The wastewater sample was collected from the Changqing oilfield in Shaanxi Province of China. A 10 ml sample was taken and introduced into a 250 ml conical flask containing 90 ml of sterile Luria–Bertani (LB) liquid medium consisting of (in per liter) 10 g of Na₂HPO₄, 5 g of yeast extract powder and 10 g of NaCl (pH 7.0), which was cultured at 30°C for 24 h under shaking condition at 120 rpm (Huang et al., 2010). The 10 ml of bacterial suspension was inoculated into 150 ml conicalflask containing 90 ml LB medium for one subculture. During cultivation, the medium gradually changed from clear to turbid and then to orange color, indicating the presence of bacteria in the culture. After three successive subcultures, the logarithmic-phase cultures were diluted to 10⁻⁸ by successive 10-fold dilution steps. Taking 100 μl of suspension with dilution from 10⁻⁴ to 10⁻⁸ was spread on LB agar plates and incubated at 30°C for 24 h. A single colony was selected and re-cultured and re-spread onto LB agar plates for purification. The isolation process was repeated three times to ensure the purity of the strain. The pure isolate obtained was named CQ-Y1.

Bacterial cells were collected from 1.5 ml of logarithmic-phase cultures by centrifugation at 11,100 × g for 1 min. Genomic DNA was isolated from the harvested cells using the DNA midi kit (Biomed, Beijing, China) according to the manufacturer’s instructions. PCR fragments were obtained by amplification of 16S rRNA gene using 27F (5′-AGAGTTTGATCCTGGCTCAG-3′) and 1492R (5′-GGTTACCTTGTTACGACTT-3′) primers. Each 25 μl PCR mixture contained 1.0 μl of template DNA (~50 ng), 1.0 μl of each primer (10 μM), 12.5 μl of 2× PCR master Mix (Biomed, Beijing, China) and 9.5 μl of ddH₂O. PCR was performed on the 2720 Thermal Cycler Instrument (Applied Biosystems, Foster City, United States) and the PCR followed these steps: 5 min at 94°C, followed by 30 cycles of 94°C for 30 s, 55°C for 45 s and 72°C for 1 min with final extension of 5 min at 72°C. The PCR products were quantified by electrophoresis on 1.0% agarose gel and purified by AxyPrep PCR Clean-up Kit (Axygen Inc., Union City, United States). The gel-purified PCR products were sequenced by BGI (Shenzhen, China). The ChromasPro 2.1.3 software (Technelysium Pty Ltd., Tewantin, Australia) and DNAMAN 6.0 software (Lynnon Biosoft, Quebec, Canada) were used to analyze and assemble 16S rRNA gene contiguous sequences. The nucleotide sequence of 16S rRNA gene was used to find maximum similarity to the bacterial strains in the NCBI database. MEGA11 software (Temple University, Philadelphia, PA, United States) and Clustal W (EMBL, Heidelberg, Germany) were used to perform multiple alignments and the phylogenetic tree construction by the neighbor-joining method (Koichiro et al., 2021). The ANI analysis was performed using JSpeciesWS (Ribocon GmbH, Bremen, Germany) with default settings.

Routine Gram staining reaction was performed and the cells were examined using bright-field microscopy at 100× magnification by using biological microscope Eclipse CI-L (Nikon, Tokyo, Japan). The cell morphology was examined using SEM (Hitachi S-4800, Ibaraki, Japan) and TEM (Hitachi H-7650, Ibaraki, Japan).

The Chrome Azurol S (CAS) two-layer plate method was used to directly detect siderophore according to the previous report (Chen, 1992). In brief, the agar containing the universal blue-colored ferric-CAS complex was used as the lower layer, and the LB medium containing 1.2% agar was used as the upper nutrient layer. The CQ-Y1 was delivered into the plate prepared in advance and incubated at 30°C for 3–4 days. The decolorization of CAS complex and the formation of yellow-orange halo around the upper nutrient layer colonies occurred, once the excreted siderophore penetrates from the upper trophic layer to the lower detection agar. The supernatant obtained from bacterial culture by centrifugation at 11,100 × g for 5 min was analyzed by CAS assay to quantify total siderophores (Machuca and Milagres, 2003; Wang et al., 2020). The relative content of the siderophore was calculated by the formula [(Ar − As)/Ar] × 100, where Ar and As are absorption of reference and sample, respectively.

The high-quality DNA was extracted by a DNA extraction kit (QIAGEN, Hilden, Germany). The degradation of the samples were detected by agarose gel electrophoresis, DNA purity was detected by Nanodrop, and DNA was accurately quantified by Qubit to detect DNA samples. After the DNA samples passed the quality inspection, the BluePippin automatic nucleic acid cutter was used to recover DNA with a specific fragment size, and then the DNA was damaged and repaired and end repaired. After purification of the magnetic beads, the NBD103 and NBD114 kits (Nanopore, Oxford, England) were used to connect barcode tags to the DNA ends, and the sequencing adapters included in the kits were connected. Finally, Qubit was used to accurately quantify the constructed DNA library. The complete genome sequence of CQ-Y1 was performed on PromethION (Oxford Nanopore Technology, Oxford, United Kingdom). After data quality control, the third generation data were assembled into contigs using the Flye version 2.8,² and the assembled data were rectified by using Racon version 1.4.13³ and Pilon version 1.23⁴ to get the final genome sequence.

Coding sequences (CDS) in the genome were predicted through Prodigal version 2.6.3.⁵ The prediction of tRNAs and rRNAs were conducted by tRNAscan-SE version 2.0⁶ and RNAmmer version 1.2.⁷ The gene annotation were carried by Interproscan,⁸ Tigrfams,⁹ Pfam¹⁰ and GO,¹¹ Kegg,¹² Refseq¹³ and COG¹⁴ databases. The best results with coverage >30% were retained as the annotation results. The circular map of the genome was obtained using Circos version 0.69.¹⁵

To investigate the high-temperature resistance, S. xiamenensis CQ-Y1 was cultured in LB medium whithout NaCl under different temperatures (35°C, 40°C, and 45°C), the incubation at 30°C was used as a control. NaCl tolerance was studied under 30°C using LB liquid medium containing various concentrations of NaCl (2, 4, 6, 8, 10 and 12%, w/v), the control experiment was also performed with the same medium without NaCl at 30°C. All cultures with 10% (v/v) inoculation were incubated on a shaker at 120 rpm for 60 h. To further determine the superimposed effect of high temperature and NaCl concentration on bacterial growth, CQ-Y1 was cultured in LB liquid medium plus different concentrations of NaCl (6, 8 and 10%, w/v) at 40°C and 45°C. The cultures of CQ-Y1 with the same media without NaCl at 40°C and 45°C were used as the control, respectively. The water loss by evaporation in conical bottles was replenished with distilled water by weight method. The bacterial growth for all cultures was monitored every 6 h by measuring OD₆₀₀ values spetrophtometrically. The ability of biofilm formation was assessed by crystal violet assay according to the previous study (De Jesus and Dedeles, 2020). The overnight culture of S. xiamenensis CQ-Y1 at different NaCl concentrations and temperatures were dilute 100 times by LB. Added 200 μl of culture to wells of sterile 96-well plates, and incubated at 30°C for 24 h. After the medium was gently sucked out, the bacteria were washed with PBS for 3 times to remove the cells, and then stained with 0.1% crystal violet for 20 min. The crystal violet dye negatively charged molecules and thus stains both bacteria and the surrounding biofilm matrix. The plates were washed exhaustively with distilled water and 200 μl of methanol was added to recover the absorbed stain. Finally, the optical density was measured at a wavelength of 600 nm with an enzyme-labeling instrument (Multiskan GO, Thermo Fisher Scientific, Inc., Rochester, United States). The ability of biofilm formation was assessed by crystal violet assay according to the previous study. The degree of biofilm formation were classified into strong (4 × Ac < A), moderate (2 × Ac < A ≤ 4 × Ac), weak (Ac < A ≤ 2 × Ac), or non-producer of biofilm (A ≤ Ac) based on the absorbance of culture sample (A) and LB medium (Ac) at 600 nm (Pourhajibagher et al., 2016).

Fe (III) reduction medium consisting of (in per liter) 12.8 g of Na₂HPO₄, 3 g of KH₂PO₄, 1 g of NH₄Cl, 2 g of sodium lactate, 2 g of yeast extract powder, and 6.48 g of synthetic Fe(OH)₃ (pH 7.0) was used to assay the Fe(III) reduction potential of CQ-Y1 (Tor and Lovley, 2001). The effect of various NaCl concentrations (0, 1, 2, 3 and 4%, w/v) or different temperatures (30°C, 35°C, 40°C, 45°C, and 50°C) on Fe(III) reduction were also monitored. The overnight culture of CQ-Y1 were centrifuged at 11,100 × g at 4°C for 1 min to harvest bacterial cells, washed twice with phosphate buffer (pH 8.0) and suspended again. A 0.1 ml of bacterial suspension was inoculated into 10 ml serum vial containing 8.9 ml sterile Fe(III) reduction medium. The operation under the same conditions without inoculation were used as control. Each serum vial was flushed with pure N₂, tightly sealed with stopper and aluminum crimp, and incubated on an Incubator for 60 h. The culture (1.0 ml) was sampled every 6 h and leached with 4 ml HCl (0.5 mol/L) for 24 h. Then, the Fe(II) concentration was measured by phenanthroline spectrophotometry (Han et al., 2018). The Fe(III) reduction ratio was calculated with the equation [C₂/C₁] × 100, where C₂, C₁ are the Fe(II) concentration in the reaction system and the Fe(III) concentration of the medium, respectively.

The morphology of precipitated substances at 0 and 120 h under different culture conditions of Fe(III) reduction expriments were observed by SEM. The samples were vacuum-dried in a desiccator for 7–10 days, and observed by SEM (Hitachi S-4800, Ibaraki, Japan) with a voltage of 5.0 kV. The XRD profiles of precipitates at 0, 72 and 120 h under different culture conditions of Fe(III) reduction expriments were examined by a multipurpose X-ray diffractometer (SmartLab SE, Rigaku Ltd., Japan) using Cu and Kα radiation (λ = 0.15406 nm). The scans were performed in the 10° ≤ 2θ ≤ 80° diffraction angular range at 25°C in steps of 10° with a recording time of 1 min for each step.

All bacterial culture and biochemical assays were performed with at least three independent repeats for each experiment. Data were presented as means ± standard deviations. The SPSS19.0 software (SPSS Inc., Chicago, IL, United States) was used for all data analysis and OriginPro 2018 software (OriginLab, Northampton, MA, United States) were used to draw all figures. The preference of the cell growth and Fe(III) reduction of CQ-Y1 adapted to high NaCl or high temperature was assessed by the regression using OriginPro 2018 software.

The complete genome sequence of S. xiamenensis CQ-Y1 was deposited in the GenBank database under the accession number CP092630.

The medium became turbid after incubation of LB medium inoculated with the oilfield wastewater sample at 30°C for 24 h. A large number of colonies appeared on the agar plates after 24 h of incubation at 30°C. Round shaped colonies with light brown color were selected and subcultured in LB liquid medium. A single colony was obtained by repeating the dilution-plate method above-mentioned. The colonies are 2.8 to 3.2 mm in diameter, circular, raised to slightly convex with smooth outline and smooth edges after 36 h incubation (Figure 1A). Gram staining was carried out (Kwon et al., 2019) and the bacterium was Gram-negative (Figure 1B). SEM and TEM showed that the bacterial cells were in rod shape with 0.35–0.46 μm in width and 1.12–1.93 μm in length, and appeared singly or rarely in pairs (Figures 1C,D). The TEM showed that CQ-Y1 have a cell envelope that exhibited the major ultrastructural features of Gram-negative bacteria with intact cell wall and full intracellular contents. The accessory structural of cilia and flagella were not detected in the TEM (Figure 1D).

The partial nucleotide sequence of 16S rRNA gene extracted from the whole genome of CQ-Y1 was used to conduct phylogenetic analysis with the sequences of bacterial strains in the GenBank database. The 16S rRNA gene sequence of CQ-Y1 shares a similarity of 98.96% to S. xiamenensis JCM^T 16212. The phylogenetic tree constructed by neighbor-joining method of MEGA software shows that CQ-Y1 belongs to Shewanella genus (Figure 1E). Additionally, the genomes were compared to confirm the phylogenetic result based on 16S rRNA gene alignment. CQ-Y1 shares the highest ANI value of 96.51% with S. xiamenensis NUITM-VS1 (Supplementary Table S1), further indicating that CQ-Y1 closely affiliated to S. xiamenensis because strains with ANI ≥ 95% are considered to belong to the same species (Lachance et al., 2020). It could be seen that the clear orange haloes occurred on the double-layer CAS agar plate after 36 h of incubation (Figure 1F), implying that the strain produced siderophores. Using the principle of the color change of CAS blue dye mediated by siderophores, the relative content of siderophores produced by CQ-Y1 was found to be 29.6% (Figure 1G).

The whole genome sequencing of CQ-Y1 was completed by Nanopore sequencing technology, and its sequencing depth distribution is shown in Supplementary Figure S1. The complete genome of CQ-Y1 consisted of a 4,710,887 bp in a single circular chromosome with 46.50% G + C content, harboring 4,022 predicted coding sequences (CDSs), 78 tRNAs and 21 rRNAs (9 for 16S rRNA, 6 for 23S rRNA, and 6 for 5S rRNA; Supplementary Table S1; Supplementary Figure S2). It is known that the genome size of S. xiamenensis strains is between 4.60 and 5.37 Mb with a G + C content ranging from 46.20% to 46.55% (Supplementary Table S1). The COGs analysis showed that the identified 3,959 encoding proteins could be assigned to 26 different categories (Supplementary Table S2). Among them, 157 were assigned to proteins with known functions, and the remaining 398 were regarded as hypothetical proteins. Of these, 10 proteins have no homologs found in other bacterial genomes, possibly suggesting they may be novel proteins.

It could be seen from Supplementary Table S2 that 328 (8.3%) and 274 (6.2%) genes were assigned to Signal transduction mechanisms (T) and Translation, ribosomal structure and biogenesis (J), respectively. It was also found that 263 genes were designated as the amino acid transport and metabolism (E, 6.6%), 238 genes involved in energy production and conversion (C, 6.0%) and 124 genes related to defense mechanisms (V, 3.1%), suggesting that S. xiamenensis CQ-Y1 might be able to adapt to extreme environments through metabolism regulation.

Shewanella xiamenensis CQ-Y1 was isolated from oilfield wastewater consisting of groundwater and drilling fluid, which had NaCl concentration of 2.65% (w/v). Therefore, CQ-Y1 might also evolved the potential of NaCl tolerance to deal with such extreme environments. Actually, the genome of S. xiamenensis CQ-Y1 harbors the genes encoding Na⁺/H⁺ antiporter, K⁺/Cl⁻ transporter and small molecules of intracellular compatibility which associated with NaCl tolerance, such as nhaB, nhaA, chaA, trkA, kdpD, proA, gltB, betT, and so on (Supplementary Table S3). A large number of thermotolerant microbes could be detected in the high temperature oilfield wastewater. Genes encoding proteins associated with high-temperature tolerance were also found on the genome of CQ-Y1, such as dnaK, grpE, aceE, lpdA, ackA, holC, priA, etc. (Supplementary Table S4). Many genes in the CQ-Y1 genome were identified to encode proteins involved in biofilm formation, such as fliH, flhG, hfq, cps, gumC, etc. (Supplementary Table S5). Environmental tolerance of microbes could be improved by biofilm formation. Additionally, We found genes encoding flavin compounds and Cytochrome c protein could involve in Fe(III) reduction, such as ribA, ribB, ribM, ribH, omcA, mtrC, mtrA, mtrB and cymA, on the genome of CQ-Y1 (Supplementary Table S6). These results indicated that S. xiamenensis CQ-Y1 might possess the similar characteristics.

In order to confirm the results of the genomic analysis mentioned above, the tolerance tests were conducted. The degree of NaCl tolerance was determined by the culture of S. xiamenensis at different NaCl concentrations (0, 2, 4, 6, 8, 10 and 12%, w/v). As shown in Figure 2A, 2% NaCl (w/v) appeared to have no effect on the growth of CQ-Y1 compared to the control group without any salt, which is consistent with the S. xiamenensis strain JCM^T 16212 (Huang et al., 2010). CQ-Y1 grows well within a wide NaCl range of 0 to 10% (w/v), although the lag phase was extended by about 6 h when NaCl concentration exceeded 4% (w/v; Figure 2A). The maximum OD₆₀₀ of 1.27 occurred at a 6% NaCl (w/v) concentration, indicating that NaCl within a specific concentration range (0–6%, w/v) could improve the CQ-Y1 growth (Figure 2A). However, the growth of CQ-Y1 was significantly inhibited (p < 0.01) when the NaCl concentration reached 12% (w/v). Crystal violet staining assay showed that CQ-Y1 exhibited a high degree of biofilm formation (Supplementary Figures S3A,B) and the biofilm formation ability increased continuously with NaCl concentration from 0 to 12% (w/v) (Supplementary Table S7). CQ-Y1 have exceptional NaCl tolerance when compared with other shewanella strains, which might be due to the presence of abundant genes in genome for NaCl resistance (Supplementary Table S8). The growth and metabolism of CQ-Y1 were not inhibited when the NaCl concentration increased to 2% (w/v).

At present, most studies on the temperature tolerance of Shewanella spp. only focus on its cold tolerance. Here, we tested the high-temperature tolerance of CQ-Y1 from the oilfield wastewater. The results indicated that the OD₆₀₀ of S. xiamenensis reached about 1.1 after 30 h incubation at 30°C, but showed no significant change with increasing temperature from 30°C to 40°C. However, the bacterial growth declined sharply once the temperature reached 45°C, and the maximum OD₆₀₀ of bacterial culture at 45°C reached only half of that at 30°C (Figure 2B). As shown in Supplementary Table S9, the biofilm formation ability was enhanced with the rise of temperature from 30°C to 45°C. Once the temperature was above 45°C, the cell growth was severely inhibited and the biofilm formation significantly weaken (p < 0.01; Supplementary Figures S3C,D).

The superimposed effect of high temperature and NaCl concentration on bacterial growth and biofilm formation were further examined. As shown in Figure 2C, the growth of S. xiamenensis CQ-Y1 was not significantly affected by the NaCl in the concentration range of 0–10% (w/v) at 40°C. However, the cell growth was severely suppressed by the increase of NaCl concentration at 45°C (Figure 2D). Because of the foregoing, the maximum NaCl concentration and temperature for bacterial growth were 10% and 45°C, respectively. These results showed that S. xiamenensis CQ-Y1 exhibted better NaCl and temperature tolerance than other Shewanella spp. (Supplementary Table S10).

Temperature and salinity are considered the major controlling factors in the distribution of most bacteria in natural environment (Wang and Gu, 2005). The relationship of bacterial growth with NaCl concentration and temperature were assessed by regression analysis, respectively. The results showed that bacterial growth were both significantly associated with NaCl concentration (r² = 0.923, p < 0.01; Supplementary Figure S4A) and temperature (r² = 0.954, p < 0.01; Supplementary Figure S4D), respectively. The correlation coefficient of the latter exceeded that of the former, and its values increased from 0.923 to 0.992 with temperature in the range of 30°C–45°C (Supplementary Figures S4A–C). This finding indicated that temperature displayed a better impact on bacteril growth of CQ-Y1 than NaCl concentration.

Shewanella spp. use FeCl₃, FeC₆H₅O₇ and Fe(OH)₃ as the electron acceptor, of which synthetic Fe(OH)₃ has a high specific surface area (Li et al., 2012). In addition, the suitable electron donors are critical for microbial Fe(III) reduction, and lactate was reported to be the dominant electron donor for Shewanella species (Hong and Gu, 2010; Wu et al., 2013). In order to assess the Fe(III) reduction by S. xiamenensis CQ-Y1, the Fe(III) reduction experiments were carried out using synthetic Fe(OH)₃ as substrate and sodium lactate as the electron donor. Figure 3 shows the dynamic variations of Fe(III) reduction at different temperatures (30°C, 35°C, 40°C, 45°C, and 50°C) and NaCl concentrations (0, 1, 2, 3 and 4%, w/v). Fe(II) concentration decreased with an increase of NaCl concenrations at 30°C (Figure 3A), while only trace amounts of Fe(II) were produced without inoculation (Figure 3B). Correspondingly, the reduction ratio of Fe(III) decreased significantly (p < 0.05), the minimum value of 36.93% were observed at 4% w/v NaCl (Figure 3C), while the values were <0.01% without inoculation (Figure 3D). Under NaCl-free conditions, the final concentrations of Fe(II) decreased from 0.45 to 0.32 as the temperature increased to 50°C (Figure 3E), but remained low content in the no-inoculation group (Figure 3F). The lowest and highest percentages of Fe(III) reduction (50.3% and 70.1%) were observed at 50°C and 30°C, respectively (Figure 3G), but the reduction ratios were below 0.01% in the control group (Figure 3H). The maximum Fe(III) reduction ratio of S. xiamenensis CQ-Y1 was similar to that of S. oneidensis (Supplementary Table S10; O'Loughlin et al., 2007).

In order to further understand the maximum potential of Fe(III) reduction under extreme environment, the Fe(III) reduction capacity was determined by the superposition of high NaCl concentration (0, 1, 2 and 3%, w/v) and temperature (40°C and 45°C). The final Fe(II) concentration decreased from 0.37 to 0.18 mg/L with increasing NaCl concentration at 40°C (Figure 3I) and the Fe(III) reduction ratios decreased to 27.99% at 3% NaCl (w/v); (Figure 3K). However, only trace amounts of Fe(II) were produced within 24 h when the NaCl concentration did not excess 2% (w/v) at 45°C, and the Fe(II) was barely detectable after further incubation (Figure 3I). Meanwhile, the Fe(III) reduction ratio of CQ-Y1 were significantly inhibited (p < 0.01; Figure 3K). Only trace amounts of Fe(II) were detected in Fe(III) reduction incubation without inoculation (Figure 3J) and the reduction ratios were below 0.01% (Figure 3L).

The precipitated substances changed from brown to gray green or white during Fe(III) reduction by S. xiamenensis CQ-Y1. The substances after 0 and 120 h were collected and mineralogically characterized using SEM. Under NaCl-free conditions at 30°C, it was observed that the substance presented smooth surfaces with small conglomerates at the starting of experiment (Figure 4A), which is the Fe(OH)₃ (Carrasco et al., 2021; Zhan et al., 2021). A large number of lenticular shape and plate-like structure, similar to Fe₃(PO₄)₂, were observed at 120 h (Figure 4B). In contrast, the formation of grainy surface were observed Figure 4C probably due to NH₄FePO₄ formation in the control group (An et al., 2011). The substance under high NaCl concentration (4%, w/v) condition changed from a smooth surface at the beginning (Figure 4D) to a spongy microstructure at 120 h, which was analog to the structure of FeCl₂ (Figure 4E; Andrés et al., 2013), while the control was consistent with the substance at 30°C under NaCl-free conditions (Figure 4F). Under NaCl-free condition at 40°C, the smooth form of Fe(OH)₃ were observed at 0 h (Figure 4G), and the large-scaled film appeared at 120 h, which might be the Fe(OH)₂ precipitates (Figure 4H; Ma et al., 2013). No significant changes were observed in the control group (Figure 4I). Additionally, we monitored the sbustance at 0 and 120 h during incuabatrion under the superimposed effect of high temperature and NaCl. The smooth substance (Figure 4J) tend to form a spongy microstructure similar to FeCl₂ (Figure 4K) at 120 h, while the control group did not change significantly (Figure 4L).

The mineralogical phases and structural properties of products collected at 0, 72 and 120 h under different culture conditions were analyzed with XRD. Fe(OH)₃ and NaCl peaks occurred at the beginning, the additional peaks of Fe(PO₃)₃ and Fe₃(PO₄)₂ appeared after 72 h, and Fe₃(PO₄)₂ with 2θ peaks 11.0, 13.0, 17.9, 19.2, 21.6, 22.9, 26.5, 27.6, 29.6 were mainly observed at 120 h under 30°C NaCl-free reduction condition (Figure 5A), Fe(III) analogs were mainly observed in the absence of inoculation (Figure 5B). Fe(III) reduction products under 30°C and 4% NaCl (w/v) were determined. It could be observed that the additional peaks of FePO₄ and Fe(H₂PO₃)₃ appeared after 72 h incubation, three wide 2θ peaks 31.5, 45.2, and 75.1 and several weak 2θ peaks 15.0, 35.3, 46.4, 50.6, and 73.2 were captured at 120 h, which were indexed to the phase of NaCl and FeCl₂ (Figure 5C). The results showed that the reduced products might contain excessive NaCl in the medium and a small amount of FeCl₂, while no Fe(II) compounds were observed in the uninoculated group (Figure 5D). The XRD pattern of the substance produced by the reduction of Fe(III) at 40°C was similar to that at 30°C, but additional peaks for Fe(OH)₂ were found (Figure 5E). Only phosphate analogs could be found when inoculated without bacteria (Figure 5F). The dynamics of the product formation under 40°C and 3% NaCl (w/v) was shown in Figure 5G, the characteristic peaks of 31.5, 45.2 and 75.1 were consistent with PDF No. 05-0628 of NaCl standard diffraction card after 120 h incubation. Additionally, the weak peaks of FePO₄, Fe(OH)₂ and FeCl₂ were captured (Figure 5G), which also did not change in the control (Figure 5H). After 120 h incubation, characteristic peaks of NaCl such as 31.5, 45.2 and 75.1 appeared.

To evaluate the relevance of Fe(III) reduction to NaCl concentration and temperature, a relationship of them was established by regression analysis. Results showed the Fe(III) reduction by CQ-Y1 were significantly negatively correlated to NaCl concentration (r² = 0.996, p < 0.01; Supplementary Figure S5A) and temperature (r² = 0.945, p < 0.01; Supplementary Figure S5D). It also could be seen that correlation coefficients of regression curves for Fe(III) reduction at 30°C, 40°C, and 45°C (Supplementary Figures S5A–C) kept almost constant with increasing temperature. These results indicated NaCl concentration conferd more effct on microbial Fe(III) reduction than temperature.

The characteristics of CQ-Y1 are very similar to the type strain S. xiamenensis JCM^T 16212 isolated from coastal sea sediment of Xiamen in China (Huang et al., 2010). Phylogenetic analysis result showed that CQ-Y1 belongs to Shewanella spp. It is worth noting that the ANI between S. xiamenensis CQ-Y1 and S. putrefaciens ATCC 8071 was found to be 96.45%. A similar finding was reported by Kim et al., in which the ANI values of different species within the same genus were 62%–100% (Kim et al., 2014). The reason might be the occurrence of horizontal gene transfer between species witin the genus of Shewanella, which requires further investigation (Cruz and Davies, 2000). As a strategy for importing and utilizing Fe(III), siderophore is widely formed by Shewanella species, but its relative content has not been reported (Trindade et al., 2019). The relative content of siderophore produced by CQ-Y1 was 29.6%, which is significantly lower than that of high siderophore-yielding strain Penicillium chrysogenum (73%; Chowdappa et al., 2020).

Some Shewanella species isolated from the ocean can thrive under 0%–10% NaCl (w/v) conditions (Kuribayashi et al., 2017; Song et al., 2022; Zhang et al., 2022). Simliarly, CQ-Y1 grows well within a wide NaCl range of 0 to 10% (w/v). However, the lag phase of CQ-Y1 was extended by about 6 h when NaCl concentration exceeded 4% (w/v). The extension of the lag phase might be due to the adverse effects of high osmotic stress, ionic toxicity and nutritional imbalance caused by excessive Na⁺ accumulation on microbial cells. Growth of CQ-Y1 could be improved at salt levels of no more than 6%. This might be attributed to the presence of the halophilic enzyme, which could prevent the denaturation and inactivation of bacterial proteins only at high NaCl concentrations (Mariana et al., 2012). Appropriate NaCl concentration could prompt the expression of some genes encoding Na⁺/H⁺ transporters, proline, betaine and other intracellular compatible small molecules to maintain osmotic balance (Qin et al., 2020). Additionally, NaCl within a certain concentration range could stimulate the EPS production and biofilm formation of S. xiamenensis CQ-Y1 (Boels et al., 2001). However, once NaCl concentration has reached a critical value or salinity exceeds the tolerance range of CQ-Y1, high concentrations of Na⁺ produce toxicity, leading to impaired metabolism and decreased cellular metabolic activity (Peng et al., 2021).

Although the cold-adapted Shewanella spp. are widely found, a few strains can grow at high temperatures (Tseng et al., 2018). The strong tolerance to high temperature stress makes CQ-Y1 stand out in the shewanella spp. due to the presence of a large number of high temperature tolerance genes in genome. Tolerance tests result showed that CQ-Y1 could survive at 30°C–45°C. The total lipid content and saturated fat have higher melting points, and the expression of related genes enhances the heat stress resistance of bacteria (Urbieta et al., 2015). Additionally, biofilm plays an important role in the whole process of microbial resistance to high-temperature stress (Prasad et al., 2019). The CQ-Y1 growth was severely inhibited when the temperature was above 45°C. High-temperature stress can cause the irreversible denaturation of bacterial proteins, which would lead to the disintegration of cell structure, resulting in bacterial growth inhibition (Murata et al., 2011). The growth of CQ-Y1 was not significantly affected by the NaCl in the concentration range of 0–10% (w/v) at 40°C. Heterologous stress pretreatment could induce higher resistance toward osmotic challenges. Shewanella putrefaciens grown at low temperatures can provide cross-protection against lethal NaCl treatment, and whether a similar mechanism can arise when CQ-Y1 produces mild heat shock remains to be confirmed (Leblanc et al., 2010).

Fe(III) reduction experiments results indicated that CQ-Y1 could still perform Fe(III) reduction under certain NaCl concentrations and/or temperatures. Bacteria tend to aggregate during freezing damage, cold shock and thermal stresses, and the rate of bacterial death could be moderated by higher bacterial densities to some extent (Rong et al., 2022). However, if high bacteria densities sustain for a long time, the organisms would die quickly caused by hypoxia. Similarly, the bacteria died rapidly when the organism was stressed enough to break the osmotic barrier (Konings et al., 2002). CQ-Y1 increased the survival rate by aggregating together, and continued to exert its Fe(III) reducing capacity under a certain stress. The low concentration of Fe(III) could promote the secretion of EPS in the biofilm, which is more conducive to microbial aggregation, external pressure resistance and Fe(III) reduction reaction (Hu et al., 2016). In addition, the chemical speciation governs Fe(III) bioavailability, changes in Fe(III) morphology during reduction might prevent the continuation of Fe(III) reduction (Zhang et al., 1995; Haas and Dichristina, 2002). However, trace amounts of Fe(II) could also be detected in Fe(III) reduction culture without inoculated. It has been reported that these Fe(II) might derived from the substance produced by chemical reactions and ion exchange in the solution, for instance, formation of NH₄FePO₄ (An et al., 2011).

Combining the SEM and XRD analysis results, we found that the increase in NaCl concentration was accompanied by the accumulation of chloride ions in solution and the Fe(II) generated by reduction tended to combine with chloride ions rather than phosphate, resulting in a predominance of FeCl₂. With temperature rises, there is a tendency for Fe(OH)₂ formation. Observation of the combined effect of high temperature and NaCl concentration on the reduction products revealed that their precipitation tended to form a spongy microstructure similar to FeCl₂, implying that Fe(II) tended to bind with the excess chloride ions in solution. In the control group, we found the amorphous organization predominated at 72 h, and Fe(III) tended to combine with phosphorous acid radicals to form FePO₄, Fe(H₂PO₃)₃, Fe₃(OH)₃(PO₄)₂ and other substances with the time increased to 120 h. Trace amounts of Fe(II) could be detected without inoculation during the determining of Fe(III) reduction ratio of CQ-Y1. However, no Fe(II) compounds were detected in the XRD and SEM results under the identical condition. This suggested that the trace amounts of Fe(II) did not influence the macroscopic product changes and had no contribute to the Fe(III) reduction reaction of CQ-Y1. Furthermore, the Fe(III) reduction ability of S. xiamenensis CQ-Y1 was weak under the superimposed effect of high temperature and NaCl concentration. Fe(III) could be reduced to Fe(II), and the Fe(II) combined with PO₄³⁻ to form a complex similar to Vivianite [Fe₃(PO₄)²·8H₂O], indicating the gray-green precipitates might be Fe₃(PO₄)₂ (John et al., 2001). It is generally believed that lower phosphate concentrations are conducive to green rust formation, and green rust is more easily converted into Fe₃(PO₄)₂·8H₂O at high phosphate concentrations (Li et al., 2012). Shewanella spp. can transform hematite to iron oxide hydroxide and iron phosphate (Zhao et al., 2020), and was able to reduce akaganeite to form magnetite (Roh et al., 2006).

Based on the genomic analysis and Fe(III) reduction performance, a hypothetical Fe(III) reduction mechanism of CQ-Y1 was proposed. As shown in Figure 6, CQ-Y1 might have two main electron transport pathways associated with Fe(III) reduction under favorable conditions. (1) The electrons transfer from the cytochromes around the cytoplasm to the terminal reductases on the cell surface (Ghasemi et al., 2020), and transfer into the quinone pool via the NADH carrier, the CymA protein on the inner membrane transfers electrons to the periplasm, Fcc₃ in the periplasmic space transfers electrons to a transmembrane protein complex of MtrA, MtrB and MtrC on the outer membrane, which in turn transfers electrons to the cell surface, and electrons were transferred directly to iron oxide through the contact of OmcA and MtrC on the cell surface (Sheng et al., 2016). (2) Bacteria transport intracellular FAD to periplasmic space, and then hydrolyze to FMN, which is released into the extracellular environment through pore-forming proteins in the outer membrane and acts as electron shuttle carrier to transport electrons (Grill et al., 2007). In addition, siderophores synthesized intracellularly by bacteria could be secreted to capture Fe(III) in the extracellular environment, enabling them to achieve specific chelation of Fe(III) even under low Fe(III) conditions (Nie et al., 2021).

However, under high temperature and high NaCl conditions, CQ-Y1 might highly express Na⁺/H⁺ antitransporters and K⁺/Cl⁻ transporters to achieve strict regulation of Na⁺ concentration, induce the expression of heat shock proteins to protect the cells, including GrpE, DnaK, DnaJ and other chaperones (Figure 6). At the same time, under such conditions, CQ-Y1 is easy to develop biofilm on iron oxides, which enable more accessible to cell metabolism (Zinicovscaia et al., 2020). The biofilms can confer protection by limiting the access of dissolved oxygen to the metal and releasing intracellular electrons to solid metallic structures (Zinicovscaia et al., 2020). In this process, the QS system can help bacteria adhere and secrete a large number of EPS, which specifically binds to metal ions, promoting Fe(III) reduction reaction (Rong et al., 2022). Once the temperature exceeds 45°C and the NaCl concentration exceeds 10% (w/v), all aspects of the biological function of the bacteria would be damaged (Fox-Powell and Cockell, 2018). The Fe(III) reduction capacity inevitably decreases when cells cannot form biofilms to sustain the nutrients and energy transport needed for normal growth.

Microbially induced corrosion is a complex problem causing many hazards. Shewanella spp. are regularly found on corroded metal structures. The corrosion effects of Shewanella, a typical iron reducing bacterial genus, on steel have been extensively studied, but the evidence for these effects remains controversial (Robert et al., 2016). It has been suggested that Shewanella could reduce insoluble Fe(III) to soluble Fe(II) and consume oxygen, thus, limiting the interaction between O₂ and the metal surface (Jo et al., 2018). On the other hand, Fe(III) respiration may contribute to the reductive dissolution of the Fe(III) (hydr)oxide protective layer, promoting the detstruction of the steel (Li et al., 2017). The research of biofilms in microbially induced corrosion is also crucial, microbial activity in biofilms may cause changes in chemical conditions. Biofilms may induce corrosion by forming different aeration and concentration cells on metal surfaces (Robert et al., 2016). This information can be used to develop preventive methods and monitoring tools for microbially induced corrosion.