an intermediate lyophilizing step and then dissolved in 700 µL of D2O. Chemical shifts (δ) are reported in parts per million (ppm) relative to sodium 2,2,3,3-d 4-(trimethylsilyl) propanoate for 1H NMR and to CDCl3 for 13C NMR66.

ELISA-based solid-phase assay

50 µL of EP solution from strain SST1 prepared in PBS (10 mM, pH 7.4 at a concentration of 10 µg mL−1), were used to coat Nunc MaxiSorp wells overnight at 4 °C. After discarding the solution, wells were washed twice (2 × 150 µL) with calcium- and magnesium-containing buffer TSM (20 mM tris(hydroxymethyl)aminomethane (Tris)-HCl, pH 8.0; 150 mM NaCl; 1 mM CaCl2; 2 mM MgCl2). Wells were then blocked with 80 µL of 1% BSA solution (Sigma-Aldrich, lyophilized powder, ≥ 96%, agarose gel electrophoresis) prepared in TSM for 30 min at room temperature. After blocking, the solution was discarded and 50 µL of different C-type-lectins including human-Fc (DC-SIGN, Langerin, MGL and MR) at 1 µg mL−1, were added. Following 1 h incubation at room temperature, wells were washed twice with TSM (2 × 150 µL) and 100 µL of anti-human horseradish peroxidase (Goat anti-human IgG-HRP from Jackson ImmunoResearch, 0.3 µg mL−1) was added. After 30 min incubation at room temperature, wells were washed with TSM (2 × 150 µL). Subsequently, 100 µL of substrate solution (3,3′,5,5′- tetramethylbenzidine, TMB, in citric/acetate buffer, pH 4, containing H2O2) was added. After 5 min incubation at room temperature, the reaction was stopped by adding 50 µL of 0.8 M H2SO4, and OD was measured at 450 nm using an ELISA reader. The experiment was performed in duplicate and data were normalized to the 450 nm signal obtained from the respective positive controls for each C-type lectin. BSA was used as blocking agent throughout the assay and also served as the negative control. Polyacrylamide polymers functionalized with different glycans were purchased from Lectinity (molecular mass ~ 20 kDa, carbohydrate content ~ 20 mol%) and used as positive controls: Galβ1-4(Fucα1-3)GlcNAcβ-OCH2CH2CH2NH2 0044-PA (PAA-LeX), positive control for DC-SIGN; GalNAcα-OCH2CH2CH2NH2 0030-PA (PAA-Tn), positive control for MGL; Fucα1-2Galβ1-4(Fucα1-3)GlcNAcβ-OCH2CH2CH2NH2 0045-PAA (PAA-LeY), positive control for Langerin. PAA-glyco-conjugates were coated onto ELISA well at concentration of 20 μg mL−1. Mannan from Saccharomyces cerevisiae (Sigma-Aldrich) was used as the positive control for MR at a concentration of 10 μg mL−167.

Statistical analysis

Data are presented as the mean ± standard error of the mean. Statistical analyses were performed using paired t-tests, with significance defined at a P < 0.05. For the C-type lectin solid-phase assay, a two-way ANOVA with Tukey’s multiple comparison test (α = 0.05) was conducted using GraphPad Prism 10.

Nucleotide sequence accession numbers

The draft genome sequence data has been deposited in GenBank under the accession number JBLMKY000000000.

Results

Strain’s phenotypic characteristics

Five halophilic archaeal strains were isolated from salt stone samples collected from a subterranean salt deposit in Avan, Armenia. Among these isolates, one strain was identified as an EPS producer, as it was able to grow on medium B, forming colonies with a mucous consistency. This strain was designated SST1 and deposited at the Microbial Depository Center of Armenia under accession number MDC11864.

On solidified medium A, SST1 formed colonies approximately 1–1.5 mm in diameter after seven days of incubation at 37 °C. The colonies were transparent, entire, smooth, and exhibited red–orange pigmentation (Figure S2A, B). Cells were predominantly triangular or rhomboidal in shape, measuring approximately 0.82–0.92 µm in liquid medium (Figure S2C, D). Motility was observed by phase-contrast microscopy and confirmed by the identification of flagellar gene clusters.

The growth temperature range for SST1 was 21–45 °C, with an optimum at 37 °C. Growth occurred within a pH range of 4.0–12.0 at 37 °C, with an optimal pH of 7.0–7.5. The optimal NaCl concentration for growth was 20%, though growth was observed over the range of 5–25%. Notably, no growth occurred at NaCl concentrations below 5%. Growth was also supported with Mg2⁺ supplementation in the range of 0–200 mM, with maximum growth at 50 mM Mg2⁺.

The strain is chemoorganotrophic with an aerobic respiratory metabolism. It utilizes d-glucose, galactose, arabinose, xylose, sucrose, maltose, lactose, l-arginine, and casamino acids as sole carbon and energy sources but does not utilize fructose, Na-glutamate, l-propionate, or d-cellobiose. SST1 produces catalase and oxidase but does not reduce nitrate with gas production. The strain demonstrates amylase activity but cannot hydrolyze casein or gelatin. TLC of polar lipids revealed methylated derivatives of diether phospholipids, including phosphatidylglycerol, and phosphatidylglycerol phosphate, characteristic of C20C20 archaeol (Figure S3).

Phenotypic characteristics were compared with the following halophilic reference strains: H. japonica TR-1T DSMZ 613137, H. quadrata DSM 11927T68, and H. argentinensis arg-1T69. Results are summarized in Table 1. Physiological and biochemical analyses revealed greater similarity between strain SST1 and H. japonica DSM 6131T. Although strain SST1 shared common traits with other members of Haloarcula, it was distinguishable from its closest relatives by certain phenotypic features. Specific traits differentiating strain SST1 from H. japonica DSM 6131T include a lower optimum temperature, broader pH range, inability to reduce NO3 to NO2, and a distinct carbon utilization profile (e.g. hydrolysis of starch, utilization of lactose, inability to utilize fructose and cellobiose).

Table 1 Comparison of the phenotypic properties of halophilic archaea from Avan subterranean deposit with those of the closest Haloarcula species (1, H. japonica TR-1T; 2, H. quadrata DSM 11927T; and 3, H. argentinensis arg-1T.

Genome features of H. japonica SST1 and phylogeny

Taxonomic identification of strain SST1 was initially performed using 16S rRNA gene sequencing. However, the chromatograms revealed overlapping peaks due to intragenomic heterogeneity among multiple 16S rRNA gene copies, which is a known characteristic of Haloarcula species. The high sequence variability among paralogous 16S rRNA genes (differing by up to several percent) hindered the generation of a single consensus sequence suitable for submission to public databases. To resolve this ambiguity and ensure accurate identification, whole-genome sequencing was carried out.

Genome completeness assessed using CheckM indicated > 90% completeness and a < 5% contamination, suggesting minimal missing genes. A draft genome sequence of ~ 4.1 Mb across 14 contigs was obtained, containing 4,440 coding DNA sequences (CDSs) and 50 RNA genes (Table 2). The G + C content was 61.4 mol%. Identified CDSs were classified into subsystems: carbohydrates (73), amino acids and derivatives (177), protein metabolism (128), RNA metabolism (38), secondary metabolism (58), cell wall and capsule (7), membrane transport (28) and stress response (7).

Table 2 Genome features and statistics of strain SST1.

Pairwise genome sequence similarity calculated using the GGDC revealed 83.7% similarity between SST1 and DSM 6131T. dDDH values compering SST1 with 16 Haloarcula strains (Table S1) showed SST1 shared an ANI of 97.63% with H. japonica DSM 6131T. These values exceed the separation thresholds70, supporting that SST1 represents H. japonica. A genome-based phylogenetic tree (Fig. 1) confirmed SST1 clusters with H. japonica DSM 6131T, distinct from all other Haloarcula species.

Fig. 1
figure 1

The phylogenetic tree of the strain SST1 constructed using FastME version 2.1.6.1 based on GBDP distance calculations from genome sequences. Branch lengths were scaled according to GBDP distance formula d5. The numbers above branches represent GBDP pseudo-bootstrap support values (> 60%) from 100 replications, with an average branch support of 91.5%. The tree was rooted at the midpoint and the δ statistics were 0.163.

EPS production

Optimization of microbial biomass and EPS production was evaluated by varying incubation time and carbon sources. Culture were incubated at 37 °C with 20% NaCl, pH 7.0, for 120 h. Growth (OD540) peaked at 120 h in stationary phase. EPS production followed a similar pattern starting in mid-exponential phase and peaked in stationary phase (Figure S4), indicating growth-associated EPS production. Carbon source influence was tested using glucose, sucrose, or maltose in medium B at 37 °C for 96 h (late exponential/early stationary phase) (Table 3). All sugars supported growth. Glucose and maltose yielded high total EP (280 mg L−1), while sucrose produced slightly less (240 mg L−1) but with the highest carbohydrate fraction (133.4 mg L−1, 55.6%). In contrast, maltose yielded 96.4 mg L−1 (34.4%) and glucose only 61.0 mg L−1 (21.8%) carbohydrate fraction. The highest yield occurred with sucrose, where EP and EPS reached 628.4 mg L⁻1 and 304.4 mg L⁻1 (48.5%, w w−1), respectively (p < 0.0015), after 120 h. Thus, optimal EPS production was obtained in sucrose medium at 37 °C for 120 h.

Table 3 Production of EP (mg L−1) and EPS (mg L−1) by H. japonica strain SST1, obtained from different carbon sources after 96 h of incubation at 37 °C in the medium B.

Molasses was also tested as an alternative sucrose source. EP yield reached 275.2 mg L⁻1 in the stationary phase, with a lower carbohydrate fraction (57.2 mg L⁻1, 20.8%, w w−1) (Fig. 2).

Fig. 2
figure 2

EP and EPS production by the H. japonica strain SST1 from sucrose and molasses at the stationary phase of growth (120 h).

EP from sucrose medium (Table 4) contained relatively low levels of protein (4.32%) and nucleic acids (0.96%) but a high proportion of uronic acids (10.8%). Molasses-derived EP showed lower carbohydrate (20.8%) content, higher protein (10.3%) content, and reduced levels of uronic acids (1.9%) and nucleic acids (3.8%).

Table 4 EP composition of the H. japonica SST1 extracted from different carbon sources.

EPS structural and chemical characterization

EPS samples harvested after 120 h in sucrose medium at 37 °C, pH 7.0. 150 rpm were analyzed. Gel filtration chromatography (Sephacryl 300 HR) revealed heterogeneous molecular mass ranging from 10–100 × 103 Da.

Monomer composition was determined by TLC, HPAE-PAD and GC–MS of acid hydrolysates (Fig. 3A,B). EPS was heteropolymeric, with mannose as the main monomer, followed by glucose and galactose, in a 1:0.85:0.39 molar ratio (Fig. 3A).

Fig. 3
figure 3

EPS monomer composition: (A) HPAE-PAD analysis of the monosaccharide composition of EPS produced by H. japonica SST1 revealed the presence of galactose (RT: 12.97 min), glucose (RT: 14.03 min), and mannose (RT: 15.7 min). Man, mannose; Gal, galactose; Glc, glucose. (B) GC–MS analysis of the corresponding methyl glycoside acetates.

NMR confirmed heterogeneity of EPS. In the 1H-NMR spectrum, three main signals appeared at 5.31, 5.14, and 5.07 ppm, with weaker peaks at 5.20, 5.18, 5.12, 5.09, and 4.93 ppm (Fig. 4A). In the 13C-NMR spectrum, a dominant signal at 103.0 ppm was accompanied by weaker signals at 101.5 and 100.3 ppm (Fig. 4B). These results, combined with composition and size data, indicate EPS is a mixture of α-linked mannose, glucose and galactose polysaccharides with diverse glycosidic linkages.

Fig. 4.
figure 4

1H NMR (A) and 13C NMR (B) spectra of EPS-SST1.

Genomic insight into EPS biosynthesis

The circular genome map of strain SST1 highlighting EPS clusters is shown in Fig. 5A. Selected cluster structures are presented in Fig. 5B. Strain SST1 encodes eight clusters related to EPS biosynthesis, four of which cover key steps: sugar transport, precursor synthesis, elongation, modification, and export. Strain SST1 encodes transporters including ABC transporters (glucose: GtsA, GtsB, GtsC; maltose: MalK, MsmX, MsmK, MsiK) and a phosphotransferase system (PTS) for fructose (IIA, IIB and IIC components). Uridylate kinase (EC 2.7.4.22) catalyzes UMP phosphorylation using ATP.

Fig. 5
figure 5

A. Genome map of the incomplete assembly of strain SST1 visualized using the Proksee platform. The outermost circle (blue) highlights regions identified by Alien Hunter (v1.3.0). The next concentric circle displays the annotated coding sequences (CDSs). The gray circle represents the contigs, while the black middle circle indicates the percentage of GC content (v1.0.2). The innermost circle depicts the GC skew [(G–C)/(G + C)] with positive skew shown in green and negative skew in purple (v1.0.2). EPS clusters are highlighted in green, with notable clusters numbered (v2.0.3). The scale is presented in megabases (Mbp). B. Organization of the main EPS biosynthetic gene clusters in H. japonica SST1. Gene clusters were identified using antiSMASH server version 7.0 and visualized with the Proksee platform. The scale is given in kilobases (Kbp).

Clusters 1–3 contain glycosyltransferases and transcriptional regulators (IclR, LuxR, HxlR, TrmB, PadR). Enzymes for precursor synthesis are in Cluster 4 (EC 3.13.1.1, EC 5.1.3.2, EC 1.1.1.22, EC 2.7.7.9) and Cluster 2 (EC 5.1.3.14 EC 1.1.1.336 EC 5.1.3.2). Clusters 1 and 3 encode lipopolysaccharide biosynthesis proteins, while transporters occur in Clusters 1, 2, and 4. Cluster 3 also includes histidine kinase, succinoglycan biosynthesis protein, and transmembrane oligosaccharyl transferase. Cluster 2 encodes HPr phosphocarrier, alginate lyase and methyltransferase. Cluster 1 contains sulfatases, Gfo/Idh/MocA family proteins, GAF-domain proteins, and HalOD1 output-domain protein. Genetic determinants are detailed in Table S2. Table S3 presents the functional annotation data of proteins potentially involved in EPS production and transport in H. japonica strain SST1, as identified using RAST and BLASTp.

Biological activity: C-type lectins recognition

Raw EPS from H. japonica SST1 was recognized by DC-SIGN, Langerin, and MR, consistent with its expected carbohydrate patterns (Fig. 6). In contrast, MGL, which recognizes terminal α-GalNAc, served as a negative control and showed no significant binding. EPS from maltose-, glucose- or sucrose-grown cultures showed no major differences in lectin binding, except reduced MR binding in sucrose-derived EPS.

Fig. 6
figure 6

Solid-phase binding of EPS from H. japonica strain SST1 to human lectins. EPS produced with different single sugars was tested for binding to DC-SIGN, Langerin, MGL, and MR in calcium-containing buffer using Fc-tagged lectins. Detection was performed via HRP–TMB reaction and OD450​ measurement. Results in the table are shown as % binding relative to positive glycan-PAA controls. Data are mean ± SD from duplicate experiments. **** = p ≤ 0.0001; ns = p > 0.05.

Discussion

EPS-producing halophiles belonging to the genera Haloferax, Haloarcula, Halococcus, Natronococcus, Halorubrum, Haloterrigena, and Halobacterium have been isolated from diverse saline environments, including solar salt ponds, solar salterns, salt lakes, marine sediments, and saline soil crusts in regions such as Spain, Kenya, The USA, Japan, Israel, Tunisia, Turkmenistan, Iran, and Australia17,23,24,25,26,27,28,29,30,31,71,72. For some of these species, the chemical composition, structure, biosynthesis and functional properties of their respective EPSs have been described. However, little is known about the diversity of EPS-producing halophilic archaea in subterranean salt deposits in Armenia. In this study, we report for the first time the isolation and chemical characterization of the EPS produced by H. japonica strain SST1 from salt stone samples collected in the Avan subterranean salt deposit, Armenia.

EPS production is influenced by several factors, including temperature, pH, incubation time, carbon and nitrogen sources, organic compounds, and NaCl concentration. In this study, the effect of different carbon sources on H. japonica strain SST1growth and EPS production was examined. The results showed that supplementation of medium B with glucose, maltose or sucrose yielded similar biomass levels, however, these carbon sources had markedly different effects on EPS biosynthesis. Sucrose supplementation resulted in the highest EPS yield, nearly doubling that obtained with glucose. Consequently, the maximum EPS yield (Fig. 2) was achieved when strain SST1 was cultivated in sucrose-supplemented medium B.

The EPS productivity of strain SST1 was considerable compared to other archaea (Table 5), particularly within the genus Haloarcula. Notably, EPS yield from H. japonica SST1in sucrose-containing medium exceeded that of most Haloarcula and related species. The maximum yield reached 628.4 mg L−1. By comparison, other strains have produced sulfated EPS in minimal media containing glucose as the sole carbon source, with yield ranging from 35 to 370 mg L−1. Only H. mediterranei ATCC 33,500 and Halococcus sp. AMS12, isolated from a solar saltern in Alicante, Spain and marine sediments in Chennai, India, respectively, exhibited higher EPS productivity71,72. Among these, Halococcus sp. AMS12 produced 3172 mg L−1, the highest EPS yield reported to date among halophilic archaea.

Table 5 Comparative table of EPS production and composition by halophilic archaea.

Growth curve analysis and EPS synthesis kinetics revealed that EPS production in strain SST1 increased with biomass accumulation, suggesting a growth-associated process. This pattern has also been observed in Haloterrigena turkmenica and H. hispanica26,31. Under the tested conditions, EPS synthesis occurred from the early growth phase through the stationary phase, with maximum production after 120 h of incubation. Remarkably, this rate is at least twice as fast as that reported for other Haloarcula strains.

After 96 h of incubation culture volume began to decrease due to evaporation. It is plausible that EPS secretion by H. japonica SST1 functions as an adaptive response to impending desiccation and increased osmolarity. During the late exponential phase, EPS concentration stabilized and then slightly declined in the stationary phase (Fig. 4SA). Despite this adaptive increase in EPS, the carbohydrate content did not change significantly. The irregular fluctuations in carbohydrate levels may result from enzymatic degradation, possibly by glucohydrolases excreted into the medium73.

Halophilic archaea generally utilize a restricted range of organic substrates and often require complex nutrients, including amino acids, yeast extract or other supplements. For example, H. turkmenica produces 296 mg L−1 EPS in minimal media supplemented with 1 g L−1 yeast extract and 1% (w v−1) glucose26. Similarly, H. hispanica ATCC 33,960 produces up to 30 mg L−1 acidic and sulfated EPS in medium containing 5 g L−1 yeast extract and sugars such as mannose, galactose, or glucose71. Therefore, using a complex medium supplemented with yeast extract and casamino acid for H. japonica strain SST1 is consistent with previous findings for other EPS-producing archaea. With respect to NaCl, EPS synthesis has been observed in media containing 150–250 g L−1 NaCl26,31,71, but the physiological role of NaCl in EPS metabolism in halophilic archaea remains poorly understood.

Carbon source selection strongly influences EPS production costs. While sugars are the most commonly used substrates, agro-industrial wastes such as molasses offer cost-effective alternatives74. In this study H. japonica strain SST1 was able to grow in molasses containing medium, although EPS yield was halved compared to sucrose, and carbohydrate content was 4.7-fold lower. By contrast, maltose supported EPS yields and carbohydrate contents similar to glucose. These findings highlight the potential of using agro-industrial residues as inexpensive, sustainable carbon sources. The utilization of molasses thus positions H. japonica strain SST1 as a promising cost-effective EPS producer, suitable for circular bioeconomy approaches. However, further studies are needed to optimize EPS production using renewable substrates.

To evaluate the physicochemical and biological properties of archaeal EPS and their potential applications, structural characterization is essential. Archaeal EPSs are complex heteropolysaccharides composed of repeating sugar units, with glucose, galactose, and mannose being most common. Additionally, sulfate groups, glycoproteins and amino acids have been identified in EPSs in halophilic archaea19. In this study, the carbohydrate composition of the EPS synthesized by H. japonica strain SST1 was similar to that of other Haloarcula species, although the molar ratios of monomers differed from previously reported EPS31,71. Most microbial EPSs are heteropolymers containing two to five sugars71,75,76. EPSs with larger number of monomers have been described in species such as H. turkmenica and Halorubrum sp. (Table 5). In Haloarcula and Haloferax mannose, glucose and galactose are the dominant monomers.

EPS containing uronic acids are particularly valuable in the cosmetics industry, due to their hydrating properties76. The EPS from H. japonica strain SST1 contained significant levels of uronic acids (10.8% of EPS on sucrose medium). Protein content (4.3–10.3%) likely reflects extracellular enzyme synthesis.

Previous studies reported molecular masses of archaeal EPS ranging from 300 to 1000 kDa77. For example, H. turkmenica, produces two polymers of 801.7 kDa and 206.0 kDa26, while H. hispanica ATCC33960 secretes an acidic EPS of ~ 1.1 × 103 kDa71. H. mediterranei produces EPS > 100 kDa78 and the lowest reported molecular mass is ~ 5.05 kDa in Halorubrum sp. TBZ11229. The EPS of H. japonica strain SST1 exhibited a heterogeneous distribution, ranging from 10–100 × 103 Da, consistent with reported archaeal EPSs ranges. Broad size range of EPS indicates multiple polysaccharide fractions with distinct chain lengths and substitutions, reflecting structural heterogeneity typical of microbial EPS. High-molecular-weight fractions are generally linked to viscosity, gel formation, and matrix protection, whereas low-molecular-weight components contribute to signaling, adhesion, and molecular interactions18,20,22,77. Such size diversity may enhance the biological versatility of EPS, facilitating the adaptability of H. japonica to environmental conditions and interspecies interactions.

Most bacterial EPS are synthesized intracellularly and secreted extracellularly via well-characterized pathways, including the Wzx/Wzy-dependent pathway, ATP-binding cassette (ABC) transporter-dependent pathway, synthase-dependent pathway, and extracellular biosynthesis mediated by sucrase proteins19. Homopolysaccharides are generally assembled through Wzx/Wzy-dependent and ABC transporter-dependent pathways. EPS biosynthesis typically involves four steps: (1) activation of monosaccharides to sugar nucleotides in the cytoplasm; (2) assembly of repeating units on a lipid carrier by glycosyltransferases at the plasma membrane; (3) polymerization of repeating units at the periplasmic side; and (4) export of the polymer to the cell surface79,80,81,82. Although EPS biosynthesis has been genetically well characterized in bacteria such as those producing xanthan, levan and dextran, genetic data for halophilic archaea remain scarce83. Here, whole-genome sequencing and bioinformatics analysis were used to predict EPS biosynthesis in H. japonica strain SST1. KEGG analysis showed that the strain possesses the transporters and enzymes necessary for maltose, glucose and fructose utilization. The genome encodes the full set of enzymes required to convert these sugars into nucleotide sugar precursors for EPS biosynthesis (Fig. 7). Genes encoding glycosyltransferases, polymerases, branching enzymes, and enzymes responsible for sugar modification and substituent addition were identified.

Fig. 7
figure 7

Metabolic pathway map for EPS production in H. japonica strain SST1, illustrating the biosynthesis of activated nucleotide sugars, polymerization, and export, based on genome functional annotation. Enzyme identifications are provided by EC numbers. The enzyme marked in red (EC 5.3.1.8), and the sucrose transport system were not identified in the genome.

Comparative analysis of EPS-related gene clusters in H. japonica strain SST1, H. hispanica ATCC 33,96071, H. gibbonsii ATCC 33,95925, and H. mediterranei ATCC 33,50084 revealed conserved and divergent features (Figure S5). The highest protein identity was observed between SST1 Cluster 4 and H. hispanica Cluster 1, particularly in Tungstate ABC transporters and several EPS biosynthesis enzymes (e.g., UDP-glucose 4-epimerase, UDP-glucose 6-dehydrogenase, UTP-glucose-1-phosphate uridylyltransferase, sugar transferases). Cluster 4 also exhibited weaker similarities with H. gibbonsii Cluster 1 in EPS precursor biosynthesis enzymes. Cluster 3 displayed low similarity to H. gibbonsii Cluster 1 proteins, especially in polysaccharide biosynthesis proteins, flippases, and succinoglycan biosynthesis proteins. However, similarities were found with H. gibbonsii Cluster 2 in glycosyltransferases and precursor biosynthesis proteins. Additional overlap was observed between SST1 Cluster 2 and both H. gibbonsii Cluster 2 and H. mediterranei Cluster 1, particularly in ABC transporters and glycosyltransferases. These results underscore the key role of glycosyltransferases and ABC transporters in archaeal EPS biosynthesis and export.

Finally, analysis of C-type lectin interactions suggests that EPS from H. japonica strain SST1 may influence host immune response. C-type lectins expressed by immune cells, recognize self and non-self ligands and mediate diverse physiological processes85. C-type lectins typically recognize heteropolymeric polysaccharides containing sugar residues such as mannose, galactose, fucose, glucose, and GlcNAc. The EPS from strain SST1 contains mannose, galactose, and glucose, all of which are known ligands for several C-type lectins. In the presence of calcium and magnesium ions, binding of the EPS to DC-SIGN, Langerin, and MR was observed, indicating that these interactions are mediated by recognition of the corresponding sugar residues. In contrast, the C-type lectin MGL, which preferentially binds terminal GalNAc, a residue absent from SST1 EPS, showed no binding, underscoring the specificity of these interactions.

The presence of galactose has been linked to immunomodulatory activity86. EPS molecular mass has also been associated with immune effects, with lower molecular mass polysaccharides eliciting stronger responses87. The EPS produced by H. japonica strain SST1, composed of mannose, galactose, and glucose and exhibiting a molecular mass distribution of 10–100 × 103 Da, is consistent with these findings and suggests potential immunomodulatory properties.

Conclusions

In conclusion, this study reports the first isolation and characterization of EPS produced by H. japonica strain SST1 from subterranean salt deposits in Avan, Armenia. The findings expand the diversity of known EPS-producing halophilic archaea. The EPS yield of H. japonica strain SST1 (628.4 mg L⁻1) exceeded that of most previously studied Haloarcula species, suggesting its potential as a valuable biopolymer producer. This study also provides insights into the influence of different carbon sources on EPS production. Sucrose was identified as the optimal carbon source for EPS biosynthesis, while alternative sources, such as molasses, also supported EPS production, albeit at reduced levels. The ability of H. japonica strain SST1 to utilize inexpensive agro-industrial by-products highlights its potential for sustainable and cost-effective EPS production within a circular bioeconomy framework.

Chemical analysis showed that the EPS of strain SST1 closely resembles those of other halophilic archaea in carbohydrate composition, uronic acid content, and protein components. Genomic analysis of strain SST1 revealed key genes involved in EPS biosynthesis, including those encoding glycosyltransferases, polymerases, and sugar-modifying enzymes. Comparative genomics with other EPS-producing archaea revealed conserved biosynthetic pathways, while also uncovering distinct genomic features unique to strain SST1. These findings enrich our understanding of the genetic basis of EPS biosynthesis in halophilic archaea.

Overall, this study provides valuable insights into the EPS-producing potential of H. japonica strain SST1 and lays a solid foundation for future investigation into its physicochemical properties and biotechnological applications. The observed binding of the EPS to human C-type lectins suggests a previously unreported biomedical potential in modulating innate immune responses. Future research should explore the functional properties of this EPS and evaluate its applications across pharmaceuticals, food, and cosmetics, while also deepening our understanding of the metabolic and regulatory mechanism driving EPS biosynthesis in halophilic archaea.