Gene Identification and Substrate Regulation Provide Insights into Sulfur Accumulation during Bioleaching with the Psychrotolerant Acidophile Acidithiobacillus ferrivorans

Abstract

The psychrotolerant acidophile Acidithiobacillus ferrivorans has been identified from cold environments and has been shown to use ferrous iron and inorganic sulfur compounds as its energy sources. A bioinformatic evaluation presented in this study suggested that Acidithiobacillus ferrivorans utilized a ferrous iron oxidation pathway similar to that of the related species Acidithiobacillus ferrooxidans. However, the inorganic sulfur oxidation pathway was less clear, since the Acidithiobacillus ferrivorans genome contained genes from both Acidithiobacillus ferrooxidans and Acidithiobacillus caldus encoding enzymes whose assigned functions are redundant. Transcriptional analysis revealed that the petA1 and petB1 genes (implicated in ferrous iron oxidation) were downregulated upon growth on the inorganic sulfur compound tetrathionate but were on average 10.5-fold upregulated in the presence of ferrous iron. In contrast, expression of cyoB1 (involved in inorganic sulfur compound oxidation) was decreased 6.6-fold upon growth on ferrous iron alone. Competition assays between ferrous iron and tetrathionate with Acidithiobacillus ferrivorans SS3 precultured on chalcopyrite mineral showed a preference for ferrous iron oxidation over tetrathionate oxidation. Also, pure and mixed cultures of psychrotolerant acidophiles were utilized for the bioleaching of metal sulfide minerals in stirred tank reactors at 5 and 25°C in order to investigate the fate of ferrous iron and inorganic sulfur compounds. Solid sulfur accumulated in bioleaching cultures growing on a chalcopyrite concentrate. Sulfur accumulation halted mineral solubilization, but sulfur was oxidized after metal release had ceased. The data indicated that ferrous iron was preferentially oxidized during growth on chalcopyrite, a finding with important implications for biomining in cold environments.

INTRODUCTION

Biomining is the industrial application of acidophilic Fe²⁺- and inorganic sulfur compound (ISC)-oxidizing microorganisms to facilitate metal recovery from sulfide minerals. Industrial biomining has, for example, been developed for a number of metals, such as nickel, zinc, cobalt, and copper, at the Talvivaara mine in Finland, as well as for gold at the Fairview mine in South Africa. Metal recovery (solubilization) is achieved through bacterial regeneration of Fe³⁺ (equation 1). Fe³⁺ oxidizes the metal sulfide bond (equation 2, using chalcopyrite as an example) and forms a catalytic cycle between biotic and abiotic steps (reviewed by Sand et al. [1]). The initial sulfur product produced by oxidation of the metal sulfide bond is thiosulfate (S₂O₃²⁻) for acid-insoluble metal sulfides, such as pyrite (FeS₂), or S⁰ for acid-soluble metal sulfides, such as chalcopyrite (CuFeS₂) (2). The intermediary ISC generated from chalcopyrite oxidation is oxidized to sulfuric acid in equation 3 (3).

$$4{\text{Fe}}^{2+}+{\text{O}}_2+4{\text{H}}^{\text{+}}\to 4{\text{Fe}}^{3+}+2{\text{H}}_2\text{O}$$ (1)

$${\text{CuFeS}}_2+4{\text{Fe}}^{3+}\to {\text{Cu}}^{2+}+5{\text{Fe}}^{2+}+2{\text{S}}^0$$ (2)

$$2{\text{S}}^0+3{\text{O}}_2+2{\text{H}}_2\text{O}\to 2{\text{SO}}_4^{2-}+2{\text{H}}^{+}$$ (3)

The production of S⁰ can form a passivating solid layer (also possibly containing precipitates) over the exterior of the sulfide mineral that hinders Fe³⁺ access and, therefore, further mineral oxidation (4). Acidithiobacillus ferrivorans SS3 (originally named Acidithiobacillus ferrooxidans SS3) is the only psychrotolerant acidophile characterized (5) and has been shown to oxidize Fe²⁺ (6) and ISCs at low temperatures (7, 8). It is also capable of catalyzing pyrite dissolution in stirred tank reactors (STRs) at 4°C. All the soluble iron in the STR was found in the Fe³⁺ form, suggesting that microbial Fe²⁺ oxidation occurred efficiently (6). However, the overall rate of metal release was reduced due to the decreased abiotic metal sulfide oxidation rate at low temperatures (6). In addition, a low-temperature mixed culture dominated by A. ferrivorans T7 (originally identified as A. ferrooxidans clone T7) is able to mediate column bioleaching at 7°C (6). Bioleaching of low-grade metal sulfides is often carried out in carefully constructed heaps of crushed or run-of-mine ore (reviewed by Olson et al. [9]). Bioheaps of a multimetal ore from Talvivaara, Finland, showed that bioleaching continues during winter although ambient temperatures reach −30°C, limiting the growth of mesophilic, acidophilic microorganisms in exposed areas of the bioheaps (10). The temperature in the interior of the heap rises as high as 80°C due to exothermic oxidation reactions (11, 12). Thus, during lengthy winters, heap bioleaching in low-temperature environments represents a gradient of temperatures where psychrotolerant A. ferrivorans is suggested to contribute to metal recovery (6).

The genetic determinants and molecular biology of Fe²⁺ and ISC oxidation have been studied extensively for Acidithiobacillus ferrooxidans, a species closely related to A. ferrivorans SS3 (reference 13 and references therein). A. ferrooxidans genes upregulated during growth on Fe²⁺ are grouped into three operons. The first, the rus operon, encodes a cytochrome oxidase, rusticyanin, and two cytochromes that transport electrons from Fe²⁺ to O₂. The second, the petI operon, encodes the bc₁ complex, and the third operon is responsible for the biogenesis of the cytochrome c oxidase complex. A different set of genes is upregulated during growth on elemental sulfur (S⁰) (13). These are the genes coding for the expression of ubiquinol-cytochrome c reductase (petII), the heterodisulfide reductase (hdr) complex, sulfide-quinone reductase (sqr), cytochrome bd and bo₃ ubiquinol oxidases, a thiosulfate-quinone oxidoreductase (TQO) complex (doxDA), and tetrathionate hydrolase (tetH). In contrast to sulfur oxidation by A. ferrooxidans, Acidithiobacillus caldus oxidation of S⁰ is suggested to be mediated by sulfur oxygenase reductase (Sor) and thiosulfate oxidation by the periplasmic thiosulfate-oxidizing multienzyme system (core TOMES) (14). A. ferrooxidans^T cultured in a mixed Fe²⁺-S⁰ medium immediately transcribed Fe²⁺-associated genes, whereas S⁰-related genes were first transcribed when all the Fe²⁺ was oxidized (15). RegAB, a redox-responsive two-component system of A. ferrooxidans^T, downregulates genes encoding S⁰ oxidation proteins in the presence of Fe²⁺ (15).

The A. ferrivorans SS3 genome sequence contains the rus and petI operons, associated with Fe²⁺ oxidation, and genes for the ISC-metabolizing enzymes tetrathionate hydrolase (tetH), sulfide quinone reductase (sqr), thiosulfate-quinone oxidoreductase (doxDA), sulfur oxidation complex SOX (soxYZ-hypB), and Sor (16). However, the role of these genes, suggested to be associated with Fe²⁺ and ISC oxidation, has not been experimentally confirmed.

Because Fe²⁺ and ISC oxidation by A. ferrivorans is vital for bioleaching at low temperatures, we have carried out a bioinformatic investigation of genes encoding Fe²⁺- and ISC-oxidizing proteins. In addition, we have investigated the expression of selected genes coding for Fe²⁺ and ISC oxidation and their relation to the fate of ISCs in low-temperature bioleaching.

MATERIALS AND METHODS

Microorganisms.

A culture of the psychrotolerant acidophile A. ferrivorans SS3 (17) and a low-temperature mixed culture dominated by A. ferrivorans T7 (6) were used in this study.

Bioinformatic reconstruction of A. ferrivorans SS3 Fe²⁺ and ISC oxidation genes.

Genes described in models for Fe²⁺ and ISC oxidation in A. ferrooxidans (13) and A. caldus (14) were used as references for gene prediction in the A. ferrivorans SS3 genome (16) by BLASTP (http://www.ncbi.nlm.nih.gov). Only genes with significant sequence identity (>50%) and an E value of >0.05 were considered. The A. ferrivorans SS3 genome annotation process is ongoing; therefore, pseudogenes and genes with unknown functions were also included if they were associated with candidate Fe²⁺- and ISC-oxidizing genes in clusters. Phylogenetic analyses were carried out using ClustalW2 (http://www.ebi.ac.uk/Tools/msa/clustalw2/) with UPGMA (unweighted-pair group method using average linkages) clustering and default settings (18, 19).

Chemostat culture, RNA preparation, and Q-PCR.

To avoid growth phase-dependent gene expression, A. ferrivorans SS3 cells for quantitative PCR (Q-PCR) were grown in a substrate-limited continuous culture with gradually changing substrate concentrations. The vessel contained minimal salts medium (MSM) with sterile filtered trace elements (20) adjusted to pH 2.5 or 1.6 with H₂SO₄ for tetrathionate or Fe²⁺, respectively. The vessel containing MSM was sterilized (121°C for 30 min) before supplementation with sterile filtered (0.22-μm Sartorius cellulose acetate filter) tetrathionate (Sigma-Aldrich) or FeSO₄·7H₂O (21) at 5 and 50 mM, respectively. The cultures were incubated with stirring at 22 ± 1°C and were sparged with 300 ml 2% (vol/vol) CO₂-enriched air min⁻¹. The continuous culture was maintained (dilution rate [D], 0.0083 h⁻¹) on tetrathionate, and the feed was then replaced by a mixture of 5 mM tetrathionate and 50 mM Fe²⁺, followed by 50 mM Fe²⁺ alone, before the vessel was finally fed with 100 mM Fe²⁺. The theoretical tetrathionate and Fe²⁺ concentrations in the continuous-culture vessel in the absence of oxidation (i.e., the concentrations if no microorganisms were present) were calculated as described previously (22). Culture samples were analyzed by Q-PCR according to the work of Rzhepishevska et al. (23) by using primers described in Table S1 in the supplemental material, except that RNA was prepared using the Viogene (Sunnyvale, CA) Total RNA Extraction Midiprep system. Portions of the putative petA1, petB1, and cyoB1 genes were amplified and sequenced with primers derived from available sequences from A. ferrooxidans^T. Primers for A. ferrivorans SS3 Q-PCR were designed using Primer3 (http://frodo.wi.mit.edu/). The A. ferrivorans SS3 16S rRNA gene sequence (NCBI GenBank accession no. AY960976) was used as a reference gene in the Q-PCR and did not change over the course of the experiment. Q-PCR was carried out in duplicate except for the final measurement of 100 mM Fe²⁺, which was a single measurement.

Substrate competition assays.

A. ferrivorans SS3 was grown in MSM containing trace elements and a substrate (ferrous iron, tetrathionate, or both) at pH 2.5. Cells were pregrown on either a 50 mM Fe²⁺ or a 5 mM tetrathionate medium and were then inoculated into MSM containing both 50 mM Fe²⁺ and 5 mM tetrathionate in order to determine which substrate was utilized first. Also, cells pregrown on the respective substrates were inoculated into a 4% (wt/vol) chalcopyrite medium (without added trace elements; the mineralogy of the chalcopyrite is described below) and were then further subcultured into a medium containing both Fe²⁺ and tetrathionate. Ferrous iron consumption was monitored by ceric sulfate titration (24). Tetrathionate was analyzed by a modified cyanolysis protocol (25) where the sample was filtered after the addition of phosphate buffer to remove ferric iron precipitates. All samples were diluted 10-fold to ensure proper removal of iron. Experiments were carried out in triplicate, and means ± standard deviations (SD) are presented.

Fe²⁺ and ISC oxidation during bioleaching.

The mineral concentrates (provided by Boliden Mineral AB) were pyrite (FeS₂) from Kristineberg, Sweden, and chalcopyrite (CuFeS₂) from Aitik, Sweden, and have been described previously (6) (see Table S2 in the supplemental material). Leaching was carried out in 1.5-liter STRs containing 4% (wt/vol) mineral concentrates in MSM without trace elements at pH 2 (20). Although the STR and MSM were sterilized (121°C for 20 min) and sampling was carried out aseptically, the sulfide minerals were not sterilized. Where indicated, the STRs were inoculated with a total of 1.3 × 10¹¹ cells (measured in a Coulter counter) of either A. ferrivorans SS3 or the T7 mixed culture. Thymol (0.08% [wt/vol]; Sigma) was added to the uninoculated controls to ensure that no microbial growth occurred. Leaching was carried out according to the work of Dopson et al. (6) and Dopson and Lindström (20) at either 6 ± 1° or 21 ± 1°C. Leaching was described (20) by measuring total metal (Metal_Tot; i.e., metal in solution as well as secondary precipitates redissolved by treatment with 5 M HCl for 30 min at 65°C) and soluble metal (Metal_Sol; i.e., soluble metal without HCl treatment), by atomic adsorption spectroscopy; soluble Fe²⁺ (Fe²⁺_Sol), by titration with ceric sulfate (24); tetrathionate and S⁰, by cyanolysis (25); redox potential (by use of a Pt electrode against an Ag⁰-AgCl electrode with 3.5 M KCl); and pH. The low-temperature experiments were carried out as duplicates, while the room temperature experiments were carried out with 4 replicates. The results are presented as averages for 2 to 4 replicates ± SD.

RESULTS

Bioinformatic reconstruction of A. ferrivorans Fe²⁺ and ISC oxidation pathways.

By using existing models for substrate oxidation in the closely related species A. ferrooxidans^T (13) and A. caldus^T (14), putative genes involved in Fe²⁺ and ISC oxidation in A. ferrivorans SS3 were identified (Fig. 1). Homologues for all known A. ferrooxidans genes involved in Fe²⁺ oxidation were also found in A. ferrivorans (see Table S3 in the supplemental material). However, some differences occur. The rus operon (Acife_1864 to -1872) includes genes coding for a Fe²⁺ oxidase (Cyc2) that passes the electron generated to the transfer proteins Rus and Cyc1, as well as to a cytochrome aa₃ oxidase (Cox) predicted to be involved in the downhill electron pathway to produce ATP. The A. ferrivorans rus operon also contains a gene encoding a small hypothetical protein (Acife_1865), which is present in A. ferrooxidans ATCC 53993 and not in the type strain. Additionally, two further copies of the rus gene (Acife_0470 and -1718) and an iron oxidase gene (iro; Acife_1427) were present in the A. ferrivorans genome. A deeper analysis of the Rus protein sequences revealed that Acife_1864 (present in the rus operon) encoded a RusA-type protein, whereas Acife_0470 and -1718 encoded RusB-type proteins (see Fig. S1 in the supplemental material). Neither rusB nor iro is found in the A. ferrooxidans^T genome (13). The A. ferrivorans rus gene cluster was located immediately upstream of the cta operon (Acife_1859 to -1863) and the two regulatory genes regAB (Acife_1857 and -1858). However, ctaR was not present in the A. ferrivorans cta gene cluster, although candidate genes can be found dispersed elsewhere in the genome (Acife_0172, -1670, and -1098). The A. ferrooxidans genome also includes the petI gene cluster (Acife_0613 to -0617), predicted to be involved in the uphill electron pathway to produce NADH at the expense of energy. This gene cluster encodes a c₄-type cytochrome (CycA1), ubiquinol-cytochrome bc₁ reductase (PetA1B1C1), and the Sdr1 oxidoreductase.

Fig 1.

Gene clusters of enzymes potentially involved in Fe²⁺ (A) and ISC (B) metabolism, as well as gene clusters of the NADH quinone oxidoreductase and the ATP synthase complex (C), as predicted by bioinformatic analysis of the A. ferrivorans genome sequence. Abbreviations: hyp, hypothetical protein; pseudo, pseudogene; Acife_1383, the A. ferrivorans genome locus tag for a periplasmic solute-binding protein; and Acife_2603, Tat (twin-arginine translocation) pathway signal sequence domain-containing protein.

The identification of putative A. ferrivorans genes involved in ISC oxidation revealed a genomic configuration distinct from those in A. ferrooxidans and A. caldus (see Tables S4 and S5 in the supplemental material). A. ferrivorans contained two copies of the gene encoding sulfide-quinone reductase (sqr1 [Acife_2028] and sqr2 [Acife_2600]), which oxidizes sulfide to S⁰. Phylogenetic analysis of the Sqr protein sequences showed that Sqr1 belongs to type I, whereas Sqr2 belongs to type V (data not shown) (26). The A. ferrivorans genome included a hdr gene cluster (Acife_2519 and Acife_2471 to -2479) similar to that of A. ferrooxidans, which, among other genes, encodes heterodisulfide reductase, predicted to oxidize S⁰, as well as putative sulfur transfer proteins, such as DsrE (13). In addition, there was a sulfur oxygenase reductase (Sor; Acife_2601) similar to that found in A. caldus (14). It was also noted that a putative thiol:disulfide interchange protein (Acife_1847) and several rhodanase-like sulfur transferases (Acife_1630, -0967, -2415, -2589, and -0854) were present. The A. ferrivorans genome also contains two genes coding for tetrathionate hydrolase (tetH1 [Acife_0037] and tetH2 [Acife_2602]). Phylogenetic analysis of TetH protein sequences showed that TetH1 grouped with previously described tetrathionate-hydrolyzing enzymes, while TetH2 grouped with previously uncharacterized TetH sequences and Acidianus ambivalens TetH2 (data not shown). A. ferrivorans contained two systems potentially coding for thiosulfate oxidation: doxDA (Acife_1384), coding for thiosulfate-quinone oxidoreductase, and a partial sox gene cluster (Acife_2489 to -2494) that was similar to the A. caldus soxII gene cluster, except that soxA and soxX were annotated as pseudogenes in A. ferrivorans. The only A. ferrivorans gene suggested to encode part of the SO₃²⁻ oxidation system was sat (Acife_2287). A five-gene cluster including a sulfite reductase (Acife_2625 to -2626) was found. However, these genes share no sequence or structural homology with genes encoding dissimilatory sulfite reductases, suggesting that they are not involved in sulfite oxidation. Electron transport systems involved in energy conservation during ISC oxidation included (i) a cyo gene cluster (Acife_1015 to -1020) encoding a cytochrome bo₃ oxidase (CyoA1B1C1D1E1), (ii) a second cyo gene cluster (Acife_2597 to -2594), containing a pseudogene instead of cyoD, (iii) a third cyo gene cluster (Acife_1028 to -1031), containing a hypothetical protein (Acife_1032) and a putative thiol:disulfide interchange protein (Acife_1033), (iv) the petII gene cluster (Acife_3266 to -3271), which includes a further bc₁ complex (PetA2B2C2), (v) the NADH complex cluster (Acife_2541 to -2554), and (vi) an ATP synthase complex cluster (Acife_3190 to -3196). Both A. ferrooxidans and A. caldus contain bd-type cytochromes suggested to be involved in ISC metabolism, although no candidate genes were identified in A. ferrivorans. Finally, it was noted that ISC oxidation-related gene clusters seem to have an accumulation of pseudogenes, 13% compared to 6% in the overall genome.

Expression of genes coding for Fe²⁺ and ISC oxidation.

A. ferrivorans SS3 cells sampled in substrate-limited continuous cultures were analyzed by Q-PCR for expression of petA1 (Acife_0615) and petB1 (Acife_0614), genes associated with Fe²⁺ oxidation, and cyoB1 (Acife_1016), associated with ISC oxidation (13, 27). cyoB1 expression was slightly greater than petA1 and petB1 expression during cultivation on tetrathionate alone, although a basal level of petA1 and petB1 gene expression was observed (Fig. 2). A decrease in cyoB1 expression and increases in petA1 and petB1 gene expression were observed when the substrate was a mixture of tetrathionate and Fe²⁺ (Fig. 2). Expression of cyoB1 on Fe²⁺ as a single substrate decreased 6.6-fold relative to expression on tetrathionate, while petA1 and petB1 expression increased 10.6- and 10.4-fold, respectively (Fig. 2). The results suggest that the expression of petA1, petB1, and cyoB1 depends on the presence of Fe²⁺ or tetrathionate, although none of the genes were totally inhibited in the presence of both substrates. Considering that the concentration of tetrathionate was maintained at the same level in the first three samples for which results are shown in Fig. 2, the expression of petA1 and petB1 was likely to be induced by the presence of Fe²⁺. Whether changes in the cyoB1 level were caused by induction with tetrathionate or by inhibition by Fe²⁺ cannot be determined.

Fig 2.

Expression of A. ferrivorans SS3 genes cyoB1 (filled bars), petA1 (open bars), and petB1 (shaded bars) (right-hand y axis) in tetrathionate- and Fe²⁺-limited continuous culture. Relative gene expression was calculated by comparison to the expression of the rrs gene, which was defined as 1.0 (23). (rrs gene expression was consistent during the course of the experiment.) Lines represent the theoretical maximum tetrathionate (solid) and Fe²⁺ (dashed) concentrations (left-hand y axis) in the continuous culture if no substrate oxidation occurred (see Materials and Methods for details).

Substrate competition assays.

A. ferrivorans SS3 was incubated with Fe²⁺ and tetrathionate in order to elucidate which substrate was preferentially oxidized. Whether Fe²⁺ or tetrathionate was consumed first depended on the substrate on which the cells had been pregrown (see Fig. S2 in the supplemental material). However, if the cells were pregrown on either Fe²⁺ or tetrathionate, subcultured once in a chalcopyrite medium (where both substrates are available) until the mid-exponential-growth phase, and finally inoculated into a mixed Fe²⁺- and tetrathionate-containing medium, then Fe²⁺ was always preferentially oxidized (Fig. 3).

Fig 3.

Consumption of tetrathionate and Fe²⁺ by A. ferrivorans SS3 in a mixed-substrate medium containing both 50 mM Fe²⁺ and 5 mM tetrathionate. The inoculum was preincubated with cells grown on either Fe²⁺ (A) or tetrathionate (B) and was subcultured once in a 4% (wt/vol) chalcopyrite mineral medium before being inoculated into a mixed-substrate medium. Symbols: ◆, Fe²⁺; ■, tetrathionate. Data points are averages ± SD for three replicates.

ISC oxidation during bioleaching.

The leaching rates in uninoculated controls were significantly lower than those in inoculated STRs (see Table S6 in the supplemental material). In the absence of microorganisms, the Fe²⁺_Sol concentration increased to 16.3 ± 6.2 and 7.3 ± 1.8 mM Fe²⁺ for the pyrite and chalcopyrite concentrates, respectively, after 108 days, suggesting that Fe²⁺ oxidation did not occur. This resulted in significantly lower Fe_Tot concentrations in the uninoculated controls than in the inoculated STRs after 108 days of leaching (see Table S6 in the supplemental material). In all inoculated STRs, the Fe²⁺_Sol concentration was ≤1.0 ± 0.1 mM after 108 days of leaching. The low Fe²⁺_Sol concentration (suggesting that all the available Fe was in the Fe³⁺ form) was reflected in the redox potential, which increased to ≥584 ± 14 mV for all inoculated STRs, and a higher metal leaching rate (Fig. 4; see also Table S6 and Fig. S3 in the supplemental material).

Fig 4.

Leaching of pyrite (A) and chalcopyrite (B) concentrates by A. ferrivorans SS3. The figure shows the leaching of metals and the redox potential (left) and ISCs (right). Symbols: ■, Fe_Tot; ▲, Cu_Tot; ●, redox potential (versus Ag-AgCl electrode); ◆, tetrathionate; □, S⁰. Data points are averages ± SD for two replicates.

Pyrite is leached via the thiosulfate mechanism (2), and therefore, the concentration of solid S⁰ produced was ≤1.3 ± 0.4 mM in the presence of either A. ferrivorans SS3 or the T7 mixed culture (Fig. 4; see also Table S6 and Fig. S3 in the supplemental material). In contrast, chalcopyrite bioleaching follows the polysulfide mechanism (2), and S⁰ increased to levels higher than those for pyrite bioleaching. For chalcopyrite bioleaching by A. ferrivorans SS3, the increased S⁰ concentration reached a maximum of 9.7 ± 1.3 mM. However, during chalcopyrite bioleaching catalyzed by the T7 mixed culture, the S⁰ concentration increased rapidly to 12.3 ± 1.1 mM after 21.0 days and remained at 12.0 ± 0.4 mM between 21 and 62 days (5 sampling points) before decreasing again (see Fig. S3 in the supplemental material). The Fe_Tot and Cu_Tot concentrations increased continuously until 62 days, reaching 78 ± 11 and 30 ± 7 mM (corresponding to 38% and 16% of the available Fe and Cu, respectively). The available S⁰ was consumed only once the Fe and Cu concentrations no longer increased (after 62 days). The fact that the S⁰ concentration increased until metal solubilization halted suggested that S⁰ passivation (passivation reviewed in reference 4) of the chalcopyrite occurred and that S⁰ was consumed only once no Fe²⁺ was available. Finally, once the S⁰ concentration decreased, bioleaching continued (see Fig. S3 in the supplemental material). Similar trends were observed with an accumulation of S⁰ during chalcopyrite bioleaching by A. ferrivorans SS3 at room temperature: almost six times as much sulfur was accumulated with chalcopyrite as with pyrite (see Fig. S4 in the supplemental material).

DISCUSSION

The ability to switch between substrates (Fe²⁺ and ISC) is important for the ecological flexibility of A. ferrivorans SS3 and hence for its applications in biomining (28) and the bioremediation of wastewaters from mineral processing (8). Therefore, we investigated genetic determinants of Fe²⁺ and ISC oxidation in A. ferrivorans SS3, its growth on different substrates, and the implications for bioleaching.

All bioinformatic comparisons of A. ferrooxidans^T genes suggested to encode Fe²⁺ oxidation enzymes (13) identified homologues in the A. ferrivorans SS3 genome (16). However, the genome of A. ferrivorans SS3 also reveals important differences: (i) the presence of iro, a gene encoding a high-potential iron-sulfur protein previously reported to be the primary Fe²⁺-oxidizing enzyme in A. ferrooxidans JCM 7811 (29), (ii) the presence in the rus gene cluster of a hypothetical protein that is absent in A. ferrooxidans^T, (iii) the absence of ctaR in the A. ferrivorans SS3 cta gene cluster, and (iv) the presence of genes encoding two b-type rusticyanins in the genome. These differences, taken together, indicate that A. ferrivorans might use a Fe²⁺ oxidation pathway similar, but not identical, to that of A. ferrooxidans^T. The pathway for A. ferrivorans ISC oxidation remains obscure, since the genome contained genes encoding predicted ISC-oxidizing proteins from both A. ferrooxidans and A. caldus (13, 14, 30). This apparent functional redundancy includes the two copies of both sqr and tetH; the Sox complex and DoxDA (TQO), both of which are reported to oxidize thiosulfate in other microorganisms (reviewed in reference 31); and Hdr and Sor, which are predicted to oxidize S⁰ (13, 14). Type I Sqr has been functionally and structurally characterized in several species, including A. ferrooxidans (32), while no bacterial type V Sqr has been shown to oxidize sulfide. Similarly, TetH1 grouped with previously described tetrathionate-hydrolyzing enzymes, and TetH2 grouped with Acidianus ambivalens TetH2, previously shown not to have tetrathionate-hydrolyzing activity (33). These analyses suggested that sqr1 and tetH1 were more likely to be involved in ISC metabolism. Also, the sox gene cluster is incomplete (containing pseudogenes) and possibly nonfunctional, suggesting that thiosulfate oxidation may be mediated by the DoxDA system, as in A. ferrooxidans (13). However, more work is needed to fully elucidate the A. ferrivorans SS3 ISC metabolic pathway.

Fe²⁺- and S⁰-induced gene expression by A. ferrivorans SS3 was further investigated by using cyoB1, coding for a terminal quinol oxidase subunit proposed to reduce O₂ during ISC oxidation (34), and petA1B1, coding for bc₁ complex proteins involved in uphill electron transfer during cultivation on Fe²⁺ (27, 35). Expression of petA1B1 during cultivation of A. ferrivorans SS3 indicated that they were induced by the presence of Fe²⁺. This is consistent with previous studies showing petA1B1 upregulation on Fe²⁺ as a sole energy source (34, 35). cyoB1 expression levels decreased when Fe²⁺ was introduced to the medium, indicating downregulation of ISC genes. Further, Fe²⁺ was preferred in a mixed Fe²⁺–tetrathionate medium inoculated with cells pregrown on chalcopyrite. These findings are consistent with available information on A. ferrooxidans^T showing utilization of Fe²⁺ in preference to S⁰ (15, 27) and downregulation of S⁰ oxidation genes in the presence of Fe²⁺ (13), and they support the view that Fe²⁺ is the preferred substrate among Fe²⁺-oxidizing acidithiobacilli (15). However, reports of the preferred substrate for A. ferrooxidans are conflicting, since S⁰ was oxidized by A. ferrooxidans FC1 on the surface of zinc sulfide in the presence of Fe²⁺ concentrations ranging from 1 to 9.5 g/liter (36). The discrepancies may be due to differences between strains in the A. ferrooxidans species.

Oxidation of Fe²⁺ and ISCs at low temperatures by cold-tolerant microorganisms such as A. ferrivorans SS3 (7, 8) is important for heap bioleaching in cold environments. Despite the temporally low temperatures during winter, oxidation can proceed, aiding heat generation within the heap to sustain the activity of mesophilic and thermophilic microorganisms (11). The initial ISC compound during chalcopyrite bioleaching is polysulfide sulfur, and the accumulation of solid S⁰ in cultures grown on the chalcopyrite concentrate was in agreement with this mechanism (2). Maximum S⁰ accumulation in both STRs (one containing A. ferrivorans SS3 and one containing the T7 mixed culture) on a chalcopyrite concentrate occurred when the metal solubilization rates were decreased and before all the metal was released (Fig. 4; see also Fig. S3 in the supplemental material). This phenomenon was especially prominent in T7 mixed cultures and suggests that S⁰ passivation may have taken place (see Fig. S3 in the supplemental material). The S⁰ concentration at which passivation occurred was less than the approximately 30 mM concentration observed for moderately thermophilic bioleaching of arsenopyrite (20). S⁰ accumulation during bioleaching supports the notion that Fe²⁺ is preferentially oxidized by A. ferrivorans SS3 during chalcopyrite bioleaching. This would explain the sequence of events during chalcopyrite bioleaching in that S⁰ passivates the mineral surface (36), resulting in a lack of Fe²⁺ such that the bacteria then oxidize S⁰ because it is the only accessible substrate.