Abstract

   Anoxygenic phototrophic sulfur bacteria flourish in contemporary and
   ancient euxinic environments, driving the biogeochemical cycles of
   carbon and sulfur. However, it is unclear how these strict anaerobes
   meet their high demand for iron in iron-depleted environments. Here, we
   report that pyrite, a widespread and highly stable iron sulfide mineral
   in anoxic, low-temperature environments, can support the growth and
   metabolic activity of anoxygenic phototrophic sulfur bacteria by
   serving as the sole iron source under iron-depleted conditions.
   Transcriptomic and proteomic analyses revealed that pyrite addition
   substantially up-regulated genes and protein expression involved in
   photosynthesis, sulfur metabolism, and biosynthesis of organics. Anoxic
   microbial oxidation of pyritic sulfur and consequent destabilization of
   the pyrite structure were postulated to facilitate microbial iron
   acquisition. These findings advance our understanding of the survival
   strategies of anaerobes in iron-depleted environments and are important
   for revealing the previously underappreciated bioavailability of
   pyritic iron in anoxic environments and anoxic weathering of pyrite.
     __________________________________________________________________

   Anoxygenic phototrophic sulfur bacteria acquire iron from pyrite under
   iron-depleted conditions via oxidizing pyritic sulfur.

INTRODUCTION

   Oceanic euxinia (anoxic and sulfidic conditions) ubiquitously occurred
   in paleo-ocean, particularly during the Proterozoic ([36]1–[37]3).
   Because of the high toxicity of hydrogen sulfide (H[2]S), euxinia are
   believed to be an important reason for biotic crises such as the
   largest known Phanerozoic extinction—the end-Permian mass extinction
   (~252 million years ago) ([38]4–[39]6). Seasonal euxinia also
   frequently develop in modern environments such as the Black Sea,
   fjords, meromictic lakes, and the Benguela Upwelling System ([40]5,
   [41]7–[42]9). Although euxinia are well known for containing highly
   toxic H[2]S, with a concentration varying from several hundreds to
   thousands of micromolar, such hostile environments generally
   accommodate unique microbial communities such as anoxygenic
   phototrophic purple or green sulfur bacteria (PSB or GSB) ([43]5,
   [44]6). These bacteria can produce long-lasting organic biomarkers,
   such as carotenoid isorenieratene, which can be used for robust
   identification of photic zone euxinia in ancient environments ([45]4,
   [46]10, [47]11).

   Prosperous growth of anoxygenic phototrophic sulfur bacteria requires
   illumination, carbon source, and reduced sulfur compounds ([48]12). As
   research progresses regarding the reaction center of anoxygenic
   photosynthesis, the importance of trace metals, particularly iron, in
   the growth of anoxygenic phototrophic sulfur bacteria has been revealed
   recently ([49]12, [50]13). For instance, three key Fe[4]S[4] clusters
   were identified within the electron transfer chain of the
   photosynthetic reaction center in the representative strain of GSB
   Chlorobaculum tepidum ([51]12). During anoxygenic photosynthesis, light
   energy captured by chlorosome (i.e., the light-harvesting structure of
   C. tepidum) is transferred by Fenna-Matthews-Olson proteins to its
   membrane-embedded photosynthetic reaction center to initiate charge
   separation and electron transfer reactions ([52]12, [53]13). A
   similarly pivotal role of iron in the growth and carbon fixation of
   phytoplankton has also been reported ([54]14, [55]15). In addition,
   iron acts as a crucial cofactor for key enzymes involved in the sulfur
   oxidation processes of C. tepidum ([56]16). Therefore, fluctuations in
   iron concentrations over geological timescale are often closely linked
   to changes in ocean primary productivity and biogeochemical cycles of
   many elements such as carbon and sulfur ([57]15, [58]17).

   However, euxinic environments are iron depleted and usually
   characterized by a high degree of pyritization [e.g., reactive aqueous
   iron or iron (hydro)oxides react with free H[2]S to form fine-grained
   pyrite in the water column] ([59]18–[60]20). This creates an apparent
   paradox between iron scarcity and widespread distribution of
   GSB-derived biomarkers (i.e., requiring iron for GSB growth) in both
   the geological past and modern euxinic environments ([61]8, [62]10),
   prompting inquiries about the iron sources and the survival strategies
   of GSB in sulfidic settings. Given that pyrite is considered as the
   most abundant iron form in euxinic environments ([63]2, [64]20), it may
   be possible that pyrite-associated iron supports the growth of
   anoxygenic phototrophic sulfur bacteria, despite the usual belief that
   pyrite is stable under anoxic, low-temperature conditions ([65]21). It
   is long believed that pyrite is not bioavailable in the absence of
   oxygen or alternative electron acceptors such as nitrate and ferric
   iron ([66]22–[67]24). Nevertheless, such belief is challenged by recent
   lines of evidence of the utilization of sulfur within pyrite by
   anaerobic PSB and the utilization of both iron and sulfur within pyrite
   by anaerobic methanogenic archaea ([68]21, [69]25, [70]26). Although
   the underlying mechanism remains unclear, these studies suggest that
   anaerobic weathering of pyrite might be possible in the environment,
   especially under iron- or sulfur-limited conditions. Thus, we speculate
   that pyrite may serve as an efficient iron source to promote the growth
   of anoxygenic phototrophic sulfur bacteria.

   The primary objectives of this study, therefore, were to investigate
   whether anoxygenic phototrophic sulfur bacteria, using GSB C. tepidum
   as a representative bacterium, can use pyrite as the sole iron source
   under iron-depleted conditions, and if so, what the potential
   underlying mechanisms are. Understanding the metabolic flexibility and
   adaptation strategies of anoxygenic phototrophic sulfur bacteria in
   response to iron deficiency will shed light on microbial weathering of
   pyrite under anoxic condition, which is of great significance to the
   biogeochemical cycles of sulfur, carbon, and iron in euxinic
   environments ([71]27).

RESULTS

Inhibited cell growth and sulfur oxidation under iron-depleted conditions

   Cultivation experiments showed that C. tepidum grew steadily in the
   canonical medium regardless of cell transfer cycles, reaching an
   optical density at 750 nm (OD[750]) value of around 1.7 at the plateau
   stage (fig. S1). In contrast, the growth of C. tepidum in the
   iron-depleted medium was increasingly inhibited after multiple
   transfers, and the OD[750] decreased from 0.9 after the first transfer
   to 0.5 after the third transfer (fig. S1). Although the growth of C.
   tepidum was not completely suppressed under iron-depleted conditions,
   probably because of a trace amount of iron derived from chemical
   impurities, these results suggested that iron deficiency limited the
   growth of C. tepidum.

   When cells collected from the third transfer in the iron-depleted
   medium were inoculated into a modified medium amended with aqueous
   Fe^2+ solution (concentration: 0.2, 2, or 20 μM), the cell growth of C.
   tepidum was restored ([72]Fig. 1A). For instance, the addition of 2 and
   20 μM aqueous Fe^2+ notably enhanced the growth of C. tepidum, as
   indicated by the greater values of OD[750] ([73]Fig. 1A) and higher
   concentrations of total protein (fig. S2A) and bacteriochlorophyll c
   (BChl c) (fig. S2B), compared with that of the iron-depleted treatment,
   with a higher promoting effect with addition of 20 μM Fe^2+ than with 2
   μM Fe^2+ ([74]Fig. 1A and fig. S2). However, there was little effect on
   cell growth when 0.2 μM Fe^2+ was amended ([75]Fig. 1A and fig. S2),
   suggesting that an iron concentration higher than 0.2 μM was required
   to sustain the active metabolism of C. tepidum.

Fig. 1. Growth curves and time-course changes of sulfide concentration.

   [76]Fig. 1.
   [77]Open in a new tab

   (A) Growth curves of C. tepidum incubated under Fe-depleted conditions
   or with the addition of various concentrations of aqueous Fe^2+ (0.2,
   2, and 20 μM) and solid-phase iron (i.e., pyrite). (B) The
   corresponding time-course decreases of sulfide concentration.

   Accompanied by the inhibited cell growth of C. tepidum, microbial
   sulfur oxidation rate decreased obviously under iron-depleted
   conditions ([78]Fig. 1B). In general, C. tepidum oxidizes sulfide to
   extracellular elemental sulfur globules, and subsequent oxidation of
   elemental sulfur and thiosulfate only occurs when sulfide is depleted
   ([79]28–[80]30). Such sequential sulfur oxidation was observed in this
   study ([81]Fig. 1B and fig. S3). For instance, sulfide concentration
   decreased slowly in the treatments with no or 0.2 μM Fe^2+ ([82]Fig.
   1B), concomitant with a gradual accumulation of elemental sulfur (fig.
   S3A). Neither consumption of thiosulfate nor production of sulfate was
   observed in these two treatments (fig. S3, B and C) because of the
   remaining sulfide in the medium ([83]Fig. 1B). In comparison, active
   sulfur oxidation proceeded in the presence of 2 or 20 μM Fe^2+ as
   indicated by complete consumption of sulfide after 30 and 36 hours of
   inoculation, respectively ([84]Fig. 1B). Once sulfide was depleted,
   elemental sulfur (produced from sulfide oxidation) and initially added
   thiosulfate began to decrease (fig. S3, A and B), accompanied by
   sulfate production (fig. S3C). The calculated electron transfer rate
   within the logarithmic growth phase (i.e., 12 to 36 hours) increased
   from about 0.04 mM/hour under iron-depleted conditions to 0.45 mM/hour
   in the presence of 2 μM Fe^2+ and further to 0.77 mM/hour with 20 μM
   Fe^2+ addition (fig. S4).

Enhanced cell growth and sulfur oxidation with pyrite as the sole iron source

   Results of both OD[750] and total protein concentration showed that the
   addition of pyrite (1 g/liter) stimulated the growth of C. tepidum
   irrespective of particle size of pyrite (fig. S5A), with an enhancement
   effect even slightly better than that with 20 μM aqueous Fe^2+ addition
   ([85]Fig. 1A and fig. S2). This slightly higher promotion of pyrite
   suggests that in addition to serving as an iron source, pyrite might
   stimulate the growth of C. tepidum by providing photoelectrons due to
   its semiconductive property. Such extracellular photoelectrons have
   been shown to be used by anoxygenic phototrophs (e.g., PSB
   Rhodopseudomonas palustris) in CO[2] fixation to biomass ([86]31).
   However, the promoting effect did not occur when pyrite was replaced
   with similarly sized quartz (fig. S5A), thus excluding the possibility
   that the provision of physical attachment sites by solid mineral was
   responsible for the enhanced cell growth. A more likely reason was the
   lack of bioessential elements such as iron in quartz ([87]32).

   Similar to the growth pattern, pyrite accelerated sulfur oxidation,
   while quartz did not ([88]Fig. 1B and fig. S5B). In the pyrite
   treatment, sulfide was completely consumed in approximately 30 hours
   ([89]Fig. 1B), in conjunction with a rapid accumulation of elemental
   sulfur, which, along with thiosulfate, was subsequently oxidized to
   sulfate (fig. S3). In contrast, about 25% of the initially added
   sulfide remained by the end of the reaction (72 hours) in the treatment
   with quartz (fig. S5B), and no sulfate production was detected. This
   pattern was similar to the iron-depleted treatment (fig. S5). Moreover,
   the electron transfer rates during the logarithmic growth period
   increased from 0.04 mM/hour under iron-depleted conditions to 0.81
   mM/hour in the presence of pyrite (fig. S4), indicating that pyrite
   served as the sole iron source to sustain active microbial sulfur
   oxidation.

   Cultivation of C. tepidum in pyrite leachate (i.e., soaking pyrite in
   the medium for 7 days before inoculation) showed similar growth and
   sulfide-oxidation kinetics to those in the iron-depleted medium (fig.
   S6). This result was expected because pyrite is stable under euxinic
   conditions ([90]20) and led to a negligible amount of iron release into
   the leachate (i.e., ~0.10 μM), which was not sufficient to meet the
   cellular iron demand of C. tepidum ([91]Fig. 1). Thus, abiotic
   dissolution of pyrite could not account for the growth promotion of C.
   tepidum. In addition, when pyrite was sequestered in a dialysis bag
   with a molecular weight cutoff of 500 Da, the accelerating effects on
   both growth and sulfur-oxidation process of C. tepidum were still
   observed but were much weaker than those without a dialysis bag (fig.
   S7). These results indicated that either direct contact with pyrite or
   high–molecular weight (>500 Da) extracellular proteins and metabolites
   were essential for the microbial acquisition of iron from pyrite.

Cellular iron uptake during cell growth

   In the iron-depleted and 0.2 μM Fe^2+ treatments, both aqueous iron
   ([92]Fig. 2A) and cell-associated iron ([93]Fig. 2B) were hardly
   detectable over the course of the experiments. In contrast, aqueous
   iron concentration in the 20 μM Fe^2+ treatment declined gradually
   ([94]Fig. 2A), while the cell-associated iron concentration increased
   over time and reached a plateau after 36 hours ([95]Fig. 2B), in
   accordance with the growth profile of C. tepidum ([96]Fig. 1A),
   suggesting substantial cellular uptake of iron during the growth of C.
   tepidum. In the 2 μM Fe^2+ treatment, aqueous iron in the medium
   declined near the beginning ([97]Fig. 2A), and a lower accumulation of
   cell-associated iron was observed ([98]Fig. 2B), in alignment with the
   weaker growth enhancement of C. tepidum in this treatment ([99]Fig.
   1A). Under the same Fe concentration, the biomass-normalized
   cell-associated iron concentration did not change obviously over time
   for all treatments ([100]Fig. 2C), which might reflect the rapid
   metabolic response of C. tepidum to iron availability and a dependence
   on the bacterial physiological needs ([101]33). The lower concentration
   of cell-associated iron per unit OD in the 2 μM Fe^2+ treatment than
   that in the 20 μM Fe^2+ treatment ([102]Fig. 2C) suggested that the
   optimal iron concentration required by C. tepidum was greater than 2
   μM.

Fig. 2. Fe concentrations.

   [103]Fig. 2.
   [104]Open in a new tab

   Time-course concentration changes of iron (A) in culture supernatant,
   (B) in association with cells, and (C) in cell-associated iron per unit
   OD under Fe-depleted conditions or with the addition of various
   concentrations of aqueous Fe^2+ (0.2, 2, and 20 μM) and solid-phase
   iron (i.e., pyrite).

   In the pyrite treatment, aqueous iron in the medium was hardly
   detectable over the whole reaction period ([105]Fig. 2A), due to the
   high stability of pyrite in anoxic environments ([106]34). However, the
   cell-associated iron concentration increased substantially, even
   slightly higher than that in the 20 μM Fe^2+ treatment by 72 hours
   ([107]Fig. 2B). Similar to the aqueous Fe^2+ treatments, the
   biomass-normalized cell-associated Fe concentration did not change over
   time ([108]Fig. 2C). Dialysis experiments showed that both absolute and
   biomass-normalized concentrations of cell-associated iron were
   substantially diminished when pyrite particles and cells were
   physically separated (fig. S8). Overall, these data demonstrated that
   cellular iron uptake by C. tepidum did occur when pyrite was used as
   the sole, solid-phase iron source, which was responsible for the
   enhanced cell growth and sulfur oxidation ([109]Fig. 1).

Pyrite surface modification by C. tepidum

   Scanning electron microscopy (SEM) observation displayed a close
   physical association between C. tepidum cells and pyrite particles
   ([110]Fig. 3), verifying the importance of direct contact for microbial
   iron acquisition (figs. S7 and S8). Although x-ray diffraction (XRD)
   analysis showed no bulk mineralogical change of pyrite or the formation
   of secondary minerals (fig. S9), the variations in the relative peak
   intensities between the untreated and treated pyrite suggested that C.
   tepidum may have modified the surface structure of pyrite to facilitate
   iron acquisition. Meanwhile, time-of-flight secondary ion mass
   spectrometry (TOF-SIMS) analysis revealed chemical modification (i.e.,
   elemental composition changes) of pyrite surface after its interaction
   with C. tepidum under iron-depleted conditions ([111]Fig. 4).
   Specifically, principal components analysis (PCA) showed a clear
   separation between untreated and microbially treated pyrite along the
   PC1 dimension in both positive and negative ion spectra, with
   quintuplicate sites from the same sample clustering together (fig.
   S10). Further identification showed an increase in several biologically
   relevant organic molecular fragments after the biotic treatment [e.g.,
   mass/charge ratio (m/z) 63 (CH[5]NO[2]^+), 149 (C[9]H[9]O[2]^+), and 72
   (C[2]H[2]NO[2]^+)] ([112]Fig. 4 and table S1). The last one was
   previously identified on the surface of extracellular S^0 globules
   generated from sulfur oxidation by C. tepidum ([113]35). In contrast,
   peak intensities at m/z 56 (Fe^+) and 32 (S^−) were decreased
   ([114]Fig. 4).

Fig. 3. SEM images and elemental maps.

   [115]Fig. 3.
   [116]Open in a new tab

   (A to C) SEM images indicative of a close physical association between
   C. tepidum and pyrite; (B) and (C) are enlargements of the yellow and
   blue rectangular areas in (A), respectively. (D to F) The elemental
   maps of C, Fe, and S for the area are shown in (A).

Fig. 4. TOF-SIMS spectra.

   [117]Fig. 4.
   [118]Open in a new tab

   Chemical modification of pyrite surface after incubation with C.
   tepidum for 72 hours as indicated by the differences in normalized
   intensities (Δ normalized intensity) between treated and untreated
   pyrite samples as a function of m/z in both (A) positive and (B)
   negative ion spectra of TOF-SIMS. Positive “Δ normalized intensity”
   values indicate an increase in intensity after treatment with C.
   tepidum, while negative values indicate a decrease.

   X-ray photoelectron spectroscopy (XPS) analysis showed that iron on the
   surface of pyrite was still primarily in the form of ferrous iron
   [Fe(II)] after incubation with C. tepidum for 72 hours (i.e., reaction
   endpoint) (fig. S11), which aligned with previous findings that C.
   tepidum cannot use Fe(II) as an electron donor ([119]36). However, the
   valence state of sulfur on the surface of pyrite changed after
   incubation with C. tepidum for 24 hours, as indicated by the detection
   of SO[4]^2− ([120]Fig. 5). At this time point, sulfide had not been
   fully consumed ([121]Fig. 1B), and, thus, production of sulfate from
   oxidation of elemental sulfur or thiosulfate could not have occurred
   (fig. S3). Therefore, the presence of SO[4]^2− on the pyrite surface
   was attributed to the direct oxidation of sulfur within pyrite by C.
   tepidum ([122]26).

Fig. 5. XPS spectra.

   [123]Fig. 5.
   [124]Open in a new tab

   The S 2p XPS spectra of (A) untreated and (B) treated (incubated with
   C. tepidum for 24 hours under the iron-depleted condition) pyrite. This
   time point (i.e., 24 hours) was selected to avoid the interference from
   the oxidized sulfur compounds generated from microbial oxidation of
   sulfide and thiosulfate.

Transcriptomic response of C. tepidum to iron scarcity

   To reveal microbial coping strategies to iron scarcity, transcriptomic
   analysis was performed and compared across three treatments:
   iron-depleted treatment, aqueous Fe^2+ amendment (20 μM), and pyrite
   addition. The genome of C. tepidum is 2,154,946 base pairs in size,
   encoding about 2288 genes ([125]37). Results of transcriptomics showed
   that approximately 2103 genes were detected across the three
   treatments. PCA displayed a clear separation among the three treatments
   along the PC1 and PC2 dimensions, indicating treatment-specific gene
   expressions (fig. S12). Following normalization and statistical
   analysis using the Student’s t test, a comparative transcriptomic
   analysis was conducted with thresholds setting at log[2] fold change
   (log[2]FC) > 1 or log[2]FC < −1 to identify up-regulation and
   down-regulation of genes, respectively. P < 0.05 was taken as
   statistically significant. Overall, the addition of pyrite resulted in
   an up-regulation of 729 genes and a down-regulation of 446 genes
   compared with the iron-depleted treatment (fig. S13A). Furthermore,
   relative to the aqueous Fe^2+ treatment, the addition of pyrite
   resulted in an up-regulation of 549 genes and a down-regulation of 215
   genes (fig. S13B). These results indicated that the presence of pyrite
   substantially affected the transcriptomic profile of C. tepidum.

   The Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment
   analysis revealed that sulfur metabolism pathways were substantially
   up-regulated in the presence of pyrite relative to the iron-depleted
   treatment ([126]Fig. 6A), consistent with the enhanced sulfur oxidation
   in this treatment ([127]Fig. 1B). In addition, pyrite addition resulted
   in the up-regulation of pathways related to vitamin metabolism, such as
   biosynthesis of folate (vitamin B9) and thiamine (B1) ([128]Fig. 6A).
   Folate is critical for DNA and RNA synthesis and cell proliferation
   ([129]38, [130]39), while thiamine diphosphate (i.e., biologically
   active form of thiamine) serves as a crucial cofactor in the
   tricarboxylic acid (Krebs) cycle ([131]40, [132]41). Furthermore, the
   pyrite treatment enriched the pathway involved in lipopolysaccharide
   biosynthesis ([133]Fig. 6A). Pathways related to the biosynthesis of
   amino acids (e.g., phenylalanine, tyrosine, tryptophan, arginine, and
   proline) were also enriched in the pyrite treatment when compared to
   either the iron-depleted ([134]Fig. 6A) or aqueous Fe^2+ treatment
   ([135]Fig. 6B). These molecular insights were well in line with the
   extensive detection of organics on pyrite surface ([136]Fig. 4).

Fig. 6. KEGG pathway enrichment.

   [137]Fig. 6.
   [138]Open in a new tab

   Comparison of KEGG pathway enrichment of up-regulated gene clusters (A)
   pyrite treatment versus iron-depleted treatment and (B) pyrite
   treatment versus aqueous Fe^2+ treatment (20 μM). Dots represent
   pathway enrichment with color coding: Orange indicates high enrichment,
   and green indicates low enrichment. The size of the dots represents the
   gene number (count). Rich factor refers to the ratio of the
   up-regulated genes to all genes within a certain pathway. The larger
   value of a rich factor represents the higher level of enrichment.

   Compared to the treatment of aqueous Fe^2+, the riboflavin (B2)
   biosynthesis pathway was enriched in the pyrite treatment ([139]Fig.
   6B). Riboflavin is a key component of the coenzymes flavin adenine
   dinucleotide and flavin mononucleotide, which is essential for energy
   production and electron transfer ([140]42–[141]44). In addition, two
   distinctive pathways related to nucleotide metabolism (i.e., pyrimidine
   metabolism and nucleotide excision repair) were also up-regulated in
   the pyrite treatment ([142]Fig. 6B). This phenomenon suggested an
   increased demand for DNA replication and repair, probably due to DNA
   damage caused by the cytotoxic reactive oxygen species produced from
   pyrite ([143]45, [144]46).

   From the view of gene expression, cytochrome-related genes such as
   CT1734 (cytochrome c), CT0073 (membrane-bound cytochrome c-[555]), cydA
   (cytochrome d ubiquinol oxidase), and CT0188 (c-type cytochrome) were
   among the most up-regulated genes in the pyrite treatment when compared
   with either the iron-depleted or aqueous Fe^2+ treatment (fig. S14).
   The up-regulation factor could reach up to 200-fold (log[2]FC = 7.66)
   in the pyrite treatment relative to the iron-depleted treatment (fig.
   S14). The substantial up-regulation of cytochrome-related genes implied
   that C. tepidum may secrete more cytochromes to enhance extracellular
   electron transfer (EET) ([145]47), possibly facilitating sulfur
   oxidation on the surface of pyrite under iron-depleted condition. In
   contrast, genes related to iron transport and iron storage were among
   the most down-regulated ones in the pyrite treatment relative to the
   iron-depleted treatment (fig. S14). For example, genes encoding ferrous
   iron transport protein such as CT1744 [ferrous iron transport protein A
   (feoA)], CT1741 (feoC), and feoB-1 were down-regulated up to 90-fold
   (log[2]FC = −6.52) in the pyrite treatment (fig. S14). Such substantial
   down-regulations likely reflected the rapid and efficient accumulation
   of cell-associated Fe in the presence of pyrite by the time of sampling
   for transcriptomic analysis (i.e., the log phase) ([146]Fig. 2, B and
   C), reducing the need for iron transport. In addition, the
   representative iron storage gene ftn, which encodes ferritin and is
   typically overexpressed in response to iron deficiency ([147]33), was
   also substantially down-regulated when pyrite was provided (fig. S14).
   This down-regulation also suggests that pyrite served as the primary
   and effective iron source, ensuring sufficient intracellular iron
   levels for microbial growth. Moreover, the relatively minor differences
   in the abundance of iron transport and storage genes between the
   aqueous Fe^2+ and pyrite treatments further support this possibility.

Proteomic response of C. tepidum to iron scarcity

   Of a total of 2288 proteins predicted from the C. tepidum genome
   ([148]37), 371 proteins were shared across all three treatments (i.e.,
   iron-depleted treatment, 20 μM aqueous Fe^2+ amendment, and pyrite
   addition). Thresholds of log[2]FC > 1 or log[2]FC < −1 and P < 0.05
   were set to identify significant changes in protein expression (fig.
   S15). Results showed that relative to the iron-depleted treatment, 119
   proteins were up-regulated and 39 proteins were down-regulated in the
   pyrite treatment (fig. S15A). When comparing pyrite treatment with
   aqueous Fe^2+ treatment, 57 proteins were up-regulated, and 19 proteins
   were down-regulated (fig. S15B).

   In particular, the expression of iron-related proteins (i.e., requiring
   iron as a cofactor or binding with iron) was analyzed because the
   primary aim of this study was to investigate whether C. tepidum can use
   iron within pyrite under iron-depleted conditions. Compared with the
   iron-depleted treatment, proteins involved in sulfur oxidation and
   photosynthesis, particularly those requiring iron or iron-sulfur
   clusters as cofactors, were up-regulated when pyrite was provided
   ([149]Fig. 7), consistent with the transcriptomic data ([150]Fig. 6)
   and the enhanced sulfur oxidation in this treatment ([151]Fig. 1B). For
   example, proteins such as aspB (adenylyl-sulfate reductase, beta
   subunit) and dsrK (DsrK protein) exhibited substantial up-regulations
   with log[2]FC > 2 ([152]Fig. 7). Notably, most of these proteins were
   also up-regulated when comparing pyrite treatment with aqueous Fe^2+
   addition ([153]Fig. 7), which corresponded to the slightly faster cell
   growth ([154]Fig. 1A) and sulfur oxidation rate observed in the pyrite
   treatment ([155]Fig. 1B). In agreement with the transcriptomic results
   (fig. S14), iron transport (e.g., feoA-1 and feoB-1) and storage
   proteins (e.g., ferritin) were substantially down-regulated in the
   presence of pyrite compared with iron-depleted treatment ([156]Fig. 7).
   This likely reflected a dynamic response of cells to iron. Cells were
   initially forced to up-regulate iron transport and storage proteins
   under iron-depleted conditions to efficiently acquire iron from
   surrounding environments ([157]48, [158]49) but by the time of sampling
   for proteomic analysis (i.e., the log phase), intracellular iron may be
   sufficient to down-regulate these iron transport and storage proteins.
   In addition, the expression differences of proteins related to iron
   transport and storage were quite small between aqueous Fe^2+ and pyrite
   treatments ([159]Fig. 7), further suggesting that pyrite, similar to
   aqueous Fe^2+ with a high bioavailability, is an efficient iron source
   for C. tepidum in iron-depleted, euxinic environments.

Fig. 7. Differential protein expressions.

   [160]Fig. 7.
   [161]Open in a new tab

   Differential expression of proteins related to sulfur oxidation,
   photosynthesis, and iron transport and storage in C. tepidum incubated
   under varying conditions. For proteins without official gene symbols,
   the corresponding gene ordered locus names are provided. Red color
   indicates log[2]FC > 0, while blue color indicates log[2]FC < 0. No
   pathway enrichment of down-regulated gene clusters was found to be
   significant (P < 0.05).

DISCUSSION

Importance of iron in the growth of anoxygenic phototrophs

   Given the pivotal roles of iron in photosynthetic reaction centers of
   oxygenic phototrophs, it has been well recognized that iron is one of
   the commonly limited nutrients for the growth of marine phytoplankton
   because aqueous iron tends to precipitate in oxic environments.
   Fluctuations of iron concentration largely control the magnitude and
   dynamics of oceanic primary production ([162]14, [163]15) and thus are
   of significance to both modern and paleoclimate changes because the
   metabolic activities of phytoplankton regulate the carbon exchange
   between atmosphere and ocean ([164]50, [165]51). However, before the
   advent of oxygenic photosynthesis, anoxygenic phototrophs such as GSB
   dominantly contribute to the primary production in oceans, especially
   during the Proterozoic ([166]17, [167]52). In general, anaerobes,
   especially these early-originated species, are not believed to suffer
   from iron deficiency because of the presumably high solubility and
   concentration of reduced iron in paleo-oceans. However, widespread and
   recurrent euxinia in Proterozoic oceans, where aqueous iron tends to
   precipitate with H[2]S as solid iron sulfide minerals ([168]2,
   [169]20), poses a great challenge of iron deficiency to these
   anoxygenic phototrophs thriving in sulfidic environments ([170]18,
   [171]20).

   Because of the importance of oxygenic phototrophs in oxygenating the
   early Earth and in controlling carbon sequestration in the modern
   ocean, extensive investigations about how oxygenic phototrophs cope
   with iron deficiency have been conducted. Nevertheless, little is known
   about how iron deficiency affects the growth of anoxygenic phototrophs.
   Such knowledge gap largely hinders our understanding of the
   biogeochemical cycles of carbon, sulfur, and nitrogen in sulfidic
   environments. Our results here demonstrate that iron deficiency
   suppressed the growth of C. tepidum, as evidenced by a reduction in
   biomass ([172]Fig. 1A and figs. S1 and S2) and impairment of sulfur
   oxidation ([173]Fig. 1B and fig. S3). Furthermore, cultivation
   experiments also suggest that iron concentrations above 0.2 μM are
   required to sustain active growth and sulfur oxidation of C. tepidum
   ([174]Fig. 1). However, such iron concentrations are higher than those
   typically found in natural euxinic environments where aqueous ferrous
   iron is scarce, yet GSB often thrive ([175]18, [176]20). This apparent
   paradox suggests that alternative iron sources such as
   mineral-associated iron may support the growth of anoxygenic
   phototrophs. In euxinic environments, reactive iron may be incorporated
   into fine-grained pyrite, creating a spatial copresence between the
   suspended pyrite and the niches of anoxygenic phototrophs ([177]1).
   Therefore, one possible explanation is that pyritic iron supports the
   growth and metabolism of GSB, considering the dominance of pyrite in
   the iron pool of euxinic settings. Our results demonstrated that C.
   tepidum can directly use pyritic iron to meet its cellular needs
   ([178]Fig. 2), enabling active growth and sulfur oxidation even when
   aqueous iron is scarce ([179]Fig. 1). However, note that a highly
   dynamic iron-sulfur cycle occurs in euxinic environments, particularly
   near or below the chemocline where other transient iron reservoirs
   (e.g., mackinawite, greigite, and FeS generated from the progressive
   sulfidation of iron oxides) may be present before complete pyritization
   ([180]20, [181]25). Given that GSB C. tepidum is capable of using iron
   in the pyrite (i.e., the thermodynamically stable phase of iron sulfide
   minerals), these metastable iron sulfide minerals possibly also serve
   as the key iron sources for the growth of anoxygenic phototrophs.
   Overall, our findings suggest that mineral-bound iron plays a
   previously underappreciated role in sustaining anoxygenic phototrophs
   in euxinic environments.

Possible mechanisms of iron acquisition by GSB

   While iron is an irreplaceable element of many proteins and enzymes,
   previous studies have primarily focused on the iron acquisition
   strategies of aerobic microorganisms. In response to low iron
   availability, aerobic bacteria can secrete high-affinity extracellular
   ligands such as siderophores to bind and solubilize iron to meet their
   cellular iron demands ([182]49). In addition, aerobes can also uptake
   more iron than immediately needed, storing the excess for later use
   depending on the iron concentration in environments ([183]49). However,
   siderophores are generally not produced by anaerobes with a few
   exceptions ([184]53, [185]54). The genome of C. tepidum (i.e., the
   strictly anaerobic phototrophs applied in this study) does not contain
   any genes related to siderophore biosynthesis ([186]37). Consistently,
   no siderophores or siderophore-complexed iron was detected in any
   treatments, implying that C. tepidum adopts different strategies to
   meet its cellular iron requirements under aqueous iron-depleted
   conditions.

   Our findings suggest that C. tepidum acquired iron via oxidation of
   sulfur on pyrite surface. Specifically, C. tepidum modified the surface
   of pyrite with organic compounds such as lipopolysaccharides and amino
   acids, as supported by the presence of organic coatings on pyrite
   surface ([187]Fig. 4) and up-regulation of relevant pathways, genes,
   and proteins ([188]Figs. 6 and [189]7 and fig. S14). Such surface
   modifications may facilitate oxidation of pyritic sulfur through EET
   mediated by cytochromes ([190]Fig. 4 and fig. S14). It is known that C.
   tepidum can oxidize sulfide to S^0 globules and sulfate ([191]55,
   [192]56), consistent with our results (fig. S3). However, when
   solid-phase S^0 purchased from reagent companies was directly added
   into the medium, C. tepidum could not use these exogenous S^0 globules,
   despite their similar composition and structure to biogenic S^0
   globules ([193]56). This biooxidizable property of biogenic S^0
   globules has been attributed to their surface modifications by
   microbial metabolites ([194]35, [195]56) [e.g., molecular fragment with
   a m/z 72 (C[2]H[2]NO[2]^+)]. A similar organic coating was also
   observed on the pyrite surface as shown by the TOF-SIMS results in this
   study ([196]Fig. 4), suggesting that C. tepidum may modify the pyrite
   surface with comparable organics. This modification could facilitate
   the oxidation of sulfur within pyrite ([197]Fig. 5), thereby
   destabilizing the pyrite structure and promoting the release of ferrous
   iron into the surrounding environment.

   Furthermore, the requirement for direct contact between C. tepidum and
   pyrite surface ([198]Fig. 3 and fig. S8) aligns with the broader
   understanding that microorganisms often need to establish a physical
   connection with a mineral surface to exchange matter and energy
   ([199]21, [200]25, [201]47). Such mineral-microbe interaction may
   involve EET, where c-type cytochromes or other redox-active proteins
   mediate electron flow between the mineral and the microbial cell
   ([202]47). A recent study demonstrated that anoxygenic phototrophic PSB
   can oxidize sulfur within pyrite using certain monoheme c-type
   cytochromes as electron carriers under sulfur-limited conditions
   ([203]26). In addition, membrane-associated cytochromes play critical
   roles in the oxidization of extracellular solid-phase S^0 ([204]30).
   Therefore, together with the results that cytochrome-related genes were
   among the most up-regulated genes in the presence of pyrite (fig. S14),
   we speculate that C. tepidum can oxidize pyritic sulfur via EET through
   cytochromes, as indicated by the early and direct formation of sulfate
   on pyrite surface (i.e., not from oxidation of sulfide and thiosulfate;
   [205]Fig. 5), thereby destabilizing the pyrite structure and
   potentially facilitating its iron utilization. Although the exact
   mechanisms of how surface modification and EET facilitate microbial
   iron acquisition remain unclear, our findings offer insights into
   survival strategies of anaerobic microbes in iron-depleted
   environments.

Implications for bioavailability of pyritic iron in anoxic environments

   Minerals and microbes have coevolved throughout much of Earth’s history
   ([206]32, [207]57). Various metabolic processes, including anoxygenic
   photosynthesis and sulfur oxidation, require enzymes that use specific
   transition metals such as iron for catalytic functions ([208]57,
   [209]58). Consequently, the origin and evolution of microbial
   metabolisms are tightly regulated by metal availability in the
   environment ([210]32, [211]57). The capability of microorganisms to use
   mineral-associated metals is a distinct ecological advantage that
   enables them to survive and function in diverse environments ([212]53,
   [213]59, [214]60). Our study emphasizes that pyrite can potentially
   alleviate the iron limitation faced by anaerobic microbial communities
   during oceanic euxinia in paleo-ocean.

   It has traditionally been believed that pyrite is stable under anoxic,
   low-temperature conditions, and chemical weathering of pyrite typically
   occurs in oxic environments, resulting in a large-scale input of
   sulfate to oceans ([215]22–[216]24). Nevertheless, a recent study
   showed that anoxic photochemical weathering of pyrite induced by
   sunlight exposure in the presence of aqueous Fe^2+ contributes
   substantial amounts of sulfate to the oceans during the anoxic late
   Archean ([217]61). In this reaction, aqueous Fe^2+ was photooxidized as
   Fe^3+, which can lead to the chemical oxidation of sulfur within
   pyrite, releasing Fe^2+ and sulfate into solutions ([218]61).
   Similarly, the traditional view that pyrite is only accessible to
   aerobic microbes or anaerobic microbes in the presence of oxidizing
   agents (e.g., Fe^3+, NO[3]^–, and MnO[2]) ([219]22–[220]24) is
   challenged by emerging evidences ([221]21, [222]25, [223]26). For
   example, anaerobic methanogenic archaea can use iron and sulfur from
   pyrite, possibly via reductive dissolution of pyrite ([224]21,
   [225]25). Under sulfur-limited conditions, anoxygenic phototroph PSB is
   capable of oxidizing pyritic sulfur, via EET, to polysulfide or
   elemental sulfur, but not to sulfate ([226]26). However, under the
   iron-limited condition applied in this study, pyrite surface–bound
   sulfur was likely oxidized to sulfate by GSB, facilitating its iron
   acquisition. In sulfidic habitats where anoxygenic phototrophs PSB and
   GSB flourish, their metabolic activities should be limited by iron, not
   sulfur. Therefore, our observed microbial oxidation of pyritic sulfur
   under iron-depleted conditions may be more environmentally relevant in
   both ancient and contemporary euxinia, enhancing the bioavailability of
   pyritic iron toward anoxygenic phototrophs. These results imply that
   microbial weathering of pyrite may be more prevalent on early Earth
   than previously thought and may have contributed to the biogeochemical
   cycles of elements such as carbon and sulfur in the paleo-ocean.

MATERIALS AND METHODS

Bacterial strain and cultivation medium

   The representative bacterial strain of GSB C. tepidum was obtained from
   the American Type Culture Collection (ATCC 49652) and used as a model
   anaerobic photosynthetic sulfur-oxidizing bacterium due to its
   well-characterized metabolic pathways and ecological relevance
   ([227]12, [228]13, [229]37). To investigate the role of iron in the
   growth and sulfur oxidation of C. tepidum, an iron-depleted medium was
   prepared, which consisted of the following (per liter): 0.5 g of
   KH[2]PO[4], 0.5 g of MgSO[4]·7H[2]O, 0.05 g of CaCl[2]·2H[2]O, 0.025 g
   of KCl, 1.0 g of NH[4]Cl, 2 ml of trace metal solution SL-6, 0.5 ml of
   Na-resazurin solution (0.1% w/v) as a redox indicator, and 100 mM
   1,4-piperazinediethanesulfonic acid (PIPES) to buffer pH around 6.8.
   The trace metal solution SL-6 contained the following (per liter): 0.1
   g of ZnSO[4]·7H[2]O, 0.03 g of MnCl[2]·4H[2]O, 0.3 g of H[3]BO[3], 0.2
   g of CoCl[2]·6H[2]O, 0.01 g of CuCl[2]·2H[2]O, 0.02 g of
   NiCl[2]·6H[2]O, and 0.03 g of Na[2]WO[4]·2H[2]O. Although no Fe
   compounds were added to this iron-depleted medium, a trace amount of
   aqueous iron might present because of certain impurities in chemicals.
   In addition, a canonical medium with the addition of 5 ml of 0.1%
   ferric citrate solution (~20 μM Fe) was also prepared for initial
   bacterial growth and cell harvest.

   The prepared medium was aliquoted into 250-ml serum bottles (200 ml of
   solution and 50 ml of headspace) and purged with high-purity nitrogen
   (N[2], 99.999%) in a water bath (100°C) for 2.5 hours to remove
   dissolved oxygen. After autoclaving, the medium was amended with 1%
   (v/v) B[12] stock solution (0.0002%, filter sterilized and anoxic) in
   an anoxic glovebox (95% N[2] and 5% H[2], Coy Laboratory Products,
   Grass Lake, Michigan, USA). Before inoculation, sulfide (Na[2]S·9H[2]O)
   and thiosulfate (Na[2]S[2]O[3]·5H[2]O) were added from their anoxic,
   filter-sterilized stock solutions in an anoxic glovebox to achieve a
   final concentration of 2 mM each. In addition, NaHCO[3] (1.5 g/liter)
   was added as a carbon source for photoautotrophic growth of C. tepidum.

Pretreatment of pyrite particles

   Pyrite (FeS[2]) was selected as a representative iron-containing
   mineral because of its abundance in ancient and contemporary euxinic
   environments, where anoxygenic phototrophic sulfur bacteria flourish
   ([230]20). A specimen pyrite (FeS[2]) sample, purchased from a mineral
   company (Pinquan, Zhejiang, China), was manually ground and size
   fractionated to obtain a 75- to 150-μm fraction. In addition, a smaller
   size fraction of 35.5 to 75 μm was separated to investigate the effect
   of particle size on cell growth and sulfur oxidation. Subsequently, the
   sieved pyrite particles were washed to remove any possible impurities
   according to the protocol described in a previous study ([231]21).
   Briefly, in an anoxic glovebox, the particles underwent four rinses
   with 1 N HCl, one rinse with 6 N HCl, three rinses with acetone, and,
   last, four rinses with Milli-Q water (18.2 megohms). The cleaned pyrite
   particles were freeze dried and preserved in an anoxic glovebox.

Experimental setup

   All experiments were conducted in 50-ml serum bottles, which were
   presoaked in 5% nitric acid and rinsed three times with Milli-Q water
   to remove any impurities. In the following experiments, C. tepidum was
   inoculated into the serum bottles with a volume of 2%, and the
   experimental bottles were incubated at 47°C for 72 hours under
   white-light illumination (~2400 lux) ([232]12). Before the experiments
   with pyrite as the sole iron source, C. tepidum cells in the canonical
   medium were transferred three times into the iron-depleted medium to
   minimize any cellular carryover of iron. Subsequently, cells collected
   from the iron-depleted medium were inoculated again into the
   iron-depleted medium, followed by various treatments: (i) without the
   addition of iron; (ii) with the addition of various amounts of aqueous
   FeCl[2] stock solution (anoxic, filter sterilized) to achieve a final
   iron concentration of 0.2, 2, and 20 μM, respectively; and (iii) with
   the addition of pyrite particles (1 g/liter). These experiments were
   set up to investigate the importance of iron in cell growth and sulfur
   oxidation and to test whether solid-phase pyrite can serve as the sole
   iron source to support the metabolic activities of C. tepidum.

   In addition, three control experiments were designed: (i)
   quartz-substitution experiment (i.e., replacing pyrite with
   similar-sized, chemically inert quartz) to exclude the possibility that
   pyrite could enhance C. tepidum growth via the provision of attachment
   sites instead of serving as an iron source; (ii) pyrite-leachate
   experiment (i.e., pyrite was first soaked into the iron-depleted medium
   for 7 days and removed by filtering in an anoxic glovebox before
   inoculation of C. tepidum) to test whether iron from the abiotic
   dissolution of pyrite accounted for the growth promotion of C. tepidum;
   and (iii) dialysis bag experiment (i.e., pyrite was enclosed in a
   dialysis bag with a molecular weight cutoff of 500 Da) to examine the
   importance of direct contact between cell and pyrite to microbial
   growth and sulfur oxidation.

OD, BChl c, and total protein concentration measurements

   At selected time points, subsamples were taken from the reaction
   bottles using a sterilized syringe inside an anoxic glovebox for the
   measurement of OD[750] to record cell growth ([233]13, [234]62). A
   previous study has demonstrated that the formation of extracellular
   elemental sulfur by C. tepidum did not interfere with OD measurement at
   this wavelength ([235]63). In addition, BChl c (a primary
   light-harvesting pigment produced by GSB) was extracted using methanol
   and quantified spectrophotometrically at 669 nm, applying an
   established extinction coefficient ([236]64). At the end of the
   reaction, the total cellular protein concentration was measured with
   the Bradford microassay to measure any biomass difference across
   various treatments ([237]65).

Sulfide, thiosulfate, elemental sulfur, sulfite, and sulfate measurements

   At selected time points, subsamples were collected for concentration
   measurements of reactants (i.e., sulfide and thiosulfate) and products
   (i.e., elemental sulfur, sulfite, and sulfate) to monitor microbial
   sulfur oxidation. Specifically, sulfide concentration was measured with
   the methylene blue colorimetric assay ([238]66). Thiosulfate, sulfite,
   and sulfate concentrations were determined using ion chromatography
   equipped with an AS-14A anion-exchange column (4 mm by 250 mm; Thermo
   Fisher Scientific, Dionex IonPac), using 0.8 mM bicarbonate and 4.5 mM
   carbonate buffer as the eluent at a flow rate of 1 ml/min ([239]29).
   Extracellular elemental sulfur was firstly extracted with toluene
   ([240]67) and then measured with high-performance liquid chromatography
   (HPLC) equipped with a C[18] column (4.6 mm by 150 mm; Agilent, Eclipse
   Plus). The analysis was performed at a flow rate of 1 ml/min with a
   sample size of 20 μl, using a methanol/water mixture (98:2 v/v) as the
   eluent ([241]68, [242]69).

Measurements of aqueous iron in culture supernatant and cell-associated iron

   To reflect cellular iron uptake during cell growth, concentrations of
   aqueous iron in culture supernatant and cell-associated iron were
   monitored with inductively coupled plasma mass spectrometry (ICP-MS;
   PerkinElmer, NexION 300X). Briefly, samples were taken from the
   reaction tubes using a sterilized syringe and centrifugated at a speed
   of 3500g for 10 min to separate supernatant from cell biomass. Iron
   concentration in the culture supernatant was measured directly, while
   the cell-associated iron was measured after digesting the collected
   cell pellets with the concentrated hydrochloric acid (i.e., 12 N HCl)
   for 15 min. In addition, Chrome Azurol S assay was conducted to check
   whether C. tepidum produced iron-binding ligands—siderophores
   ([243]60), and the presence or absence of siderophore-complexed iron
   was further discerned with LC-ICP-MS ([244]60).

SEM observations

   To observe any morphological change of pyrite and its spatial
   relationship with C. tepidum cells, SEM observations were made for
   cell-pyrite suspensions. At the end of the reaction, suspensions of
   cell and pyrite were fixed for 90 min in a mixture of 2%
   paraformaldehyde and 2.5% glutaraldehyde buffered with phosphate buffer
   solution (130 mM NaCl and 10 mM sodium phosphate), followed by
   sequential dehydration using 25, 50, 75, 95, and 100% ethanol,
   successively ([245]70). Subsequently, samples were dried using a
   critical point dryer and coated with platinum (Pt) using a Quorum
   SC7620 sputter coater. SEM observations were conducted with a Zeiss
   SUPRA 55 SAPPHIRE scanning electron microscope equipped with a Genesis
   2000 x-ray energy dispersive spectroscopy. The accelerating voltage was
   set to 10 to 15 kV, and the working distance was 15 cm.

XRD analysis

   XRD was conducted to confirm the purity of the cleaned pyrite and to
   check whether pyrite was altered or any secondary minerals were formed
   at the end of reaction. Analysis was performed with a Rigaku SmartLab
   x-ray powder diffractometer using a rotating anode generator (Cu Kα
   radiation) with a power of 9000 W (200 kV, 45 mA). Samples were scanned
   over a 2θ range of 3° to 90°, stepping at 0.02° with a counting time of
   1 s per step. MDI Jade 6 software was used to identify mineral phases.

TOF-SIMS analysis

   To reveal any compositional changes on the surface of pyrite at the end
   of the reaction, static TOF-SIMS, a surface-sensitive technique with a
   penetration depth of 1 to 2 nm ([246]71), was applied with a TOF.SIMS5
   spectrometer (IONTOF GmbH, Münster, Germany). A pulsed Bi[3]^+ primary
   ion beam of 25 keV was used to collect the TOF-SIMS spectra. Five
   positive ion spectra and five negative ion spectra were collected per
   sample from randomly selected surface sites, with minimum counts of
   1000 and a minimum signal-to-noise ratio of 3. After background
   correction and normalization of obtained mass spectrum data, PCA was
   performed to evaluate any differences in molecular clusters between the
   untreated and microbially treated pyrite.

XPS analysis

   XPS analysis, a surface-sensitive technique with a penetration depth up
   to 10 nm ([247]72), was applied to determine the valence states of iron
   and sulfur on the surface of microbially reacted pyrite. The XPS data
   were collected using a Thermo Fisher Scientific ESCALAB 250Xi
   spectrometer with a monochromatic Al Kα x-ray source. The samples were
   prepared within an anoxic glovebox to avoid any potential air oxidation
   of pyrite. High-resolution spectra were recorded in 0.1-eV steps for
   iron and 0.05-eV steps for sulfur with a pass energy of 35 eV. The
   spectra were analyzed with Thermo Fisher Scientific Avantage software.
   The collected binding energy values were calibrated using carbon (C 1s)
   as the reference (i.e., 284.60 eV) ([248]73). The Fe 2p[3/2] and Fe
   2p[1/2] peaks were fitted with binding energies of 707.15 and 720.05
   eV, respectively ([249]73). The S 2p spectra were fitted using
   spin-orbit split doublets, with a separation of 1.18 eV between the
   peaks ([250]74). Both components were constrained to have the same full
   width at half maximum, and the area of the S 2p[3/2] peak was set to be
   twice that of the S 2p[1/2] peak ([251]26, [252]74). The S 2p[3/2]
   binding energies for monosulfide (S^2−), disulfide (S[2]^2−),
   polysulfide (S[n]^2−), and sulfate (SO[4]^2−) were assigned to 161.14,
   162.10, 164.10, and 168.10 eV, respectively ([253]26, [254]75).
   Accordingly, the S 2p[1/2] binding energies for S^2−, S[2]^2−, S[n]^2−,
   and SO[4]^2− were assigned to 162.32, 163.28, 165.28, and 169.28 eV,
   respectively ([255]26, [256]75).

Transcriptomic and proteomic analyses

   Cells at the logarithmic phase (30 hours) were harvested for
   transcriptomic and proteomic analyses. These analyses were performed
   and compared across three treatments: the iron-depleted treatment, the
   20 μM aqueous Fe^2+ treatment, and the pyrite treatment. For
   transcriptomic analysis, total RNA was extracted from samples using the
   TRIzol method ([257]76). RNA quality was assessed with a Thermo Fisher
   Scientific NanoDrop One equipped with an Agilent 4200 tape station.
   Ribosomal RNA (rRNA) was removed with the Epicentre Ribo-Zero rRNA
   Removal Kit. Libraries were constructed with the NEBNext Ultra II
   Directional RNA Library Prep Kit for Illumina, followed by quality
   control and sequencing on the Illumina NovaSeq platform (PE150). Raw
   sequencing reads were obtained in FASTQ format and subjected to quality
   control using fastp ([258]77). Clean reads were aligned to the C.
   tepidum genome using Bowtie 2, and transcript abundance was quantified
   ([259]78). PCA was performed to show the overall expression patterns of
   samples across different treatments. Differential expression analysis
   was performed using DESeq2 ([260]79). An R package, clusterProfiler,
   was used to conduct enrichment analysis for the KEGG pathway ([261]80).

   For proteomic analysis, total protein was extracted from cells by
   sonication, followed by separation using SDS–polyacrylamide gel
   electrophoresis ([262]62). The protein bands were excised from the gel,
   digested with trypsin, and analyzed by electrospray ionization Orbitrap
   MS. Peptide sequences were matched against the UniProt protein database
   derived from the C. tepidum genome using MaxQuant software (version
   2.2.6) to identify and quantify proteins. Filtering of potential
   contaminants, z-score normalization, Student’s t test, and FC analysis
   were performed in Perseus (version 2.2.1).

Acknowledgments