T. aciditolerans SY113 was cultivated under anaerobic conditions in TRM medium at 85 °C (Zeng et al., 2009; Li et al., 2021). To investigate the growth of SY113 strain, culture was grown in TRM medium supplemented with S⁰ until the stationary phase, with OD_600nm values ranging from 0.5 to 0.6. Then the culture was centrifuged at 2,000 × g and 4°C for 10 min to remove sulfur powder, transfer the supernatant liquid to a new tube and centrifuged again at 5,000 × g and 4°C for 15 min. The cell pellet was collected and washed three times with fresh anoxic TRM medium in an anaerobic chamber. The cells were then resuspended in anaerobic TRM medium to an OD_600nm of ~1.0. The resuspended cells were inoculated into TRM medium at 1% (v/v), then dispensed into 11 ml serum bottles in an anaerobic chamber with complete liquid filling (no headspace). For sulfur-supplemented culture conditions, elemental sulfur was added at a final concentration of 2 g/l. High-pressure cultivation at 27 MPa was performed in a stainless steel pressure vessel (Nantong Feiyu Petroleum Technology Development Co., Ltd., China) equipped with a pressure gauge and an automated water-pumping system for pressurization control. For colony selection, the TRM medium was solidified with 1% (w/v) Gelzan CM (Sigma-Aldrich, CAS 71010-52-1), and polysulfide (10 g Na₂S·9H₂O was dissolved in 15 ml deionized water, followed by addition of 3 g sublimed sulfur powder) was used as a substitute for elemental sulfur at a final concentration of 10 μl/ml. Under these conditions, visible colonies typically developed within 2–3 days at 85 °C.

For RNA extraction, ~3 × 10⁹ cells in the mid-exponential growth phase were harvested by centrifugation at 8,000 × g for 20 min. Total RNA was extracted with a TRI reagent-RNA/DNA/protein isolation kit (Molecular Research Center, Inc.) according to the manufacturer's instructions, yielding ~20–30 μg of RNA per sample. For RNA purification and cDNA synthesis, the PrimeScript^TM II 1st strand cDNA synthesis kit (TAKARA, Shiga, Japan) was used. The RT-PCR was performed as described previously (Li et al., 2018). The expression levels of target genes were normalized to the reference gene por (pyruvate/ketoisovalerate ferredoxin oxidoreductase subunit gamma, POR, FPV09_09760). Each experiment was performed in triplicate for each sample, and a mean value and standard deviation were calculated for the relative RNA expression levels.

The absolute copy number of target genes was quantified by correlating the cycle threshold (Ct) values with a standard curve. The genes mbsL, mbhL1, and mbhL2 from SY113 strain were cloned into the pMD™19-T vector (TAKARA, Shiga, Japan). The recombinant plasmids were transformed into E. coli DH5α competent cells and verified by DNA sequencing. Plasmids containing mbsL, mbhL1, and mbhL2 were extracted and quantified using a NanoDrop 2000/2000c spectrophotometer (Thermo Scientific, Shanghai, China). Real-time quantitative PCR (qPCR) was conducted on a QuantStudio 3 system (Applied Biosystems). Gradient-diluted plasmids were used as templates to generate Ct values for the construction of standard curves, which were subsequently used to calculate the absolute copy numbers of the target genes.

To overcome the limitations of available genetic markers, gene deletions were performed by replacing the target gene with the 3-hydroxy-3-methylglutaryl coenzyme A (HMG-CoA) reductase gene, which confers simvastatin resistance (Sim^R) as a positive selection marker. The Sim^R gene cassette was derived by fusing the promoter regions of glutamate dehydrogenase (PF1602) from P. furiosus DSM 3638 and HMG-CoA reductase (PYCH_01180) from P. yayanosii CH1 (Li et al., 2014). For counter-selection, an HHP-responsive toxin-antitoxin cassette (TAC) was utilized. The toxin gene is controlled by a HHP-inducible promoter, while the antitoxin gene is constitutively expressed under the P_gdh promoter to minimize basal toxin activity and eliminate background growth (Song et al., 2019).

Using the SY113 genome as a template, ~1 kb upstream and downstream regions of the mbsL, mbhL1, and surR genes were amplified using the primer pairs 04690-up-F/R, 03665-up-F/R, and 08560-up-F/R, respectively (Supplementary Table 1). The Sim^R cassette was amplified from the pUSW01 plasmid using primers P_gdh-F and HMG-CoA-R, and the HHP-TAC cassette was amplified using primers hhp-tac-F/R (Supplementary Table 1). The upstream and downstream fragments, along with the Sim^R and HHP-TAC cassette, were assembled into a linearized pUC18 vector containing 15 bp overlapping sequences using the ClonExpress^® Ultra One Step Cloning Kit (Vazyme). The resulting suicide plasmids-pUS04690, pUS03665, and pUS08560-were introduced into strain SY113 via CaCl₂-mediated transformation method (Li et al., 2014), with a transformation efficiency of ~3.3 × 10³ CFU per μg of plasmid DNA. Transformants were initially plated on solid TRM medium supplemented with 10 μM simvastatin and incubated at 85 °C for 2–3 days. Single colonies were then transferred to 5 ml of liquid TRM medium and incubated at 85 °C for 12–16 h. Take 500 μl of the culture and inoculate it into 50 ml of TRM medium. After incubating at 85 °C for 12–16 h, harvest the cells for genomic DNA extraction. Subsequently, verify the transformants by PCR amplification. Verified transformants were then inoculated into syringes and incubated in TRM under 27 MPa for 12 h to facilitate mutant selection via secondary homologous recombination. Following high-pressure incubation, cultures were plated onto solid TRM medium supplemented 10 μM simvastatin to isolate mutants resulting from secondary homologous recombination. The Hungate tube were incubated at 85 °C for 2–3 days. For each mutant strain, we typically selected 30–50 single colonies. These were subjected to liquid culture, followed by genomic DNA extraction, PCR amplification, and restriction enzyme digestion verification to obtain the gene knockout mutants. Overall, it takes ~3 weeks from vector construction to finally obtaining the mutant strain.

For quantitative determination of H₂S, 100 μl of culture from the stationary phase was mixed with 1 ml of H₂S absorption solution, and centrifuged at 12,000 × g for 10 min at 4 °C to remove the supernatant. Subsequently, 375 μl of N, N-dimethyl-p-phenylenediamine, and 375 μl ferric ammonium sulfate solutions were added, thoroughly mixed, and incubated at room temperature for 10 min. The absorbance was measured at 665 nm using a UV spectrophotometer. The specific H₂S concentration measurement followed the protocol of the H₂S assay kit (Solarbio, BC2050). The H₂S production yield was determined by normalizing to OD_600nm-measured biomass, using the formula: [H₂S] sample—[H₂S] blank control/OD_600nm, where [H₂S] represents the H₂S concentration measured by methylene blue method.

For quantitative determination of H₂, freshly prepared TRM medium was inoculated to an initial OD_600nm of ~0.01. Then, 5 ml of inoculated culture was dispensed into 15 ml Hungate anaerobic tubes with N₂ as the headspace gas. The tubes were incubated at 85 °C until the stationary phase was reached. Headspace gas samples were collected from the top of the sealed anaerobic tubes to quantify H₂ production. Gas analysis was performed using a gas chromatograph (Trace GC Ultra, Thermo Scientific) equipped with a thermal conductivity detector (TCD) set at 250 °C. Chromatographic separation was achieved using a CBP-01 column (6 ft [1.83 m] × 1/8 inch), with the injector and column oven temperatures set at 200 and 50 °C, respectively. Nitrogen served as the carrier gas at a flow rate of 20 ml/min, and the injection volume was 0.1 ml. The H₂ production yield was determined by normalizing to OD_600nm-measured biomass.

To investigate the effects of HHP on the growth and S⁰ metabolism, SY113 strain was cultured at 0.1 MPa and 27 MPa, both with and without S⁰ supplementation. As shown in Figure 1A1, in the absence of S⁰, the SY113 strain exhibited only limited growth under both 0.1 MPa and 27 MPa (corresponding to the strain's isolation depth). The addition of S⁰ significantly promoted the growth of the SY113 strain at atmospheric pressure. Notably, application of HHP further improved the growth. At 0.1 MPa, S⁰ supplementation increased the maximum biomass of the SY113 strain by ~6.5-fold (OD_600nm: 0.26 ± 0.01 with S⁰ vs. 0.04 ± 0.02 without S⁰). At 27 MPa, the biomass increased by about 11.6-fold (OD_600nm: 0.58 ± 0.04 with S⁰ vs. 0.05 ± 0.02 without S⁰). Notably, under HHP conditions with S⁰, not only enhanced the growth rate (0.32 ± 0.04 h⁻¹ at 0.1 MPa vs. 0.47 ± 0.06 h⁻¹ at 27 MPa), but also increased the H₂S concentrations from 10,062.03 ± 461.81 mmol/OD_600nm at 0.1 MPa to 13,118.21 ± 582.30 mmol/OD_600nm at 27 MPa (Figure 1B1). These results suggest that HHP promotes the growth of SY113 strain by enhancing H₂S production.

To further evaluate the impact of HHP and S⁰ on the gene expression of SY113 strain, we first analyzed the S⁰ reduction related genes in its genome. As shown in Figure 2, SY113 strain harbors a well conserved mbs gene cluster (FPV09_03610 to FPV09_03670), with gene composition and arrangement identical to those of shallow-water model strains. All corresponding subunits exhibited high similarity (average 85.6%) to those of the shallow-sea strain T. kodakarensis KOD1 (Supplementary Table 2). In addition, SY113 strain had two copies of mbh gene clusters, which were located next to each other on the chromosome. One copy, named as MBH1 (FPV09_08505 to FPV09_08570), showed a high sequence identity (average 73.3%) with shallow-sea strain KOD1, while the other copy, MBH2 (FPV09_08575 to FPV09_08645), possessed a relatively low sequence identity (average 44.5%) (Supplementary Table 2).

Using the catalytic subunit encoding genes mbsL (FPV09_03665), mbhL1 (FPV09_08560) and mbhL2 (FPV09_08630) as reference, we analyzed the effects of HHP and S⁰ on the transcriptional levels of these gene clusters. As shown in Table 1, the expression of mbsL was primarily enhanced by S⁰, while HHP showed no significant regulatory effect. The addition of S⁰ induced the expression of mbsL gene at both 0.1 MPa (0.1S/0.1, RQ: 1.92 ± 0.13) and 27 MPa (27S/27, RQ: 4.05 ± 0.25). However, when HHP was applied, no significant expression changes in mbsL gene were observed in either S⁰-supplemented (27S/0.1S, RQ: 1.18 ± 0.07) or S⁰-free (27/0.1, RQ: 0.56 ± 0.04) conditions. Notably, combination of S⁰ and HHP produced a synergistic effect, resulting in a 2.26-fold upregulation of mbsL gene expression compared to control conditions (27S/0.1, RQ: 2.26 ± 0.14). On the contrary, HHP markedly enhanced the mbhL1 gene expression irrespective of S⁰ availability (27/0.1, RQ: 2.39 ± 0.03 and 27S/0.1S, RQ: 3.00 ± 0.07). In contrast, S⁰ had no effect on mbhL1 gene expression (0.1S/0.1, RQ: 0.99 ± 0.02 and 27S/27, RQ: 1.24 ± 0.03). Strikingly, the combined application of S⁰ and HHP synergistically enhanced mbhL1 gene expression by ~3-fold (27S/0.1, RQ: 2.97 ± 0.03). For another copy of mbhL2 gene, neither S⁰ nor HHP treatment exerted any significant regulatory effects.

We further analyzed the transcriptional abundance of these genes under different conditions by means of absolute quantitative RT-PCR. As shown in Supplementary Figure 1, the transcript abundance of mbhL1 gene at 0.1 MPa was modestly elevated relative to mbsL gene (1.4-fold and 1.9-fold with or without S⁰, respectively). At 27 MPa, this difference became more pronounced, with mbhL1 gene expression exceeding mbsL gene by 3.2-fold and 7.1-fold under S⁰-supplemented and S⁰-free conditions, respectively. In contrast to both mbsL gene and mbhL1 gene, mbhL2 gene exhibited significantly low transcriptional abundance (10-fold lower) in all tested conditions. Collectively, these findings demonstrate distinct regulatory responses among the three gene clusters, MBS expression was predominantly induced by S⁰, and MBH1 was primarily activated by HHP, while MBH2 cluster remained transcriptionally inert under all tested conditions.

To further investigate the roles of MBS and MBH1 in HHP-enhanced growth, we first inactivated the mbsL gene (Supplementary Figure 2A). As illustrated in Figure 1A2, the disruption of the mbsL gene significantly impaired the growth of the SY113 strain under S⁰-supplemented conditions, with particularly pronounced inhibition observed at 27 MPa. At 0.1 MPa, the maximum biomass decreased from OD_600nm 0.26 to 0.19 and the growth rate decreased from 0.32 h⁻¹ to 0.18 h⁻¹. While at 27 MPa, the maximum biomass decreased by 5.8 times (OD_600nm: 0.58 ± 0.04 for wild-type vs. 0.10 ± 0.01 for ΔmbsL) and the growth rate decreased by 3.4 times (0.47 ± 0.06 h⁻¹ for wild-type vs. 0.14 ± 0.04 h⁻¹ for ΔmbsL). Thus, the strain's growth phenotype shifted from piezophilic to piezosensitive. Notably, under S⁰-supplemented conditions, the ΔmbsL mutant exhibited a pronounced reduction in H₂S production at both 0.1 MPa and 27 MPa compared to the wild-type strain. Specifically, H₂S yields decreased from 10,062.03 ± 461.81 mmol/OD_600nm (wild-type) to 1,685.89 ± 307.28 nmol/OD_600nm (ΔmbsL) at 0.1 MPa, and from 13,118.21 ± 582.30 mmol/OD_600nm to 3,790.13 ± 865.36 nmol/OD_600nm at 27 MPa, representing 76.5% and 68.2% reductions, respectively (Figures 1B1, B2).

Moreover, we further disrupted the mbhL1 gene to investigate the growth and metabolic profiles of the strain when only the MBS was functional (Supplementary Figure 2B). Surprisingly, compared to the wild-type strain, the ΔmbhL1 strain exhibited enhanced growth in the presence of S⁰ (Figure 1A3). At 0.1 MPa, the maximum biomass increased by 1.8 times (OD_600nm: 0.46 ± 0.02 for ΔmbhL1 vs. 0.26 ± 0.01 for wild-type) and the growth rate increased by 1.6 times (0.50 ± 0.05 h⁻¹ for ΔmbhL1 vs. 0.32 ± 0.04 h⁻¹ for wild-type). While at 27 MPa, the maximum biomass increased from 0.58 to 0.65 and the growth rate increased by 1.8 times (0.86 ± 0.1 h⁻¹ for ΔmbhL1 vs. 0.47 ± 0.06 h⁻¹ for wild-type). The ΔmbhL1 strain produced H₂S concentrations of 5,912.02 ± 1,775.23 mmol/OD_600nm and 19,676.19 ± 455.25 mmol/OD_600nm at 0.1 MPa and 27 MPa (Figure 1B3), respectively, which were comparable to those of the wild-type strain. These results indicate that MBS alone is sufficient for HHP-enhanced growth.

Although HHP stimulated mbhL1 gene expression, disruption of this gene unexpectedly enhanced the growth of SY113 strain. To resolve this discrepancy, we analyzed the expression patterns of mbsL gene and mbhL2 gene in the ΔmbhL1 mutant. As shown in Table 2, compared to the wild-type strain, mbhL1 gene disruption led to significantly elevated mbsL gene expression under all tested conditions, except for only a 1.65-fold increase at 0.1 MPa with S⁰ supplementation. Notably, mbhL2 gene expression was also markedly upregulated in the ΔmbhL1 mutant, particularly under HHP conditions (Table 2). These results suggest a regulatory link between mbhL1 gene and the expression of mbsL gene and mbhL2 gene. Furthermore, we examined mbhL1 gene and mbhL2 gene expression in the ΔmbsL mutant. Similarly, mbsL gene disruption resulted in significant upregulation of both mbhL1 gene and mbhL2 gene across all tested conditions (Table 2). Collectively, these findings demonstrate that the expression of MBS and MBH genes is interdependent.

In shallow-water model strains, sulfur promotes growth by altering the conformation of SurR, which subsequently enhances MBS expression while repressing MBH expression (Lipscomb et al., 2009; Yang et al., 2010). However, how HHP interacts with SurR in SY113 strain is yet to be elucidated. Analysis of the SY113 genome identified a SurR homologous gene (FPV09_04690) flanked by glutaredoxin and ferredoxin, a syntenic region characteristic of Thermococcales. To investigate the potential role of SurR in HHP response, we generated a ΔsurR mutant (Supplementary Figure 2C). Surprisingly, surR gene disruption did not impair the HHP-promoted growth. On the contrary, the ΔsurR mutant demonstrated superior growth performance under HHP compared to the wild-type strain (Figure 3A). At 0.1 MPa, the ΔsurR mutant achieved a maximum biomass of OD600 = 0.36 ± 0.03 with a growth rate of 0.38 ± 0.04 h⁻¹, whereas at 27 MPa it showed increased maximum biomass (OD_600nm = 0.73 ± 0.06) and significantly enhanced growth rate (1.13 ± 0.16 h⁻¹). Consistently, H₂S production under sulfur-supplemented conditions reached 7,561.41 ± 1,448.27 mmol/OD_600nm at 0.1 MPa and 15,350.00 ± 1,024.52 mmol/OD_600nm at 27 MPa (Figure 3B).

We further analyzed the transcriptional profiles of mbsL gene, mbhL1 gene and mbhL2 gene following surR gene disruption. As shown in Table 2, surR gene disruption resulted in upregulated expression of mbsL gene under all tested conditions, particularly under HHP. In contrast, mbhL2 gene exhibited significant expression downregulation in S⁰-supplemented conditions upon surR gene disruption. Disruption of the surR gene also resulted in a slight expression upregulation of mbhL1 gene at 0.1 MPa (RQ: 2.22 ± 0.03), while causing a slight downregulation under 27 MPa with S⁰ supplementation (RQ: 0.37 ± 0.01). No significant changes in expression were observed under the other two tested conditions. These findings demonstrate that SurR does not directly regulate either HHP-promoted growth or elemental sulfur reduction.