Methanogens acquire and bioaccumulate nickel during reductive dissolution of nickelian pyrite

ABSTRACT Nickel (Ni) is a key component of the active site metallocofactors of numerous enzymes required for methanogenesis, including [NiFe]-hydrogenase, carbon monoxide dehydrogenase, and methyl CoM reductase, leading to a high demand for Ni among methanogens. However, methanogens often inhabit euxinic environments that favor the sequestration of nickel as metal-sulfide minerals, such as nickelian pyrite [(Ni,Fe)S2], that have low solubilities and that are not considered bioavailable. Recently, however, several different model methanogens (Methanosarcina barkeri, Methanococcus voltae, Methanococcus maripaludis) were shown to reductively dissolve pyrite (FeS2) and to utilize dissolution products to meet iron and sulfur biosynthetic demands. Here, using M. barkeri Fusaro, and laboratory-synthesized (Ni,Fe)S2 that was physically isolated from cells using dialysis membranes, we show that trace nickel (<20 nM) abiotically solubilized from the mineral can support methanogenesis and limited growth, roughly fivefold less than the minimum concentration known to support methanogenesis. Furthermore, when provided direct contact with (Ni,Fe)S2, M. barkeri promoted the reductive dissolution of (Ni,Fe)S2 and assimilated solubilized nickel, iron, and sulfur as its sole source of these elements. Cells that reductively dissolved (Ni,Fe)S2 bioaccumulated approximately fourfold more nickel than those grown with soluble nickel and sulfide but had similar metabolic coupling efficiencies. While the mechanism for Ni uptake in archaeal methanogens is not known, homologs of the bacterial Nik uptake system were shown to be ubiquitous across methanogen genomes. Collectively, these observations indicate that (Ni,Fe)S2 is bioavailable in anoxic environments and that methanogens can convert this mineral into nickel-, iron-, and sulfur-containing metalloenzymes to support methanogenesis and growth. IMPORTANCE Nickel is an essential metal, and its availability has changed dramatically over Earth history due to shifts in the predominant type of volcanism in the late Archean that limited its availability and an increase in euxinic conditions in the early Proterozoic that favored its precipitation as nickel sulfide minerals. Observations presented herein indicate that the methanogen, Methanosarcina barkeri, can acquire nickel at low concentration (<20 nM) from soluble and mineral sources. Furthermore, M. barkeri was shown to actively reduce nickelian pyrite; use dissolution products to meet their iron, sulfur, and nickel demands; and bioaccumulate nickel. These data help to explain how M. barkeri (and possibly other methanogens and anaerobes) can acquire nickel in contemporary and past anoxic or euxinic environments.

whole cells grown under nickel replete and deplete conditions provided with soluble or mineral [i.e., (Ni,Fe)S 2 ] sources of Ni.The results are discussed as they relate to minimum requirements for Ni and mechanisms of its acquisition from minerals by methanogens in the euxinic environments they commonly inhabit on modern Earth and during the transition from the late Archean to the early Proterozoic.

RESULTS AND DISCUSSION
Synthesis, solubility, and reactivity of synthetic nanoparticulate (Ni,Fe)S 2 Synthetic (Ni,Fe)S 2 was prepared following previously described methods for nanopar ticulate FeS 2 (17,23) but with the replacement of 5 mol% Fe with Ni during the initial mackinawite synthesis step (see Materials and Methods for details).Briefly, the synthesis reaction proceeded through the formation of nickelian mackinawite [(Ni,Fe)S] by reacting Fe(II) and Ni(II) with stoichiometric HS -followed by the polysulfide-dependent conversion of (Ni,Fe)S to (Ni,Fe)S 2 .Synthesized (Ni,Fe)S 2 was washed several times with 1N HCl and once with 6N HCl to solubilize and thus remove surface-associated and/or unreacted metals.Synthesis of (Ni,Fe)S 2 proceeded without any observed visual differences from the synthesis reaction for FeS 2 .The average size (~500 nM) and morphology (rhombohedral or framboidal) of (Ni,Fe)S 2 grains were similar to FeS 2 (Fig. 1a  and b).The mineral composition of (Ni,Fe)S 2 , as assessed by X-ray diffraction spectro scopy, showed very minor spectral differences from FeS 2 and thus was considered a 100% match (Fig. 1c and d).Using energy-dispersive X-ray spectroscopy (EDS), the weight % Ni, Fe, and S in synthetic FeS 2 and (Ni,Fe)S 2 were compared to predicted values based on the stoichiometry of reagents used in synthesis reactions.The wt % S in synthetic FeS 2 was slightly lower than expected, and the wt % of Fe slightly higher (Table 1).This was also the case for (Ni,Fe)S 2 .However, the Ni content was comparable to the predicted wt % expected for (Ni,Fe)S 2 and was below detection in FeS 2 , confirming Ni incorporation into the mineral matrix of the latter, similar to a previous report (24).Atomic absorption spectroscopy of fully acid-digested (Ni,Fe)S 2 agreed with EDS measurements and indicated that 5.1 mol% of the Fe in the mineral was replaced by Ni (data not shown).
Bulk or specimen FeS 2 is highly insoluble (9,25), yet synthetic FeS 2 with small grain size (~500 nM) and high surface area, such as synthesized herein, is likely to be more soluble (18).The same might also be expected for (Ni,Fe)S 2 .To examine anoxic dissolu tion of (Ni,Fe)S 2 , 2 mM of the mineral (as Fe) was incubated in 35 mL of anoxic ultrapure water for 48 h at 38°C under a N 2 headspace.Following incubation, Ni and Fe were detected in the supernatant (0.22 µM filtered) at a concentration of 19 nM and 16 nM, respectively, despite Fe being at ~20-fold higher atom % abundance.This represents a solubilization of ~0.1% of the Ni and 0.001% of the Fe in the incubated (Ni,Fe)S 2 .Sulfide (HS -) was not detected following incubation (detection limit = 1.5 µM).Preferential release of Ni compared to Fe may be due to sorption of Ni on the surface of the mineral or diffusion from the mineral matrix, as has been described in other (Ni,Fe)S 2 syntheses (24).Collectively, these data indicate that the synthetic (Ni,Fe)S 2 used herein has a low solubility, and the concentrations of metals released into the solution are below the amounts commonly used to cultivate methanogen cells (0.1 to 2 µM), as discussed above and further below.
Previous studies have shown that synthetic FeS 2 can be reductively dissolved by H 2 at 38°C in aqueous solutions with pH 7.0, as evidenced by the detection of the dissolution product, HS - (17,18).To determine if the heterometal Ni influences the reactivity of (Ni,Fe)S 2 , 2 mM of synthetic (Ni,Fe)S 2 or FeS 2 was individually incubated in the presence of 200 µM dissolved H 2 at 38°C under anoxic conditions, and the reaction progress was monitored by quantifying total sulfide (HS -+ H 2 S) (Fig. 1e).The rate and the maximum amount of total sulfide produced were not significantly different in reactors containing FeS 2 or (Ni,Fe)S 2 , indicating that the reactivities of the two minerals were similar.Dissolved (< 0.22 µM) Fe or Ni was not detected in the supernatants of H 2 -reacted FeS 2 or (Ni,Fe)S 2 (data not shown).This is presumably attributable to Fe(II) and Ni(II) that were released during the dissolution process forming a complex with HS -that is also released during mineral reduction (9,10).Such complexes would likely eventually precipitate as FeS and NiS phases that are larger than the 0.22 µM pore size used to filter samples prior to metal determination.

M. barkeri strain Fusaro can acquire Ni, Fe, and S from (Ni,Fe)S 2
Several methanogen strains have been shown to reductively dissolve FeS 2 to liberate Fe and S to support element demands for activity and growth (17,18,20,21).Here, we tested whether M. barkeri Fusaro can also liberate Ni from (Ni,Fe)S 2 during reductive dissolution to meet its high Ni demand (13,16).Importantly, M. barkeri Fusaro was grown in low salt medium to limit contaminant metals in salts and to improve ICP-MS detection limits for Fe and Ni by avoiding unnecessary dilution.Furthermore, cells in this study were not grown with tryptone, yeast extract, or cysteine such as to limit the source(s) of Ni, Fe, and S to only those that were provided.This had the effect of decreasing the overall growth kinetics when compared to other studies where one or more of those substrates were provided (26,27).M. barkeri Fusaro was first cultivated with soluble sources of Fe (20 µM Fe(II) added as FeCl  M. barkeri Fusaro cells were then grown with (Ni,Fe)S 2 as their sole Fe and S source with 1 µM added soluble Ni(II) or without added Ni.No significant difference in growth (CH 4 or DNA production) was observed between these conditions (Fig. 2a and b).Furthermore, CH 4 and DNA production in cultures provided with (Ni,Fe)S 2 [either with added soluble Ni(II) or no added Ni] were not significantly different from those grown with soluble sources of Fe, S, and Ni (Fig. 2a and b).This indicates that Ni and Fe were not limiting during growth on (Ni,Fe)S 2 , either in the presence or absence of added soluble Ni(II).Yet, the cells under both conditions did not enter log phase growth, as previously seen in M. barkeri Fusaro growth curves (26,27).Prior studies also typically provided M. barkeri Fusaro with cysteine as an additional sulfur source/reductant, suggesting the cells in this study might have been sulfur limited, a feature that may have been exacerbated by the cells growing as aggregates (discussed below).Limitation of sulfur solubilized from a mineral (in this case as an oxidant) has also been shown to keep the archaeon Acidilobus sulfurireducens from entering log-phase growth (28) and may point to enhanced turnover of cells during surface-dependent growth.Nonetheless, in this study the elimination of cysteine from the growth medium was necessary to demonstrate that (Ni,Fe)S 2 can be used as a source of Ni, Fe, and S for M. barkeri Fusaro.
During growth on (Ni,Fe)S 2 , HS -increased in concentration up to three days of incubation and then slowly declined (Fig. 2c) without a corresponding change in the pH of the growth medium pre-and post-growth (data not shown).The initial increase in total HS -is indicative of biologically mediated (Ni,Fe)S 2 reduction (17,18,21).The decrease in HS -after three days of incubation may be due to the rate of HS -production from mineral reduction being lower than the rate of HS -assimilation since methanogens require more S than Fe and Ni for biosynthesis (13), and this would be the timeframe when the cells would typically enter log-phase growth.Alternatively, it is possible that the HS -produced from mineral reduction complexed with metals in the growth medium and formed particles larger than 0.22 µM, the pore size used to filter the growth medium prior to HS -determination.
Previous studies have shown that, in the absence of soluble electron shuttles [e.g., hydrogen or anthraquinone-2,6-disulfonate (AQDS)], direct contact is required for methanogen-mediated FeS 2 reduction (17,18).Here, field emission scanning electron microscopy (FE-SEM) was used to visualize M. barkeri Fusaro cells provided with (Ni,Fe)S 2 as their sole source of Ni, Fe, and S (Fig. 3).Cells grew as aggregates, common for this genus (22) when grown in low salt medium (22,29), and appeared in close contact with the (Ni,Fe)S 2 framboids.Thin filaments (likely dehydrated extracellular polymeric substance) were observed between cells and (Ni,Fe)S 2 surfaces.The micro scale association observed between M. barkeri Fusaro cells and (Ni,Fe)S 2 supports macroscale observations that cells produce biofilms that encapsulate (Ni,Fe)S 2 fram boids, similar to what has been described for Methanococcus voltae and FeS 2 [see Supplemental Videos in (17)].This close contact is needed to support EET to reduce (Ni,Fe)S 2 or FeS 2 and liberate dissolution products required for biosynthesis (18).Further experiments using dialysis tubing to physically isolate cells from (Ni,Fe)S 2 and its effects on growth are described below.

M. barkeri Fusaro acquires low nM concentrations of Ni from abiotic dissolu tion of (Ni,Fe)S 2
Experiments were conducted to determine if trace Ni released from abiotic dissolution of (Ni,Fe)S 2 in growth medium (pH 7.0) at low temperature (38°C) could support metha nogen activity and growth.Our previous experiments (discussed above) revealed that ~19 nM Ni was released from (Ni,Fe)S 2 over 48 h of incubation at 38°C.To examine if this could support the growth of M. barkeri Fusaro, (Ni,Fe)S 2 was sequestered in dialysis tubing with pore sizes equivalent to 12-14 kDa.By sequestering the mineral, direct contact between M. barkeri cells and (Ni,Fe)S 2 was prohibited, thereby preventing the biological reduction of the mineral (18) and forcing cells to use only Ni leached from the mineral (Ni-leached).Production of CH 4 and DNA was determined, since the cells grew as aggregates (see Materials and Methods) making it difficult to quantify growth using microscopy.CH 4 and DNA production in these cultures was compared to (i) a negative control with no added Ni, Fe, or S; (ii) a positive control provided with 1 µM soluble Ni(II), 20 µM Fe(II), and 2 mM HS -(Ni-replete); (iii) an experimental control with 20 µM soluble Fe(II) and 2 mM HS -but no Ni (Ni-deplete); and (iv) an experimental control with 20 µM soluble Fe(II), 2 mM HS -, and 2 mM (as Fe) (Ni,Fe)S 2 that was free in solution (Ni-mineral).Although dialysis membranes and clips underwent thorough washing prior to use in dialysis experiments (see Materials and Methods), untied membranes and clips were added to the negative control conditions to account for any trace addition of Fe and/or Ni to the growth medium from these materials.
To further explore the effect of Ni, Fe, and S sources on the growth of M. barkeri Fusaro, the metabolic coupling efficiency, or the ability of cells to couple their energy metabolism to production of biomass (measured here as µg DNA/mmol CH 4 ), of cells cultivated under the varying conditions was calculated during the early (days 0-7) and late phases (days 7-13 ) of growth.Prior to this determination, however, a series of experiments was performed that showed that FeS 2 has no effect on the extraction and recovery of M. barkeri DNA (Fig. S1).During the early phase of growth (days 0-7), metabolic coupling efficiencies were similar across the three conditions where Ni was supplied either as soluble Ni (as NiCl 2 ; Ni-replete) or mineral Ni (as (Ni,Fe)S 2 ) that was sequestered (Ni-leached) or free in solution (Ni-mineral) (Fig. 4c).In contrast, cells grown with soluble Fe(II) and HS -but with no Ni (Ni-deplete) showed a decreased metabolic coupling efficiency during the early phase growth, which is attributed to Ni limitation that developed during this growth period.
During the late-phase growth (days 7-13) metabolic coupling efficiencies further diverged from those during the early phase growth but in unexpected ways.In the case of the Ni-replete-and Ni-mineral-grown cells, metabolic coupling efficiencies at late phase were threefold and fourfold higher than in early phase, respectively.In contrast, metabolic coupling efficiencies of the late-phase Ni-deplete-or Ni-leached-grown cells were not significantly different from cultures at early phase.It is not immediately clear why metabolic coupling efficiencies varied as they did in early phase versus late-phase cultures but it is unlikely to be attributable to minerals influencing DNA extraction or recovery efficiencies (Fig. S1).Comparing across treatments, Ni-replete cells had significantly higher metabolic coupling efficiencies than the Ni-deplete cells (Student's ttest, P = 8.58 × 10 −3 , n = 3), while the Ni-mineral grown cells had significantly higher metabolic coupling efficiencies than Ni-leached cells (P = 9.29 × 10 −3 , n = 3).Collectively, these results demonstrate that cells can acquire additional Ni from (Ni,Fe)S 2 through reductive dissolution, and this supports the growth that is comparable to those grown under Ni-replete conditions.

Bioaccumulation of Ni from (Ni,Fe)S 2 during reductive dissolution
The amount of Ni bioaccumulated in M. barkeri Fusaro biomass in Ni-replete and Ni-deplete growth conditions with either soluble or mineral sources of Ni was determined.However, because M. barkeri Fusaro directly attaches to (Ni,Fe)S 2 during growth, it was first necessary to develop a cultivation system whereby reductive dissolution could be driven by cells that were physically isolated from the bulk mineral to avoid mineral contamination of biomass.We have shown that the synthetic electron shuttle, AQDS, facilitates EET from M. barkeri to FeS 2 and allows for growth when the mineral is sequestered in dialysis tubing (18).Thus, growth experiments were conducted determine the Ni content of cells provided with AQDS and with Fe(II), HS -, and (Ni,Fe)S 2 sequestered in dialysis tubing (Ni-mineral), and these were compared to (i) Ni-replete cells grown with soluble Fe(II), HS -, and Ni(II); (ii) Ni-deplete cells grown with soluble Fe(II), HS -, and no Ni; and (iii) Ni-leached cells grown with Fe(II), HS -, and (Ni,Fe)S 2 sequestered in dialysis tubing without added AQDS.Again, control cultures contained untied dialysis tubing and clips to control for contaminant elements that may contribute to the growth medium.The concentration of total dissolved (0.22 µm filtered) Ni in each cultivation reactor was determined pre-(0 days) and post-cultivation (13 days) for these conditions to attempt to account for the Ni present in each.Furthermore, prior to elemental analysis, cells were washed with nitriloacetic acid (NTA) to minimize the contribution of cell surface-associated Ni.The concentration of NTA used herein (5 mM) to remove sorbed metals from cells is less than that (10 mM) has been previously used to remove sorbed metals from the methanogen, Methanocaldococcus jannaschii, while maintaining cell viability (30).That said, bioaccumulation is defined for the purposes of this study as cell-associated since it is not definitively known if the metal was associated with extracellular components (sorption) or was intracellular.
Normalization of the amount of Ni bioaccumulated to DNA (proxy for growth) only further pronounced the differences in Ni availability and accumulation in the various culture conditions, with the highest normalized bioaccumulated Ni content in Ni-mineral cells (Fig. 5).The Ni content of Ni-mineral grown cells was approximately fourfold higher than Ni-replete-grown cells (Ni-mineral: 15.70 ± 2.5 ng Ni/µg DNA; Ni-replete: 4.22 ± 0.33 ng Ni/µg DNA, n = 4), and the difference was significant (Student's t-test, P = 9.61 × 10 −6 , n = 4).Importantly, the Ni content of M. barkeri Fusaro grown in batch or continuous culture and with different carbon sources and methanogenesis substrates (i.e., methanol, acetate, H 2 /CO 2 , trimethylamine) with HS -as sulfur source/reducing agent was previously shown to vary by up to a factor of two (13).This suggests that broad changes in carbon and methanogenesis pathways are unlikely to solely account for the Ni content of cultures observed herein.Furthermore, the Ni content of dried biomass from Ni-replete, Ni-deplete, Ni-mineral, and Ni-leached cultures was 35 ± 5.0, 6.4 ± 1.0, 163 ± 12.8, and 10 ± 2.0 ppm, respectively (Table S1).Values for Ni-deplete, Ni-leached, and Ni-replete cultures were twofold to 10-fold lower than what has been measured previously (range of 60-150 ppm) for a variety of methanogens grown on various substrates and with HS - as sulfur source (13).Only Ni-mineral-grown cells fall within this previously determined range (Table S1).As mentioned above, cells in this study were washed with NTA (5 mM) and then with MQ water to dissociate and remove membrane-associated metals or metal clusters prior to element analysis.In contrast, previous reports of Ni content per cell were from methanogen cultures that were washed with only water prior to analysis (13), which may have led to an overestimation of the Ni content of methanogen biomass due to sorption of Ni or Ni sulfide clusters.Again, the concentration of NTA used in this study was half of that used previously and which was shown to not affect viability and thus membrane integrity in M. jannaschii (30).

Probing potential mechanisms of Ni uptake in methanogens
Little is known of nickel uptake in methanogens; however, in other cells, it must be modulated to avoid metal toxicity (31,32), and the same is likely true for methanogens.In E. coli, the NikR repressor binds to the promoter region of the nikABCDE operon when Ni is present, preventing transcription (33,34).A prior bioinformatics study of NikR binding sites in eight methanogen genomes identified homologs of a second puta tive ABC transporter involved in cobalt (Co) uptake (CbiMNQO) near genomic regions encoding Ni-dependent enzymes (i.e., [NiFe]-hydrogenase, CODH) (35).Due to their location near genes encoding Ni-dependent enzymes in methanogen genomes, these homologs were hypothesized to be involved in maintaining Ni homeostasis and have been biochemically shown to have preference for uptake of Ni in several bacteria.As such, the genes were designated nikMNQO to differentiate them from Co uptake systems (35).NikQ and NikM are thought to be permeases, NikN has no predicted function, and NikO is predicted to be an ATPase that together functions as a high-affinity Ni 2+ transport system (<100 nM) regulated by NikR (35).
Using a previously compiled database containing 301 methanogen genomes from the Department of Energy-Joint Genome Institute (DOE-JGI) (36), the distribution of genes encoding NikMNQOR homologs was examined (Table 2; Table S2).Every genome analyzed had at least one homolog of a component of the Nik system, with every genome encoding a homolog of NikO (ATPase) and oftentimes multiple homologs, which likely also included Co transport paralogs.Nearly all (98.0%) genomes encoded homologs of NikR.Most genomes defined as being complete (empirically defined as those with >95% estimated completeness through CheckM analyses as reported by DOE-JGI for each genome) encoded copies of both NifQ and NikM, with those that did not typically encode at least one homolog of NikQ or NikM.The genomes that did not encode a NikQ or NikM homolog were from an organism belonging to the more recently diverging order Methanocellales (Methanocella paludicola SANAE) and three organisms belonging to the early diverging order Methanopyrales (Methanopyrus kandleri AV19, Methanopyrus KOL6, and Methanopyrus SNP6).Together, these results suggest that the Nik system may be involved in Ni homeostasis in diverse methanogens, including those that are more early evolved [termed Type I based on phylogenetic branching (37,38)] and those that are more recently evolved (termed Type II).
Potentially in support of a role in Ni uptake, two gene clusters that encode homologs of Nik/CbiMNQO were identified in the Type I methanogen Methanococcus maripaludis S2 (locus tags MMP1481-1484 and MMP0885-MMP0889), one of which (MMP1481-1484) is co-localized with genes encoding the membrane-associated [NiFe]-hydrogenase, Eha (locus tags MMP1448-1467).A previous study generated transposon insertions randomly in the genome of M. maripaludis S2 and used these to identify the essentiality of encoded RNAs and proteins during growth with H 2 /CO 2 (39).Scores of <3 and <2 were deemed diagnostic of essential protein-encoding genes in rich and minimal medium, respectively, and scores of >11 and >5 were deemed diagnostic of possibly nonessential protein-encoding genes in rich and mineral medium.In rich medium containing yeast extract, casamino acids, acetate, vitamins, cysteine, and coenzyme M, the essentiality of MMP1481-1484 ranged from 18 to 48 and MMP0885-MMP0889 ranged from 9 to 46 indicating that they are potentially nonessential under this growth condition.However, in mineral medium with cysteine as the only organic amendment, the essentiality of MMP1481-1484 ranged from 6 to 14 and MMP0885-MMP0889 ranged from 0 to 18; the essentiality of Eha (MMP1448-1467) minus subunits PQRS (MMP1463-1466) was <3 regardless of conditions.This is consistent with a potential role for MMP1481-1484 in Ni acquisition and homeostasis in M. maripaludis S2.The essentiality scores (borderline nonessential) may also suggest alternative mechanisms of acquiring Ni in M. maripaludis S2 and possibly other methanogen strains such as M. barkeri Fusaro or some level of redundancy with other uptake systems.While Nik may be involved in maintaining Ni homeostasis in methanogen cells, it does not necessarily help to explain bioaccumulation of Ni in those cells, if the Ni is intracellular.The Nik system is specific for Ni 2+ (32), which, like other thiophilic metals (e.g., Fe 2+ ), is likely to be present as a nickel-sulfide in sulfidic medium or environments (8)(9)(10).It is possible that the high content of Ni is due to sorption on the cell surface, although steps were taken to minimize this possibility (i.e., wash with NTA).An alter native explanation for Ni bioaccumulation in cells grown with (Ni,Fe)S 2 sequestered in the presence of AQDS in this study comes from experiments conducted with the methanogen Methanococcus voltae when grown on FeS 2 (19).In that study, M. voltae cells grown with FeS 2 as the sole source of Fe and S exhibited approximately twofold higher Fe contents than cells grown with FeCl 2 and HS -.The excess Fe was shown to be stored as an intracellular thioferrate-like mineral (19).Despite FeS 2 -grown M. voltae cells apparently being replete with Fe, shotgun proteomics of these cells revealed up-expres sion of the Fe(II) transporter (FeoB) and the Fe(II) regulators FeoA and DtxR relative to FeCl 2 /HS --grown cells.In bacteria, the Feo system is upregulated in response to Fe(II) limitation and is involved in Fe(II) transport (40).In methanogens, homologs of genes encoding FeoAB system are ubiquitous (36), and in M. maripaludis S2, genes encoding FeoAB are essential (39).In the case of up-expression of FeoAB and DtxR in M. voltae cells grown with FeS 2 , it was posited that cells incorrectly sensed Fe limitation during growth on FeS 2 because Fe(II) was complexed with HS - (17).It was suggested that passive transport of uncharged iron-sulfur aqueous clusters resulted in the bioaccumulation of Fe.Like Fe(II), Ni(II) has a strong affinity for HS -, although it is not clear if charged [(Ni(HS) 2 ) 0 ] or neutral [Ni(HS) + ] aqueous clusters are predominant under the conditions of the experiments conducted herein (9).If the Ni-sulfide clusters are uncharged, it may lead to intracellular bioaccumulation of Ni in M. barkeri Fusaro in a manner similar to that previously observed with Fe in M. voltae.Additional work is warranted to determine (i) the mechanism of Ni uptake/transport/bioaccumulation in these cells, (ii) the form of Ni that is bioaccumulated, either through extracellular sorption or as intracellular phases, and (iii) the extent that Ni from mineral sources such as (Ni,Fe)S 2 supports methanogens (and potentially other anaerobes) in natural environments.Regardless, the observations presented herein help to explain how methanogens acquire Ni at low concentrations and from mineral sources that are expected to be predominant in sulfidic habitats, both today and in the geologic past.

Preparation of synthetic and natural minerals
Synthetic nickelian pyrite (Ni,Fe)S 2 was prepared by modifying previously described FeS 2 synthesis methods (17,23) to include a Ni impurity at a final concentration of 1.67%.Specifically, two separate 50 mL solutions (A and B) were prepared in sterile, anoxic ultrapure MilliQ water (MQ H 2 O) inside of an anaerobic chamber (97.5%:2.5% N 2 :H 2 ).Solution A contained 57 mmol FeSO 4 •7H 2 O and 3 mmol NiCl 2 •6H 2 O, while solution B contained 60 mmol Na 2 S. Solutions A and B were combined and mixed on a stir plate with a magnetic stir bar for 10 mins.Next, 2.1 g of elemental sulfur (S 0 ; previously baked at 95°C to dry and sterilize) was added to the mixture of solutions A and B, and it was stirred for an additional 15 min.The mixture was then transferred to a glass serum bottle, sealed with a blue butyl rubber stopper, and crimped closed.The bottle was removed from the anaerobic chamber, and the solution was bubbled for 1 h/L with sterile N 2 gas that had been passed over a heated (210°C), H 2 -reduced copper column.All chemicals used in the synthesis were of American Chemical Society grade, and all glassware was washed with 10% nitric acid prior to use.The synthesis reaction was then allowed to incubate in the dark in the sealed anaerobic bottle for four days at 65°C, followed by an additional four days at 85°C.After the synthesis had completed, to remove any unreacted HS -, Fe(II), FeS, Ni(II), NiS, and S 0 , the synthesized (Ni,Fe)S 2 was washed via centrifugation and decanting initially with 1N HCl, then with boiling 6N HCl, then twice with MQ H 2 O, then three times with >99.5% acetone, and finally five times with sterile, anoxic MQ H 2 O within an anaerobic chamber.The final synthetic (Ni,Fe)S 2 slurry was transferred to a sterile, anoxic, acid-washed serum bottle and then capped and sealed prior to removal from the anaerobic chamber and sparging with sterile and anoxic N 2 (see above) to remove residual H 2 .

Mineral characterization and reactivity
Washed minerals were dried under a stream of 0.22 µM filtered ultrahigh purity N 2 gas prior to characterization on a SCINTAG X-1 system X-ray powder diffraction (XRD) spectrometer (XRD; Eigenmann GmbH, Mannheim, Germany) and a Zeiss SUPRA 55VP field emission scanning electron microscope (FE-SEM; Zeiss, Oberkochen, Germany) equipped with an energy dispersive X-ray spectrometer (EDS).A small aliquot of the mineral sample (~0.1 g) was dried under a stream of 0.22 µM filtered ultrahigh purity N 2 gas and then dissolved in 40% trace metal-free HNO 3 for three days at room temperature (~21°C).The dissolved mineral was diluted with MQ H 2 O to 3% HNO 3 and passed through a 0.22 µM PTFE syringe filter prior to quantification of Fe and Ni via atomic absorption (AA) spectroscopy on an Agilent 240 FS instrument (Agilent Technologies Inc., Santa Clara, CA) equipped with Fe and Ni lamps utilizing an acetylene/air mixture (11:60 lb/in 2 partial pressures) as a fuel source.The Fe and Ni contents for the samples were determined using standard curves prepared from 1,000 ppm Fe or Ni standards (Ricca Chemical Company, Arlington, TX), respectively.All dilutions and standards were prepared in fresh 3% HNO 3 .
To compare the reactivity of synthetic FeS 2 to that of (Ni,Fe)S 2 , minerals were reacted with H 2 , and HS -, a product of FeS 2 reduction, was quantified following previously described methods (18).Briefly, 2 mM (as Fe) of FeS 2 or (Ni,Fe)S 2 was incubated in 35 mL of anoxic MQ H 2 O with either 100% N 2 headspace (negative control) or 90%:10% N 2 :H 2 headspace (experimental control).Triplicate reactors for each headspace condition and each mineral type were incubated for seven days at 38°C.Dissolved HS -was measured at 0, 2, 24, 96, and 168 h by removing liquid samples with a N 2 -purged needle and syringe, passing the sample through a 0.22 µM filter, and quantifying HS -via the methylene blue assay (41).Dissolved HS -concentrations were converted to total sulfide (HS -+ H 2 S) using Henry's law.At the end of incubation, the supernatant from each reactor was filtered through a 0.22 µM PTFE syringe filter and then acidified in 5% trace metal-free HNO 3 prior to the quantification of dissolved Fe and Ni via inductively coupled plasma mass spectrometry (ICP-MS) on a Thermo iCAP Q ICP-MS (Thermo Fisher Scientific, San Jose, CA) at the Montana Bureau of Mines and Geology Analytical Laboratory at Montana Technological University.

Cultivation conditions
M. barkeri strain Fusaro was purchased from the American Type Culture Collection (ATCC-BAA-2329).Although it has been shown that M. barkeri Fusaro disaggregates in high salinity medium, the high salt content would have prohibited the accurate quantification of Fe and Ni on ICP-MS and AA spectroscopy.Thus, the cells were grown in low salinity medium as previously described (18).All growth medium and amendments were prepared without Fe, Ni, or S in ultrapure, MQ water and using acid (10% HNO 3 )washed glassware.The base salts medium contained (in g/L): NaCl, 1.00; MgCl 2 ⋅ 6H 2 O, 0.40; NH 4 Cl, 0.50; KCl, 0.50; CaCl 2 ⋅ 2H 2 O, 0.10; KH 2 PO 4 , 0.15; NaHCO 3 , 2.0.Prior to adding the NaHCO 3 , the remaining base salts medium was boiled for 10 min, then purged with anoxic N 2 gas (see above) for 1 h/L.The sparged medium was capped, sealed, and then moved into an anaerobic chamber where the NaHCO 3 was added, and the pH was adjusted to 7.0 using 1N HCl.The medium (75 mL) was dispensed into 165 mL serum bottles, capped with a blue butyl rubber stopper, and sealed with crimp caps before removing from the anaerobic chamber.The headspace of bottles was exchanged with N 2 :CO 2 (80%:20%) gas for 15 min, before autoclaving.
Prior to inoculation, the medium was amended with 1% (vol/vol) Wolfe's vitamins, 1% SL-10 trace metals (no Fe-or Ni-containing components), 0.5% (vol/vol) or 123 mM methanol, and 4 mM acetate.All cultures were grown with a N 2 :CO 2 (80%:20%) headspace pressurized to 1.72 atm.Where indicated, Fe(II) (as FeCl 2 ) was added to a final concentration of 20 µM, Ni(II) (as NiCl2) was added to a final concentration of 1 µM, sulfide (as Na 2 S) was added to a final concentration of 2 mM, FeS 2 [as Fe(II)] was added to a final concentration of 2 mM, and (Ni,Fe)S 2 was added to a final concentration of 2 mM.Cultures were inoculated with a 5% (vol/vol) transfer of cells from a mid-log phase culture that was first pelleted via centrifugation (4,696 × g for 20 min at 4°C) and then washed with 5 mM nitriloacetic acid in sterile and anoxic medium; the nitrilotriacetic acid (NTA) was used as a chelator to remove residual metals, including Fe and Ni.
For experiments where (Ni,Fe)S 2 was sequestered to prohibit direct physical contact between the mineral and cells, 12-14 kDa dialysis tubing and clips (Spectrum Laborato ries, Rancho Dominguez, CA) were first washed in MQ water and sterilized in ethanol as previously described (17).Once washed and sterilized, one end of the tubing was clipped and then the tubing was filled with a (Ni,Fe)S 2 slurry and sealed with the other clip.Sealed tubing containing (Ni,Fe)S 2 was added to the culture bottles, and the bottles were capped and sealed within the anaerobic chamber.Other growth conditions and controls for the dialysis experiments contained empty washed and sterilized tubing and clips.Once removed from the chamber, the headspace of the bottles was exchanged with N 2 :CO 2 (80%:20%) for 15 min to remove residual H 2 prior to inoculation.Where indicated, anoxic and 0.22 µm filtered anthraquinone-2,6-disulfonate (AQDS) was added to cultures to a final concentration of 2 mM.

Quantification of cellular growth, activity, and metal content
Growth of M. barkeri Fusaro was monitored via DNA quantification, as a proxy for growth, because the cells strongly attach to FeS 2 and (Ni,Fe)S 2 that prohibit accurate cell quantification through microscopy or optical density (OD) approaches.Similar to previously described methods (18), 1 mL of culture was anaerobically and aseptically removed from bottles and centrifuged at 20,000 × g for 20 min at 20°C to pellet cells.The supernatant was discarded and the pellets were resuspended in 489 µL of sodium phosphate buffer (MP Biomedicals, Irvine, CA) and 61 µL MT buffer (MP Biomedicals).The cell suspension was subjected to three freeze (−80°C) thaw (70°C) cycles and then transferred to a sterile 2 mL screw-top tube that contained 100 mg of 0.1 mM glass beads (Biospec Products, Bartlesville, OK).The tubes were shaken on a bead beater (Biospec Products) for 40 s and then centrifuged at 14,000 × g for 15 min at 20°C to pellet cellular debris, thereby leaving DNA in the supernatant.The DNA concentration was quantified fluorometrically using a Qubit HS dsDNA kit and fluorimeter (Invitrogen, Carlsbad, CA).
To quantify CH 4 , 100 µL of headspace gas from culture bottles was sampled with a N 2 -flushed gas-tight syringe and analyzed via gas chromatograph on a Shimadzu GC-2014 (Shimadzu Scientific Instruments, Columbia, MD) equipped with a 2.0 m HayeSep Q 80/100 column operated at 30°C.A thermal conductivity detector, set at 150°C and 50 mA, was used to measure CH 4 using ultrahigh purity N 2 as a carrier gas.A standard curve for CH 4 was generated using a dilution series prepared from a 100% CH 4 gas standard (EGAS Depot, Nampa, ID), and this was used to convert the measured peak area to ppm.Total CH 4 (dissolved and gas phase) was calculated using Henry's Law.Dissolved HS -in cultures was quantified using the methylene blue assay as described above, and total HS -(dissolved and gas phase) was calculated using Henry's Law.
Biomass from cultures was analyzed for Ni content by harvesting cultures at the end of growth experiments (day 13) via centrifugation (4,696 × g for 20 min at 4°C).The supernatant was removed, and the biomass was resuspended in sterile, anoxic base salts medium with 5 mM nitriloacetic acid to chelate and remove residual metals, including Ni. Biomass was pelleted once more via centrifugation (4,696 × g for 20 min at 4°C), resuspended in a minimal amount of ultrapure water, and transferred to a pre-weighed, sterile 2 mL screw-top tube.The suspended biomass was dried on a heat block at 70°C overnight and then weighed again to determine biomass dry weight.Next, biomass was digested by dissolving in 0.75 mL of 10% trace metal-free HNO 3 in a 98°C heat block with periodic vortexing to encourage digestion.Upon complete digestion (~30 h), samples were diluted with 0.75 mL MQ H 2 O and passed through a 0.22 µm PTFE filter prior to analysis on AAS to quantify Ni content, as described above.

Field emission scanning electron microscopy
M. barkeri Fusaro cells were grown as described above with (Ni,Fe)S 2 as the sole source of Ni, Fe, and S. At mid-log phase, a subsample of cells was collected and subjected to glutaraldehyde fixation and dehydration, as previously described (17).Images were collected using a high-resolution FE-SEM (Supra 55VP, Zeiss, Thornwood, NY) with a primary electron beam energy of 1 keV at different magnifications with technical support from the Imaging and Chemical Analysis Laboratory at Montana State University.

FIG 1
FIG 1 Synthetic nickelian pyrite (Ni,Fe)S 2 has similar physical, chemical, and structural properties and reactivity as synthetic pyrite (FeS 2 ).a-b, Field emission scanning electron micrographs of synthetic FeS 2 framboids (a) and synthetic (Ni,Fe)S 2 framboids (b) (scale bar = 250 nM in both panels).(c and d), X-ray diffraction (XRD) spectra for synthetic FeS 2 and synthetic (Ni,Fe)S 2 .(e) The production of total sulfide (dissolved plus gas phase) from abiotic reduction of either synthetic FeS 2 (blue squares and line) or (Ni,Fe)S 2 (red circles and line) by H 2 incubated at 38°C.No sulfide was detected in abiotic reactors containing FeS 2 or (Ni,Fe)S 2 when the headspace was N 2 (data are not shown).

FIG 2
FIG 2 Growth and activity of Methanosarcina barkeri Fusaro with nickelian pyrite [(Ni,Fe)S 2 ] or soluble sources of nickel (Ni), iron (Fe), and sulfur (S).Production of total methane (CH 4 ) (a) and DNA (proxy for growth) (b) in cultures of M. barkeri Fusaro grown with acetate and methanol as methanogenesis substrates and with Fe(II) and HS -with and without added Ni(II), as the sole source of Ni, Fe, and S or with (Ni,Fe)S 2 with and without added Ni(II), as the source of Ni, Fe, and S, as specified in the legend.The production of total sulfide (dissolved plus gas phase) was only quantified in reactors provided with (Ni,Fe)S 2 (c).CH 4 , DNA, or sulfide was not produced in abiotic controls (data not shown).

FIG 3
FIG 3 Field emission scanning electron microscopy image of Methanosarcina barkeri strain Fusaro in close association with (Ni,Fe)S 2 rhombohedral nanoparticles during growth.M. barkeri Fusaro was grown with acetate and methanol as methanogenesis substrates and with (Ni,Fe)S 2 as the sole source of nickel, iron, and sulfur.Scale bar equals 2 µM.

FIG 4
FIG 4 Access to nickelian pyrite [(Ni,Fe)S 2 ] enhances reductive dissolution and increases growth efficiency in Methanosarcina barkeri Fusaro.Production of total methane (CH 4 ) (a) and DNA (proxy for growth) (b) in cultures of Methanosarcina barkeri Fusaro grown with acetate and methanol as methanogenesis substrates and soluble forms of Fe, S, and Ni and/or nickelian pyrite [(Ni,Fe)S 2 ] free in solution (Free) or sequestered in dialysis tubing (Seq.), as specified in the legend.(c)The metabolic coupling efficiency (i.e., µg of DNA produced per mmol of CH 4 produced) calculated during the early phase (0-7 days) and during the late phase (7-13 days) of growth for each condition.CH 4 , DNA, or sulfide were not detected in abiotic controls (data not shown).

FIG 5
FIG 5 Methanosarcina barkeri Fusaro bioaccumulates nickel released through reductive dissolution of nickelian pyrite [(Ni,Fe)S 2 ].Total dissolved (< 0.22 µM) nickel (Ni) quantified in each reactor before methanogen growth (day 0, pre-growth) or following methanogen growth (day 13, post-growth) and the total bioaccumulated Ni in M. barkeri Fusaro biomass.Cells were grown with acetate and metha nol as methanogenesis substrates and soluble forms of Fe, S, and Ni and/or (Ni,Fe)S 2 sequestered in dialysis tubing (Seq.) with or without a synthetic electron shuttle, AQDS, provided to facilitate reductive dissolution of the mineral, as specified in the legend.Normalized bioaccumulated Ni content was calculated by dividing the total bioaccumulated Ni content in the biomass for each reactor by the total amount of DNA that was recovered from the biomass of each reactor.*, not detected.

TABLE 1
The predicted and measured weight percent of sulfur, iron, and nickel in synthetic pyrite (FeS 2 ) and nickelian pyrite ((Ni,Fe)S 2 ) cSulfur predicted (%) Sulfur measured (%) aIron predicted (%) Iron measured (%) a a Data represent the mean and standard deviation of the mean (n = 3).b ND, not detected.c Predicated elemental compositions were from the stoichiometry of reagents used in mineral syntheses, while measured elemental compositions were determined by energy-dispersive X-ray spectroscopy.

TABLE 2
Distribution of genes encoding homologs of NikMNQOR among 301 methanogen genomes a Finished defined as having >95% estimated completeness by CheckM.Full-Length Text Applied and Environmental Microbiology October 2023 Volume 89 Issue 10 10.1128/aem.00991-2311