All solutions used in this study were prepared using degassed Milli-Q water (18 MΩ· CM; Millipore) inside an N₂-filled glovebox. The degassed water was produced by boiling Milli-Q water under a nitrogen gas purge for 2 hours to remove oxygen. All solutions were either freshly prepared or stored overnight in the glovebox prior to use.

A Hg(II) stock solution (100 g L⁻¹) was prepared by dissolving mercuric nitrate monohydrate [Hg(NO₃)₂·xH₂O, CAS: 7783-34-8, Aladdin, Shanghai] in 12% nitric acid. Working solutions at concentrations of 0.01, 0.1, 1 and 10 g L⁻¹ were prepared from the stock solution.

Geobacter sulfurreducens PCA (Caccavo et al., 1994) was purchased from DSMZ and cultivated under an N₂ atmosphere at 30 °C and pH 6.8 (± 0.05) in “standard” growth medium as described by Schaefer and Morel (Schaefer and Morel, 2009), with selenium omitted to prevent interference in Hg L_III-edge EXAFS measurements. The medium (pH 6.8) contained 1 g L⁻¹ yeast extract, 40 mM sodium fumarate, 10 mM sodium acetate, 10 mM MOPS, 5 mM NH₄Cl, 1.3 mM KCl, 0.25 mM MgSO₄, 0.17 mM NaCl, 80 μM nitrilotriacetic acid, 50 μM NaH₂PO₄, 8.8 μM CaCl₂, 1 mg L⁻¹ trace metals which contained 30 μM MnCl₂, 4.2 μM CoCl₂, 3.6 μM FeSO₄, 3.5 μM ZnSO₄, 0.4 μM NiCl₂, 0.4 μM Na₂MoO₄, 0.04 μM CuSO₄, and 1 mg L⁻¹ resazurin. All media and buffers were prepared in acid-cleaned serum bottles, boiled under a continuous N₂ purge, and sterilized by autoclaving.

Cultures (1−1.5 L) were harvested during the exponential growth phase (OD₆₀₀: 0.5–0.8, corresponding to OD₆₆₀: 0.4–0.7; Supplementary Figure S1) by centrifugation at 5.000 g for 10 min at 4 °C. Cell density was determined using a hemocytometer (DB-180M Series Microscope). Cultures collected after 2 days (OD₆₀₀ ~0.5) were classified as middle-exponential phase, and those after 3 days (OD₆₀₀ ~0.8) as late-exponential phase.

To assess whether cytoplasmic thiol concentrations respond to nutrient levels, additional sodium fumarate or sodium sulfate was added to the growth medium to create a high-carbon (“High C”; 3× fumarate) and high-sulfate (“High S”; 10× sulfate) growth medium.

The subcellular fractionation workflow was performed using two different protocols, as illustrated in Figure 1. Protocol 1 was adapted from established methods for isolating protein subcellular fractions in G. sulfurreducens (Inoue et al., 2010; Jahan et al., 2019). Protocol 2 was a modified version of Protocol 1, optimized for spheroplast preparation, in which the outer membrane and spheroplast were separated earlier in the process (Wang et al., 2016; An et al., 2019).

Extracellular fractions were obtained by vacuum filtration (0.2 μM) of cell culture suspensions, with the filtrate defined as the extracellular component. Cell pellets were rinsed at least twice with degassed phosphate-buffered saline (PBS; pH 7.4; 0.14 M NaCl, 3 mM KCl, 10 mM Na₂HPO₄, and 2 mM KH₂PO₄) and designated as “whole-cell” samples. A portion of these samples was resuspended in degassed Milli-Q water and subjected to disruption with a cell disruptor (SCIENTZ-IID, Ningbo) to ensure complete lysis and release of all intracellular thiols, including those in the cytoplasm, periplasm, and membranes. The remaining cells were fractionated into subcellular compartments using the two protocols described below.

Protocol 1: The whole-cell pellet was initially resuspended in 10 mL of 250 mM Tris-HCl buffer, and 1 mL of 0.5 M EDTA was added. After 10 min of incubation at room temperature, 10 mL of 700 mM sucrose was added, and the mixture was incubated for 20 min at room temperature. Following this, 150 mg of lysozyme was added, and the suspension was incubated at 30 °C for 30 min. Osmotic shock was then induced by adding 20 mL of degassed Milli-Q water for 4 min. The mixture was centrifuged at 20.000 g for 10 min at 4 °C, and the supernatant was collected and designated as the periplasmic fraction. The resulting pellet, consisting of spheroplasts and outer membrane fragments, was washed at least twice with PBS buffer to remove residual lysozyme and sucrose. The washed spheroplasts were resuspended in 40 mL of 100 mM Tris-HCl buffer (pH 7.5) containing 5 units of DNase I and disrupted by sonication on ice. The lysate was centrifuged at 8.000 g for 30 min at 4 °C to remove cell debris. The resulting supernatant was further centrifuged at 20.000 g for 30 min at 4 °C, and the final supernatant was designated as the cytoplasmic fraction. The pellet was resuspended in 5 mL of 100 mM Tris-HCl buffer (pH 7.5) containing 2% (w/v) lauroylsarcosine, followed by the addition of another 5 mL of 100 mM Tris-HCl buffer. The suspension was then centrifuged at 125.000 g for 30 min at 4 °C (Beckman, Ultra-High Speed Centrifuge, L-100XP, Type 90 Ti rotor). The supernatant was collected as the inner membrane, and the pellet was reserved as the outer membrane.

Protocol 2: Cell pellets were resuspended in 4 mL of 250 mM Tris-HCl buffer (pH 7.5), followed by the addition of 0.4 mL of 0.5 M EDTA (pH 8.0) to chelate divalent cations within the peptidoglycan layer. After 1 minute of incubation at room temperature, 4 mL of 700 mM sucrose and 75 mg of lysozyme were added. The mixture was incubated at 30 °C for 6 hours. To induce osmotic shock and release periplasmic contents, 8 mL of degassed Milli-Q water was added. The suspension was then centrifuged at 20.000 g for 10 min at 4 °C. The resulting supernatant, containing the periplasmic and outer membrane components, was subjected to ultracentrifugation at 177.500 g for 2 h at 4 °C using a Beckman Type 90 Ti rotor. The pellet was designated as the outer membrane fraction, while the supernatant was collected as the periplasmic fraction. This approach differs from Protocol 1 in that the extended lysozyme treatment allows outer membrane material to partition into the supernatant rather than the pellet. This modification was adapted from spheroplast isolation protocols (An et al., 2019; Wang et al., 2016). The initial 20.000 g pellet, corresponding to spheroplasts, was washed twice with PBS (20.000 g, 4 °C, 10 min) to remove residual lysozyme and sucrose. The washed spheroplasts were resuspended in 15 mL of degassed Milli-Q water, freeze-thawed, and disrupted on ice using an ultrasonic homogenizer (SCIENTZ-IID) to lyse the inner membrane. The lysate was centrifuged at 8.000 g for 30 min at 4 °C to remove cell debris. The supernatant was then ultracentrifuged at 125.000 g for 30 min at 4 °C to separate the cytoplasmic fraction (supernatant) from the inner membrane (pellet).

To minimize sample oxidation during subcellular fractionation, all solutions were prepared using degassed Milli-Q water, and all operations were conducted inside a glovebox, except for centrifugation. Before centrifugation, the tubes were filled with an N₂ atmosphere. Samples were protected from light by covering the vessels with aluminum foil.

The subcellular dry mass was determined after freeze-drying, calculated as the difference between the measured weight and the theoretical dry weight of the added extractant when present. Total organic carbon (TOC) and total sulfur (TS) contents in subcellular fractions were measured using an elemental analyzer (Vario MACRO cube) with a relative standard deviation (RSD) of 0.5%. Total Hg concentrations were determined by combustion atomic absorption spectroscopy using a direct mercury analyzer (DMA-80, Milestone; detection limit =0.01 ng; RSD =6%). The reference material IAEA-456 (Hg = 77 ± 5 μg kg⁻¹) was used and yielded recoveries of 95%–105%.

The thiol group concentration in each individual sample (C_RS, μmol g⁻¹) was calculated using finally determined Hg concentrations (by DMA) and coordination numbers (CNs) obtained from EXAFS, as described in Equation 1.

where C_Hgfinal is the concentration of Hg (in μmol g⁻¹), and CN_S, CN_O/N, CNHg0, and CN_HgS are the EXAFS model fitted first-shell coordination numbers of chemical bonds Hg-S in Hg(RS)₂, Hg-O/N in Hg(RO/N)₂, Hg-Hg in liquid elemental Hg [Hg(l)0] and Hg-S in metacinnabar (β-HgS) respectively, with the denominators 2, 2, 6, and 4 representing the theoretical CNs of Hg-S, Hg-O/N, Hg-Hg, and Hg-S in these chemical forms, respectively. Importantly, only data for samples in which Hg-O/N (oxygen and nitrogen bonding cannot be separated by EXAFS) were observed in the first coordination shell were used in the calculation of C_RS. The underlying assumption is that only if the much weaker bonds to O/N atoms are detected, all thiol groups are saturated by Hg(II) and thus can be determined by Hg(II) titration (Skyllberg et al., 2006). In samples without detectable O/N involvement, the calculated thiol content is underestimated. The chemical forms Hg(l)0 and β-HgS are expected to form in parallel to Hg-thiol complexes. Only Hg species (Hg(RS)₂, Hg(RO/N)₂, Hg(l)0 and β-HgS) improving the merit-of-fit, ∑(model − experiment)²/∑experiment², by more than 10% were considered significant.

It should be noted that the reported thiol concentration (C_RS), derived from Equation 1, includes mass of the extraction reagents (for the cytoplasmic and periplasmic fractions) or the mass of growth medium (for the extracellular fraction). Thus, it does not represent the thiol concentration of the “pure” compartment alone. A more appropriate measure is the concentration of thiols per cell (μmol cell^-1), where the cell densities were estimated from OD₆₀₀. Total membrane thiols (inner + outer membranes) concentration per cell were calculated by subtracting the cytoplasmic and periplasmic thiols from that of whole-cell. Thiol concentrations express per gram of TOC (μmol g^-1 C) were also reported after correcting for reagent-derived TOC, although this correction introduces greater uncertainty. In contrast, thiol concentrations normalized on a per-cell basis (μmol cell^-1) are not affected by this issue and are therefore used throughout the study.

Uncertainties in calculated thiol group concentrations were propagated from an estimated 10% uncertainty in EXAFS coordination numbers (CNs), along with the standard deviations of Hg concentrations (Supplementary Table S1) and cell densities (Supplementary Figure S1). Statistical differences in thiol content between samples were assessed using a Z-test as each Hg EXAFS dataset originated from a single sample. The Z-test was calculated as:

where μ₁ and μ₂ are thiol concentrations; σ₁ and σ₂ are their respective propagated uncertainties. A difference was considered statistically significant at the 95% confidence level when z > 1.96. All calculations and statistical analyses were performed in R (Version 4.4.3) (R Core Team, 2025), with uncertainty propagation calculated by the first-order Taylor series method, as implemented in the R package errors (Ucar et al., 2018).

Dry mass, total organic carbon (TOC), and total sulfur (TS) were quantified in extracellular and subcellular fractions at the exponential growth phase (OD₆₀₀ ~0.5). Background contributions from media and extraction reagents were subtracted to correct the values. As shown in Supplementary Tables S2–S4, the two subcellular fractionation protocols produced consistent results, with average recoveries in the range of 80%–130%.

Averaged values from both protocols are summarized in Table 1. Specifically, the dry mass values for the extracellular, cytoplasmic, periplasmic, inner membrane, and outer membrane fractions were (6.0 ± 1.7) × 10⁻¹², (1.2 ± 0.2) × 10⁻¹², (5.0 ± 0.5) × 10⁻¹³, (2.4 ± 0.4) × 10⁻¹³ and (6.0 ± 2.0) × 10⁻¹⁴ g cell⁻¹, respectively. The whole-cell dry mass was (1.6 ± 0.2) × 10⁻¹² g cell^-1, approximately five times greater than that of E. coli grown in nutrient-poor medium (2.8 × 10⁻¹³ g cell^-1) (Dennis and Bremer, 1974; Neidhardt et al., 1990), likely due to bacterial species and growth condition differences. The whole-cell dry mass was 0.36 ± 0.05 g L⁻¹ of bacterial culture volume, closely matching the previously reported value of 0.42 g L⁻¹ for G.sulfurreducens (Engel et al., 2020).

TOC content in the extracellular, cytoplasmic, periplasmic, inner membrane, and outer membrane fraction was (1.9 ± 0.9) × 10⁻¹², (4.3 ± 0.7) × 10⁻¹³, (2.6 ± 1.4) × 10⁻¹³, (1.6 ± 0.2) × 10⁻¹³ and (2.0 ± 0.3) × 10⁻¹⁴ g cell^-1, respectively. A strong correlation was found between dry mass and TOC (Pearson's r = 0.99, p < 0.01; Supplementary Figure S2). Whole-cell TOC was (7.2 ± 1.0) × 10⁻¹³ g cell^-1, representing (46 ± 2)% of the total dry mass, while total nitrogen (TN) accounted for (12 ± 1)%. Both values are consistent with those reported for E.coli (50% TOC and 14% TN) (Neidhardt et al., 1990).

When the extracellular fraction was excluded, the cytoplasm comprised 59% of whole-cell dry mass and 48% of TOC. The periplasm, inner membrane, and outer membrane contributed 25%, 12%, and 3% of dry mass, and 30%, 18%, and 2% of TOC, respectively (Table 2). These distributions are consistent with reports of Gram-negative bacteria (e.g., E. coli, Salmonella typhimurium), with the cytoplasmic and periplasmic spaces occupying approximately 60%–90% and 20%–40% of total cell volume, respectively (Prochnow et al., 2019; Stock et al., 1977).

When both extracellular and cellular components were included (Table 2), the extracellular fraction accounted for 75% of the total dry mass and 69% of TOC in the cell culture, with relatively high uncertainties (~30%) due to the correction of background contributions from the growth medium The cytoplasm, periplasm, inner membrane, and outer membrane contributed 15%, 6%, 3%, and ~1% of dry mass, and 15%, 9%, 6%, and ~1% of TOC, respectively. The debris fraction, likely composed of residual membrane material and incompletely lysed cells, contributed less than 2% to both the dry mass and TOC of the whole-cell (Supplementary Tables S2, S3).

Total sulfur (TS) accounted for 0.85% of whole-cell dry mass (Supplementary Table S4), consistent with reported values of 0.5%–1% in bacterial dry mass (Kertesz, 2000). Sulfur was predominantly associated with the inner membrane (49%), followed by the cytoplasm (32%) (Table 2).

Sulfur K-edge XANES analysis (uncertainty estimated at ± 5% in the quantity of specified S forms) was performed on the extracellular, cytoplasmic, periplasmic, and whole-cell fractions. Because the energy separation between disulfide (RSSR), monosulfide (RSR), and thiol (RSH) is very small (< 1 eV), these species cannot be individually resolved in S K-edge XANES. However, their combined signal is readily distinguishable from other sulfur oxidation states and can be quantified with confidence. For this reason, the sum of these species is reported as reduced organic sulfur (Org-S_RED). In other words, S XANES provides only a semi-quantitative estimate of thiols through the Org-S_RED fraction.

Org-S_RED accounted for ~90% of total sulfur in the cytoplasmic, periplasmic, and whole-cell samples, but accounted only for ~30% in the extracellular fraction (Table 3 and Figure 2). These findings are consistent with our previous study employing physical cell lysis (i.e., freeze-thaw and ultrasonication), in which Org-S_RED accounted for approximately 90% of total sulfur in membrane and whole-cell samples, and around 25% in extracellular components of G. sulfurreducens (Song et al., 2020). This suggests that the subcellular fractionation method used in the present study does not alter sulfur speciation or subcellular thiols.

Subcellular thiol distributions in the middle exponential-phase of G. sulfurreducens are summarized in Figure 5 and Table 1, expressed as RSH per cell (μmol cell^-1). Extracellular thiols measured (5.7 ± 2.8) × 10⁻¹¹ μmol cell^-1. Whole-cell thiols totaled (1.35 ± 0.16) × 10⁻¹⁰ μmol cell^-1, with membranes constituting the dominant pool, (8.3 ± 1.9) × 10⁻¹¹ μmol cell^-1, partitioned between the inner membrane (7.1 ± 1.6) × 10⁻¹¹ μmol cell^-1 and outer membrane (1.2 ± 0.3) × 10⁻¹¹ μmol cell^-1. Concentrations in the cytoplasmic and periplasmic fractions were (1.5 ± 0.3) × 10⁻¹¹ and (3.6 ± 1.1) × 10⁻¹¹ μmol cell^-1, respectively.

The percentage distribution of thiols is shown in Table 2. When extracellular thiols are excluded, membranes account for 62 ± 16% of cellular thiols, with 53 ± 14% from the inner membrane and 9 ± 2% from the outer membrane; the cytoplasmic and periplasmic fractions contribute 11 ± 2% and 27 ± 8%, respectively. When extracellular thiols are included, they constitute 30 ± 16% of the total; membranes thiols contribute 44 ± 14%, partitioned into 37 ± 12% in the inner membrane and 6 ± 2% in the outer membrane; cytoplasmic and periplasmic thiols contribute 8 ± 2% and 19 ± 7%, respectively.

It should be noted that after 48 h of reaction between Hg(II) and subcellular fractions, substantial losses (0%–79%) of total Hg were observed across subcellular fractions, including extracellular, cytoplasmic, periplasmic, and whole-cell samples (Supplementary Table S1, Figure 6). This observation is consistent with previous studies reporting Hg losses of 10%–64% in membranes (a mixture of inner and outer membranes) (Song et al., 2020), 10%–52% in separately isolated inner or outer membranes (Gutensohn, 2023), and ~56% in cells incubation (Zhang et al., 2023). Mercury loss (%) was correlated with the Hg-to-thiol molar ratio in both extracellular and cytoplasmic fractions (data for the periplasmic and whole-cell fractions were insufficient for statistical analysis). In both cases, Hg loss increased with rising Hg-to-thiol ratios, reaching a maximum at approximately 10 for the extracellular and 4 for the cytoplasmic fractions, then declined slightly (Figure 6).

This loss can be attributed to the reduction of Hg(II) to Hg(g)0, as evidenced by the detection of liquid Hg(l)0 by Hg EXAFS (see Section 3.3.3). The observed Hg losses pattern agrees with previous studies demonstrating that Hg(II) reduction by natural organic matter and by G. sulfurreducens membranes is promoted at higher Hg-to-thiol ratios (Jiang et al., 2015; Miller et al., 2009; Song et al., 2020). This pattern can be explained by changes in both Hg speciation and the availability of electron donors. At low Hg concentrations, electron donors from cellular components are sufficient, and the extent of Hg(II) reduction is mainly governed by Hg-ligand speciation. As the Hg-to-thiol ratio increases, Hg shifts from Hg-thiolate complexes to more reducible Hg-RO/N species, e.g., Jiang et al. (2015), enhancing reduction. However, at very high Hg loadings, electron donors become limiting, constraining the overall reduction capacity. Consequently, further increases in Hg lead to a lower percentage of Hg loss.

The uptake of Hg(II) is highly dependent on its extracellular chemical speciation. Small neutral species such as HgCl2(aq)0, Hg(OH)2(aq)0, Hg(SH)2(aq)0 (Gutknecht, 1981; Mason et al., 1996; Benoit et al., 1999a,b, 2001; Hsu-Kim et al., 2013; Zhou et al., 2017), and HgS nano-particles (Deonarine and Hsu-Kim, 2009; Zhang et al., 2012; Guo et al., 2023), are thought to cross cell membranes via passive diffusion, whereas charged or larger complexes like Hg(Cys)₂ are proposed to be taken up via active transport (Schaefer and Morel, 2009; Schaefer et al., 2011).

In this study, the extracellular thiol concentration is approximately 6 × 10⁻¹¹ μmol cell^-1, representing about 30% of the total thiols in the bacteria culture. Although the TOC-normalized extracellular RSH density is relatively low (which is strongly dependent on growth conditions), extracellular thiols can still substantially influence Hg(II) speciation by the form of Hg(RS)₂ or mixed-ligand complexes such as Hg(RS)Cl_n(OH)_1−n in the extracellular medium Previous work in similar culture systems has shown that extracellular thiols, together with outer-membrane thiols, can modify Hg(II) speciation by promoting the formation of ternary complexes such as Hg(RS)(Mem-RS) (Adediran et al., 2019; Song et al., 2020), thereby influencing the complexes presented to the cell surface and potentially the uptake pathway.

The outer membrane has approximately 1 × 10⁻¹¹ μmol cell^-1 of thiols, corresponding to 6% of total thiols in the cell culture (or 10% of whole-cell thiols when the extracellular fraction is excluded). Although the absolute amount is relatively small, the outer membrane exhibited the highest thiol density (600 μmol g^-1 C), consistent with earlier work identifying this compartment as a major site of Hg(II) retention. However, its influence on Hg(II) uptake remains debated.

One study has reported that there was no significant change in Hg(II) uptake after blocking surface thiols with the bromine-containing probe qBBr (Thomas et al., 2020). In contrast, other work suggests outer membrane thiols may serve as a barrier restricting Hg(II) uptake (Dunham-Cheatham et al., 2014; Graham et al., 2012; Hu et al., 2013; Liu et al., 2016). An alternative view considers Hg(II)-RSH surface complexation as the initial step in the uptake process (An et al., 2019; Lin et al., 2014; Zhao et al., 2017). Adediran et al. (2019) proposed that Hg(II) may form ternary complexes involving membrane thiols and extracellular ligands. The letter suggestion find support in EXAFS measurements showing that only 5% of membrane thiols (inner + outer membranes) participate in bis-thiolate Hg(LMM-RS)₂ coordination, whereas 95% form ternary species (Song et al., 2020). Similarly, Gutensohn (2023) found that 42% of outer membrane thiols take part in ternary rather than bis-thiolate complexation. We propose that the contrasting effects of outer membrane thiols, as reported in the literature, may be explained by a parallel formation of bis-thiolate and ternary Hg(II) coordinations at the cell surface. Bis-thiolate complexes may impede uptake, whereas ternary complexes may promote transfer, offering a mechanistic basis for the observed variability in Hg(II) uptake behaviors.

The periplasm contains 3.6 × 10⁻¹¹ μmol cell^-1 of thiols, representing about 20% of total thiols in the full culture (or ~30% of whole-cell thiols when the extracellular fraction is excluded). Its thiol density (140 μmol g^-1 C) is lower than that of either inner or outer membrane, between which the periplasm is situated. During Hg(II) uptake, Hg(II) passes from the outer membrane into the periplasm before reaching the inner membrane, making the periplasm the aqueous intermediary between two solid-phase membranes. Its comparatively lower thiol density likely slows the transfer of Hg(II) to the inner membrane. This aligns with findings by Schaefer et al. (2014) and Wang et al. (2020) who showed that transport into the periplasm was the rate-limiting step, and higher Hg(II) uptake rates were observed in G. sulfurreducens spheroplasts (cells lacking outer membranes and retaining only the inner membrane) compared to intact whole-cells.

Further, the EXAFS indication on the formation of Hg-RSSR and β-HgS makes the periplasm particularly noteworthy. The presence of disulfides (RSSR) in the periplasm of G. sulfurreducens is expected, as the periplasm (in prokaryotes) is the primary site of disulfide-bond formation, a process essential for proper protein folding and stability, since proteins with reduced (free) cysteines are rapidly degraded (Bardwell et al., 1991; Sevier and Kaiser, 2002). In Gram-negative bacteria such as E. coli, the Dsb (disulfide-bond formation) proteins, including DsbA and DsbB, together with related periplasmic enzymes, catalyze thiol oxidation and generate RSSR species (Nakamoto and Bardwell, 2004; Ruiz et al., 2006). The small sulfide fraction detected by EXAFS (~13%, near the fitting uncertainty) may arise as a byproduct of disulfide-bond formation/degradation, and prior studies report sulfide production in G. sulfurreducens under specific metabolic conditions (e.g., when S⁰ serves as an electron acceptor) (Caccavo et al., 1994). Regardless of its origin, the periplasm provides a redox-active environment for Hg-S reactions that likely have an impact on Hg(II) internalization and transformation.

The inner membrane poses the highest thiol concentration, approximately 7 × 10⁻¹¹ μmol cell^-1, representing about 37% of total thiols in the full culture (or ~50% of whole-cell thiols when the extracellular fraction is excluded). The relatively high thiol density of the inner membrane (450 μmol g^-1 C) further supports its role in cellular Hg(II) internalization. This interpretation is consistent with observations that spheroplasts of G. sulfurreducens accumulate more Hg and exhibit higher uptake rates than intact cells (Schaefer et al., 2014; Wang et al., 2020). These studies suggest that Hg(II) complexes such as Hg(Cys)₂ or HgCl₂ passively diffuse across the outer membrane and periplasm, followed by an active transport across the inner membrane into the cytoplasm.

A plausible explanation may relate to differences in thiol-binding behavior. In the outer membrane, where ~58% of thiols are capable of forming bis-thiolate complexes Hg(RS)₂, Hg could become sequestered in this coordination state on outer membrane, thereby reducing opportunity for ligand exchange. In contrast, the inner membrane is predominantly (~95%) associated with ternary complex formation, Hg(RS)L, which might provide greater flexibility for ligand substitution and potentially support active transport into the cytoplasm. Nonetheless, this remains a hypothesis that requires experimental validation.

The cytoplasm contains 1.5 × 10⁻¹¹ μmol cell^-1 of thiols, accounting for about 8% of total thiols in the cell culture (or roughly 10% of whole-cell thiols when the extracellular fraction is excluded). The gradual increase in thiol abundance from the cytoplasm to the periplasm and then to the extracellular space is consistent with previous studies showing that LMM-RSH species are synthesized in the cytoplasm and subsequently secreted to the extracellular environment (Adediran et al., 2019). As cells progress through mid–exponential growth, cytoplasmic thiols decrease while extracellular thiols increase, as reported in Table 5. The cytoplasm exhibits a relatively low thiol density (40 μmol g^-1 C), similar to that of the extracellular fraction.

Within the cytoplasm, thiols may function as competitive ligands that acquire Hg(II) from inner membrane complexes via ligand-exchange, thereby enabling Hg(II) entry into the cytoplasm. The comparatively low cytoplasmic thiol abundance and density may modulate the rate of Hg(II) transfer into the cytosol, influencing its intracellular availability for subsequent processes such as Hg(II) methylation.

Previous studies have characterized the thermodynamic stability of Hg(II) complexes formed with LMM-RSH (Liem-Nguyen et al., 2017), natural organic matter associated thiols (Song et al., 2018), and membranes-associated thiols (Song et al., 2020), revealing comparable binding affinities (log K values). Building on this foundation, the detailed quantification of subcellular thiol distributions in this study provides a valuable framework for modeling Hg(II) speciation and bioavailability within one of model methylating bacteria, G. sulfurreducens. This information enables thermodynamic predictions of Hg(II) speciation and concentration across subcellular compartments, although the extent to which thermodynamic equilibrium, rather than kinetically constrained conditions, applies in living cells remains uncertain. Incorporation of subcellular thiol concentrations and their binding characteristics into biogeochemical models would enable model development to calculate Hg(II) uptake, intracellular partitioning, and methylation potential under varying environmental conditions.

Because many soft metals (e.g., Cd, Pb, Zn, Ag, and Cu) exhibit similar thiol-binding behavior, these reported results also provide a basis for understanding how other thiol-reactive metals may be partitioned and transformed within cells.