Biodiversity, metabolism and applications of acidophilic sulfur-metabolizing microorganisms

Authors


E-mail bss041@bangor.ac.uk; Tel. (+44) 1248 382358; Fax (+44) 1238 370731.

Summary

Extremely acidic, sulfur-rich environments can be natural, such as solfatara fields in geothermal and volcanic areas, or anthropogenic, such as acid mine drainage waters. Many species of acidophilic bacteria and archaea are known to be involved in redox transformations of sulfur, using elemental sulfur and inorganic sulfur compounds as electron donors or acceptors in reactions involving between one and eight electrons. This minireview describes the nature and origins of acidic, sulfur-rich environments, the biodiversity of sulfur-metabolizing acidophiles, and how sulfur is metabolized and assimilated by acidophiles under aerobic and anaerobic conditions. Finally, existing and developing technologies that harness the abilities of sulfur-oxidizing and sulfate-reducing acidophiles to extract and capture metals, and to remediate sulfur-polluted waste waters are outlined.

Origin and characteristics of acidic, sulfur-rich environments

Sulfur is the 16th most abundant element in the lithosphere, with an average abundance (by weight) of about 0.05%. However, it is highly concentrated in parts of the crust, such as in metal sulfide ore deposits and coals. It can exist in nine different oxidation states, ranging from −2 (e.g. in hydrogen sulfide) to +6 (in sulfate) (Steudel, 2000). Many acidophilic species of bacteria and archaea are able to utilize this group VI element as a source or a sink of electrons in dissimilatory metabolic processes, catalysing redox transformations that involve between one and eight electrons. Figure 1 shows the various transformations of sulfur that occur in extremely acidic environments, and indicates redox reactions that are catalysed by microorganisms and those that are abiotic.

Figure 1.

Transformations of sulfur in extremely acidic environments. The various oxidation states of sulfur in the compounds and minerals listed are shown in bold text and highlighted in grey (sulfur atoms in polythionates can have different oxidation states). Reactions catalysed by acidophilic prokaryotes are indicated. Key: aSOP, acidophilic sulfur-oxidizing prokaryotes; aSRP, acidophilic sulfur-reducing prokaryotes; aSRB, acidophilic sulfate-reducing bacteria (no archaea are known to mediate dissimilatory sulfate reduction at low pH). *although the oxidation states of sulfur atoms in thiosulfate are often considered to be −2 and +6, XANES spectroscopic analysis has indicated that the two atoms have charge densities corresponding to oxidation states of −1 and +5 (Vairavamurthy et al., 1993).

Sulfur-rich environments can be acidic, sometimes extremely so (pH < 3), and may be natural features or anthropogenic in origin. The most widespread natural sulfur-rich environments, termed ‘solfatara’, are terrestrial geothermal sites that occur in the vicinities of active volcanoes and where the earth's crust is relatively thin (e.g. Yellowstone National Park, Wyoming). Sulfur is present in magmas as water-soluble-S and sulfide minerals (Métrich and Mandeville, 2010), and also occurs in volcanic gases. Crystals of prismatic sulfur, formed by the condensation of sulfur dioxide and hydrogen sulfide (Eq. 1 and Fig. S1) can occasionally be observed around the margins of fumaroles.

image(1)

Oxidation of elemental sulfur (S°; and other reduced forms of S) by acidophilic bacteria and archaea generates sulfuric acid (Eq. 2):

image(2)

Water temperatures in solfatara fields approach boiling point (∼ 85–100°C, depending on altitude) but tend to cool rapidly as the water flows from the source of the geothermal spring. Solfatara may therefore be colonized by a variety of acidophilic microorganisms that have different temperature optima. Besides Yellowstone National Park (USA), solfatara are found in Whakarewarewa (New Zealand); Krisuvik (Iceland); the Kamchatka peninsula (Russia); Sao Michel (Azores); Volcano, Naples, and Ischia (all Italy); Djibouti (Africa); and some Caribbean islands, such as Montserrat and St. Lucia. Related to these are deep and abyssal submarine hydrothermal systems, such as the Mid-Atlantic ridge; the East Pacific Rise; the Guaymas Basin; and active seamounts (e.g. around Tahiti). Because of the high pH buffering capacity of sea water, extremely low pH zones tend to be limited in scale in and around submarine hydrothermal vents, although one moderately acidophilic (pH 3.3–5.8)extremely thermophilic sulfur-reducing archaeon has been isolated, which appears to be widely distributed in hydrothermal vents (Reysenbach et al., 2006).

Biogenic sulfuric acid is also responsible for extremely acidic sulfur-rich environments that are unconnected with geothermal activity. Some subterranean environments contain sulfide-rich ground waters, and when these are in contact with oxygen, microbial oxidation of sulfide to sulfuric acid can lead to extensive dissolution of carbonate rock strata and the development of extensive cave systems, e.g. the Lechugilla cave in New Mexico. In situations where the acidity is not neutralized by basic minerals, areas of extreme acidity may develop. The most well-studied site of this kind is the Frasassi cave complex, Italy (Fig. S2). Here, viscous biofilms (‘snottites’) in water films of pH 0–1 have been described (Vlasceanu et al., 2000; Macalady et al., 2007). Community genome analysis of these snottites has shown that they have a limited biodiversity in terms of component microorganisms. The mesophilic sulfur-oxidizing bacterium Acidithiobacillus (At.) thiooxidans was identified as the most abundant (∼ 70% of cells) primary producer and ‘architect’ of the snottites. Actinobacteria related to Acidimicrobium ferrooxidans, and archaea related to the uncultivated ‘G-plasma’ clade of the Thermoplasmatales (at 5% and 15% of cells, respectively) have also been detected in Frasassi snottites (Jones et al., 2012).

A situation analogous to that in the Frasassi cave system has long been recognized as a major problem in many urban conurbations. Underground sewage systems transport water wastes that are enriched with (sulfur-containing) organic materials, and varying concentrations of dissolved sulfate. Dissimilatory sulfate-reduction and desulfurylation (hydrolysis of thiols, such as cysteine) results in the production of gaseous H2S, which migrates from the waste water to the overlying air-space, and condenses on the walls of the pipes through which the sewage is transported. Sulfide-oxidizing bacteria housed in biofilms in the upper parts of the pipes oxidize the hydrogen sulfide, ultimately to sulfuric acid (Fig. 2). In cases where the sewage pipes are constructed of concrete, acid erosion can cause weakening and ultimately collapse of the sewage system (Sand, 1987). As with Frasassi, At. thiooxidans was identified as the most abundant bacterium in a heavily corroded concrete sample in a sewage system in Japan (Okabe et al., 2007).

Figure 2.

Cycling of sulfur in an urban sewage system. Anaerobic conditions in the organic carbon-rich effluents promote the formation of hydrogen sulfide via dissimilatory sulfate reduction and desulfurylation. Hydrogen sulfide gas diffuses into the overlying air layer and microorganisms embedded in biofilms in the upper surfaces of the sewage pipes oxidize it to sulfuric acid. In situations where the sewage pipes are constructed of concrete, this can lead to severe corrosion and, in the worst scenarios, collapse of the underground sewage network.

The most ubiquitous and widely studied low pH, sulfur-rich environments are man-made, and are associated with the mining of metals and coals. Many metals of commercial value occur as, or are intimately associated with, sulfide minerals. In addition, coal deposits may contain up to 20% (by weight) of sulfur. Solid waste materials (e.g. waste rocks and mineral tailings (Fig. S3) from metal mines and mineral processing) frequently contain significant concentrations of sulfides, such as pyrite (FeS2). Pyrite is the most abundant of all sulfide minerals but is not generally considered to have any commercial value. Biologically accelerated oxidative dissolution of sulfides in rock piles and tailings generates acidity, chiefly via the oxidation of the sulfide moiety in these minerals, although hydrolysis of the ferric iron generated in the process is also a proton-generating reaction. Equation 3 illustrates the transformation of a reduced sulfide mineral (pyrite) to an oxidized oxy-hydroxide mineral (schwertmannite):

image(3)

Water percolating through waste dumps becomes acidic, except when there are sufficient basic minerals to counterbalance the generated acidity, and enriched with soluble metals (metal cations are generally more soluble in acidic than in circum-neutral pH liquors) and metalloids such as arsenic. Streams flowing from the wastes are known generically as ‘acid mine drainage’ (AMD; Fig. S4) and represent a major environmental hazard, capable of causing serious pollution to sites distantly downstream of the point of discharge. The reactions involved in AMD formation are essentially the same as those in ‘biomining’ technologies (discussed below), the major difference being that in the case of AMD they operate in an uncontrolled context, and frequently over an extended time frame (for many hundreds of years). The microflora of oxidizing mine wastes and acidic mine waters tend to be different to those of solfatara and sulfidic caves, in that iron-oxidizing bacteria are often the most numerous chemolithotrophic prokaryotes. However, some these species (such as At. ferrooxidans) can oxidize sulfur (anaerobically as well as aerobically) as well as ferrous iron. Detailed accounts of the microbiology of mine wastes and AMD can be found in reviews by Baker and Banfield (2003); Hallberg (2010); and Schippers and colleagues (2010).

Stockpiling of S° can also inadvertently generate extremely acidic liquors. Sulfur is generated as a by-product by the gas and oil industries and the current scale of S production exceeds demand (Crescenzi et al., 2006). Sulfur is stockpiled in large heaps (∼ 20 m high), which are often stored outside for economic reasons. Exposure to water and oxygen stimulates the growth of sulfur-oxidizing acidithiobacilli and other sulfur-oxidizers, again causing sulfuric acid to be generated. Transportation of such ‘bio-activated’ sulfur carries the risk of metal corrosion to, for example, the holds of ships.

Diversity of sulfur-metabolizing acidophiles

The first acidophile to be described was the sulfur-oxidizing proteobacterium At. thiooxidans (named at the time as Thiobacillus thiooxidans; Waksman and Joffe, 1922). Several decades later, the first sulfur-metabolizing archaea (although not referred to as such at the time) were isolated from geothermal sites in Yellowstone by James Brierley, and also by Tom Brock and co-workers. Brierley's isolate, obtained in 1965, was subsequently named as Acidianus (Ac.) brierleyi (Brierley, 2008) while Brock's team isolated the first Sulfolobus spp., although the organism named at the time as Sulfolobus (S.) acidocaldarius was later found not to oxidize S° when grown in pure culture. Since then, a large number of phylogenetically diverse bacteria and archaea that can mediate dissimilatory redox transformations of sulfur in acidic liquors have been described, although far more are known to catalyse oxidative than reductive reactions. Table 1 lists species of extremely acidophilic bacteria and archaea (defined as having pH optima for growth of ≤ 3) and excludes ‘moderate acidophiles’, such as Thiomonas spp., which also oxidize S° and reduced inorganic sulfur compounds (RISCs; thiosulfate, tetrathionate etc.) but which are generally not found in extremely acidic environments.

Table 1. Extremely acidophilic prokaryotes that catalyse dissimilatory transformations of sulfur.
 S0 oxidationS0 reductionT responseNotes
  • a. 

    S. Hedrich and D.B. Johnson, unpubl. data.

  • b. 

    Demonstrated in one strain only.

  • c. 

    Psychotolerant.

  • Genus abbreviations: A., Acidiphilium; Ac., Acidianus; Af. Acidiferrobacter; At. Acidithiobacillus; H., Hydrogenobaculum; S., Sulfolobus; Sg., Stigiolobus; Ss., Sulfurisphaera. Temperature response abbreviations: M, mesophilic; MT moderately thermophilic; ET extremely thermophilic.

Proteobacteria     
 At. thiooxidans+MObligate aerobe/autotroph
 At. albertensis+MForms glycocalyx; otherwise very similar to At. thiooxidans
 At. caldus+MTSome strains also use H2 as e-donora
 At. ferrooxidans++bMAlso uses Fe2+ & H2 as e-donors, and O2 or Fe3+ as e-acceptors
 At. ferrivorans+McAlso uses Fe2+ as e-donor
    O2 or Fe3+ as e-acceptor
 Af. thiooxydans+MAlso uses Fe2+ as e-donor
    O2 or Fe3+ as e-acceptor
 A. acidophilum+MFacultative chemoautotroph
 Other Acidiphilium spp.+MObligate heterotrophs; aerobic/microaerophilic
 Acd. organivorans+MTO2 or Fe3+ as e-acceptor
Firmicutes     
 Sulfobacillus spp.+M & MTAlso uses Fe2+ as e-donor
    Facultative chemoautotrophs
    O2 or Fe3+ as e-acceptor
 Alicyclobacillus spp.+MTNot all species oxidize S
Aquificae     
 H. acidophilum+ETAlso use H2 as e-donor [and one isolate uses As(III)]
Euryarchaeota     
 Thermoplasma spp.+MTObligate heterotrophs
Crenarchaeota     
 S.metallicus+ETAlso uses Fe2+ as e-donor
 S. tokodaii+ETAlso uses Fe2+ as e-donor
 Acidianus spp.++ET Ac. brierleyi can also use Fe2+ as e-donor
 Metallosphaera spp.+ETFacultative autotrophs
 Sulfurococcus spp.+ Facultative autotrophs
 Sg. azoricus+ETObligate anaerobe
 Ss. ohwakuensis+ETFacultative anaerobe

One of the key traits often used to differentiate acidophilic prokaryotes is their response to temperature. Mesophilic species are generally regarded as those with temperature optima between 20°C and 40°C, and all sulfur-metabolizing species are, at the time of writing, exclusively bacteria. Moderate thermophiles have temperature optima between 40°C and 60°C and, with the current sole exception of the sulfur-reducing euryarchaeote Thermoplasma, are also bacteria. In contrast, all known sulfur-metabolizing extreme thermophiles (temperature optima of > 60°C) are crenarchaeotes, apart from the sulfur-oxidizing autotrophic bacterium Hydrogenobaculum acidophilum. No psychrophilic sulfur-oxidizing acidophiles have been described, although the sulfur- and iron-oxidizing proteobacterium At. ferrivorans is psychro-tolerant, and grows at 4°C to 35°C (Kupka et al., 2009; Hallberg et al., 2010).

Most of the prokaryotes listed in Table 1 either oxidize or reduce sulfur. However, the three currently recognized species of the crenarchaeote genus Acidianus can both oxidize and reduce sulfur, depending on the availability of oxygen and the provision of a suitable electron donor (for sulfur reduction). Intriguingly, Ohmura and colleagues (2002) reported that one strain of the most widely studied of all acidophiles (the proteobacterium At. ferrooxidans) can, like Acidianus spp., couple the oxidation of H2 to the reduction of S°.

Sulfur-metabolizing acidophiles can also be differentiated in terms of carbon assimilation, as obligate autotrophs (e.g. At. ferrooxidans), heterotrophs (e.g. Acidicaldus organivorans), or facultative autotrophs/heterotrophs [e.g. Sulfobacillus spp. and Acidiphilium (A.) acidophilum] that preferentially use organic carbon but can also fix carbon dioxide. While many sulfur-oxidizing acidophiles appear to be obligate aerobes (e.g. At. thiooxidans and At. caldus) others (e.g. At. ferrooxidans, At. ferrivorans and Acidiferrobacter thiooxydans) can use ferric iron as an alternative electron acceptor to oxygen and grow in anoxic environments.

In many cases, sulfur-oxidizing acidophiles can use other electron donors in addition to reduced forms of sulfur. Many chemolithotrophic sulfur-oxidizing bacteria use ferrous iron, and/or hydrogen as electron donors, and heterotrophic species use organic electron donors. Curiously, the very first species that was isolated, At. thiooxidans, is one of the most specialized of all known sulfur-metabolizing acidophiles, in being an obligate autotroph that grows only by coupling the oxidation of reduced sulfur to the reduction of molecular oxygen (Johnson and Hallberg, 2009).

Table 1 does not include acidophilic sulfate-reducing prokaryotes (alluded to in Fig. 1) as there are currently no validated species known to have this trait. However, sulfate-reducing bacteria have been isolated from extremely acidic mine waters [e.g. Rowe et al., (2007)] and have also been identified in low pH (2–4) sulfidogenic bioreactors (Ňancucheo and Johnson, 2012). All ofthese are Firmicutes, identified as either species of a novel genus (‘Desulfobacillus’) or new species of Desulfosporosinus.

In both natural and anthropogenic environments, sulfur-metabolizing prokaryotes live alongside, and interact with, other acidophiles. In some cases, mutualistic interactions involving sulfur bacteria have been described. For example, neither the sulfur-oxidizing autotroph At. thiooxidans nor the iron-oxidizing heterotroph Ferrimicrobium (Fm.) acidophilum can oxidize pyrite when grown in pure culture, but can do so when grown in co-culture. Ferric iron, which is required for the abiotic oxidation of pyrite, is generated via oxidation of ferrous iron by Fm. acidophilum, which uses organic carbon (as C-source) provided by At. thiooxidans. RISCs produced from ferric iron attack on pyrite (Fig. 1) act as electron donors for At. thiooxidans, fuelling CO2 fixation and ultimately leakage of organic carbon (Bacelar-Nicolau and Johnson, 1999).

Sulfur metabolism in acidophilic prokaryotes

Aerobic RISC metabolizing enzymes and pathways

Oxidation of reduced sulfur can provide up to eight electrons (in the case of oxidation of H2S to SO42−) for energy conservation, compared with a single electron from Fe2+ oxidation (which is the other major electron donor available in many extremely acidic environments). Additionally, electrons derived from RISC oxidation enter electron transport at a lower potential resulting in more ATP being produced per electron mol than for Fe2+ oxidation (Kelly, 1999). On first glance, acidophiles appear to have an extremely large proton gradient from the external to internal milieu that may be exploited for ATP synthesis (Slonczewski et al., 2009). However, without coupling metabolic proton export (via electron transport components) to ATP production, this would rapidly acidify the cytoplasm and result in cell death.

The biochemistry of the individual enzymes involved in acidophile RISC oxidation has been described in recent review articles (Rohwerder and Sand, 2007; Johnson and Hallberg, 2009). Biochemical and inhibitor analyses of RISC metabolism in Gram-negative bacteria identified intermediates from which a general pathway for At. ferrooxidans (Pronk et al., 1990), At. caldus (Hallberg et al., 1996) and A. acidophilum (Meulenberg et al., 1992) has been proposed (Johnson and Hallberg, 2009). The cellular localization of the enzymes in this model is not the same between all acidophiles as sulfur-oxidizing enzymes in the marine At. thiooxidans strain SH are hypothesized to be located in the periplasm (Kamimura et al., 2005). Further analysis of ATP production by At. caldus showed the involvement of electron transport intermediates including the final electron acceptors bo3 and cytochrome c oxidases, and ATP production by a F0F1 ATPase (Dopson et al., 2002). In contrast to Gram-negative acidophiles, little is known about RISC metabolism in Gram-positive acidophiles, such as Sulfobacillus spp.

Many of the RISC-metabolizing enzymes from the Sulfolobales archaea have been characterized (reviewed in Rohwerder and Sand, 2007). The primary difference between archaeal and Gram-negative bacterial RISC metabolism is that S° is disproportionated in the former, as opposed to being oxidized (Kletzin, 1992). The key, initial enzyme is the sulfur oxygenase reductase (SOR) and its mechanism has been elucidated (Veith et al., 2011). In addition, other RISC intermediates (e.g. tetrathionate and thiosulfate) have been identified.

Systems biology insights in RISC metabolism

Due to the difficulties in developing genetic techniques in acidophiles, a large proportion of the hypotheses regarding RISC metabolic pathways in these prokaryotes are based on systems biology. From these studies, putative genes have been assigned to the RISC oxidation and reduction pathways in several microorganisms including At. ferrooxidans (Bruscella et al., 2007; Chi et al., 2007; Amouric et al., 2009; Quatrini et al., 2009), At. caldus (Mangold et al., 2011), Sulfolobus spp. (Chen et al., 2005; Bathe and Norris, 2007), Metallosphaera (M.) sedula (Kappler et al., 2005; Auernik et al., 2008), and Ac. ambivalens (Laska et al., 2003). Although several acidophiles have common RISC substrates, intermediates, and end-points (especially the acidithiobacilli), the genes encoding the different steps are not always common to all.

Aerobic RISC metabolism

At. ferrooxidans RISC oxidation (Fig. 3A) includes sulfide (HS-) oxidation by sulfide/quinone oxidoreductase, encoded by sqr, while genes encoding heterodisulfide reductase (hdrCBA) are upregulated during growth on S° (Quatrini et al., 2009). In addition, genes encoding the suggested rhodanese-like protein (rhd), a SirA like disulfide bond formation regulator protein (tusA), and the heterodisulfide reductase complex family protein (dsrE) are also upregulated and are hypothesized to be involved in S° oxidation and transport (Quatrini et al., 2009). Tetrathionate hydrolysis is mediated by TetH (tetH) and the thiosulfate produced is oxidized by DoxDA (Quatrini et al., 2009). The sulfite produced from S° oxidation is potentially transformed by (among other enzymes) sulfate adenylyltransferase (sat) (Quatrini et al., 2009). Three terminal electron acceptors are upregulated during RISC energy conservation: a bd ubiquinol oxidase (cydAB); a bo3 oxidase (cyoABCD); and the bc1 complex and cytochrome c4, encoded by the petII operon (Quatrini et al., 2006). This suggests a certain elasticity and potential differential regulation during energy conservation. In addition, the presence of Fe2+ downregulates the expression of RISC oxidation genes (Amouric et al., 2009). When At. ferrooxidans was grown on thiosulfate, the RISC metabolism-related periplasmic proteins tetrathionate hydrolase and cytochrome c4 (CycA2) were identified in high throughput proteomics analysis (Chi et al., 2007).

Figure 3.

Models of RISC oxidation derived from systems biology approaches in Acidithiobacillus ferrooxidans (A) and Acidithiobacillus caldus (B). In both models, electrons derived from RISC metabolism are suggested to be channelled to the quinone pool from which they pass along the electron transport chain to multiple terminal electron acceptors. These terminal electron acceptors pump protons out of the cell that can re-enter via the F0F1 ATPase to generate ATP. The figures are reprinted with permission from Quatrini and colleagues (2009) and Mangold and colleagues (2011), respectively.

The genes for At. caldus RISC metabolism share similarities to those of At. ferrooxidans (Fig. 3B) and homologues between the two species include sulfide/quinone oxidoreductase (sqr), tetrathionate hydrolysis (tetH), bd ubiquinol oxidase (cydAB), and bo3 oxidase (cyoABCD) (Valdes et al., 2009; Mangold et al., 2011). However, the At. caldus genome contains two partial sox gene clusters suggested to be involved in thiosulfate oxidation (soxABXYZ), and S° oxidation is mediated by sulfur oxygenase reductase (sor) or heterodisulfide reductase (hdrA-hyp-C-B) (Mangold et al., 2011). The DsrE (dsrE) protein suggested to play a role in S° metabolism is also present in At. caldus (Mangold et al., 2011). TetH and doxD (other dox genes are not present) from the At. caldus tetrathionate hydrolase operon are upregulated in the presence of tetrathionate, also suggesting their role in RISC metabolism (Rzhepishevska et al., 2007).

Annotation of the At. ferrivorans genome sequence identified genes potentially encoding an incomplete SOX complex (soxYZ-hypB) and a sulfur oxygenase : reductase (sor) similar to those in At. caldus (Liljeqvist et al., 2011a). Further analysis also revealed the presence of hdrABC and doxD potentially encoding S° and thiosulfate oxidizing proteins, respectively (M. Dopson and D.S. Holmes, unpubl. data). However, the expression and roles of these genes have not been confirmed experimentally. The genome of the At. thiooxidans type strain has also been sequenced and, as with At. caldus, two partial sox clusters, tetrathionate hydrolase (tetH), and thiosulfate quinone oxidoreductase (doxD) have been identified (Valdes et al., 2011).

Subtractive hybridization and sequencing of cDNAs from S. metallicus cells grown on either ferrous iron- or S° identified the dominant S° transcript to be sulfur oxygenase-reductase (sor), which had previously been identified in S. tokodaii and Acidianus spp. (Bathe and Norris, 2007). Transcript profiles of M. sedula grown heterotrophically on yeast extract, or mixotrophically on yeast extract with either S° or pyrite revealed differential transcription of respiratory complexes (Kappler et al., 2005). M. sedula grown with S° expressed higher levels of soxABCD and a cytochrome b with a possible role in iron oxidation or chemolithotrophy (Kappler et al., 2005). Further transcripts induced during growth on tetrathionate and S° included a putative tetrathionate hydrolase, a novel polysulfide/sulfur/dimethyl sulfoxide reductase-like complex, and a novel Hdr-like complex (Auernik and Kelly, 2008). Finally, the genes involved in thermoacidophile RISC oxidation are summarized in Auernik and Kelly (2010) and include a sor, tetH, and doxDA in S. tokodaii as well as doxDA in S. solfataricus.

Anaerobic RISC metabolism

A model for anaerobic H2 oxidation coupled to S° reduction in the crenarchaeote Ac. ambivalens is shown in Fig. 4. The key enzyme activities in the pathway are mediated by a putative sulfur reductase encoded by the sreABCDE cluster along with a hydrogenase encoded by a 12 gene cluster (Laska et al., 2003). The sulfur reductase and hydrogenase form the electron transport chain, probably linked by quinones, and an electrochemical gradient (to drive ATP synthesis) is likely generated by protons used to reduce Sulfolobusquinone on the cytoplasmic side of the membrane that are released to the outside upon reooxidation (Fig. 4). A putative sulfur reductase has also been identified in a gene cluster in the proteobacterium At. ferrooxidans, which has a similar order to that of Ac. ambivalens (Valdes et al., 2008). A proteomic study of anaerobic S° oxidation coupled to Fe3+ reduction in At. ferrooxidans showed upregulation of sulfide/quinone oxidoreductase (sqr) and proteins involved in cell envelope biogenesis (Kucera et al., 2012). In addition, proteins previously attributed to electron transport during aerobic Fe2+ oxidation, including rusticyanin and cytochrome c552, were also found to be upregulated (Kucera et al., 2012). This observation has been confirmed in a combined proteomic and transcriptomic study of S°-grown At. ferrooxidans carried out by the authors and co-workers (D.B. Johnson, V. Bonnefoy, M. Dopson and D.S. Holmes, unpubl. data).

Figure 4.

Anaerobic growth via sulfur respiration in Ac. ambivalens (figure redrawn from Laska et al., 2003). CM, cytoplasmic membrane; SQ, sulfolobusquinone.

Sulfur assimilation

Sulfur (usually as sulfate) is assimilated from the environment and incorporated into the amino acids methionine and cysteine, iron-sulfur centres, and other metabolites (reviewed in Aguilar-Barajas et al., 2011). Extracellular sulfate may be transported across the membrane either via an ABC uptake system (CysU and CysA) or a SulP sulfate permease (Valdes et al., 2003). The assimilation of sulfate into At. ferrooxidans bio-molecules (e.g. amino acids) has been demonstrated experimentally (Tuovinen et al., 1975) and a bioinformatic reconstruction of sulfate assimilation presented (Valdes et al., 2003). Production of cysteine is via the intermediates adenosine-5′-phosphosulfate (APS), sulfite, and sulfide and is encoded by the cysJIHDNG operon (Valdes et al., 2003). The At. ferrooxidans APS reductase was cloned and expressed and its activity confirmed (Zheng et al., 2009). In addition, a potential pathway for sulfation of metabolites by the PAPS pathway is predicted to be encoded by two non-identical fused copies of cysNC (Valdes et al., 2003) previously described in a Rhizobium sp. BR816 (Laeremans et al., 1997). Finally, APS reductase and sulfate adenylyl transferase (SAT) are present on the M. sedula genome although as a mixotroph, it is possible that these proteins act in the reverse direction to sulfate assimilation to produce ATP (Auernik et al., 2008).

Applied aspects

Role of acidophilic sulfur-metabolizing prokaryotes in mineral processing

Bio-processing of metal ores and concentrates (‘biomining’) has developed into a major area of biotechnology since it was first applied, in the 1960s, to recover copper from low-grade ‘run of mine’ ore in massive waste rock dumps in the USA (Brierley, 2008). Currently, metals including copper, cobalt, nickel, uranium and gold are recovered from primary ores and mine wastes in full-scale commercial operations. Biomining harnesses the ability of some species of acidophilic prokaryotes to generate ferric iron and sulfuric acid, and thereby create conditions that cause the oxidative dissolution of sulfides and some other minerals, such as uraninite (UO2; (Rawlings and Johnson, 2007b; Johnson, 2010). In cases where target metals are solubilized (e.g. copper) the process is known as ‘bioleaching’, whereas if the metal becomes accessible but remains in the solid phase (e.g. gold) it is referred to as ‘bio-oxidation’. While the primary microorganisms involved in current biomining operations are iron-oxidizing prokaryotes, secondary sulfur-oxidizers play a critical role by: (i) generating sulfuric acid that maintains the acidic conditions (pH ∼ 1–3) required by the iron-oxidizers (ii) removing sulfur rich layers on the mineral surface that can hinder metal dissolution, and (iii) retaining the metals released from the bioleached minerals in solution, which facilitates their recoveries in downstream processing operations (Dopson and Lindström, 1999; Rawlings and Johnson, 2007a). In addition, some mineral-degrading acidophiles, such as At. ferrooxidans, At. ferrivorans and Sulfobacillus spp. oxidize both ferrous iron and reduced sulfur and so have potential dual roles in biomining operations.

Much of the current knowledge of the transformations of sulfur that occur during the oxidative dissolution of sulfide minerals have come from the work of Sand and Schippers (Schippers et al., 1996; Schippers and Sand, 1999). They noted that metal sulfides could be categorized as ‘acid insoluble’ or ‘acid soluble’, and that these formed different sulfur intermediates when they were (microbiologically) oxidized at low pH. Acid-insoluble sulfides, such as pyrite, are attacked by ferric iron, and a total of six successive one-electron oxidation steps are required to break the sulfur-metal bonds. The initial sulfur product released from the degrading mineral is thiosulfate, and therefore this form of oxidative sulfide mineral dissolution has been described as the ‘thiosulfate mechanism’. Thiosulfate is oxidized by ferric iron to tetrathionate, which hydrolyses to form highly reactive sulfane-monosulfonic acid (HS3O3-) and sulfate. A variety of sulfur anions, and S°, are formed from reactions involving sulfane-monosulfonic acid, which are utilized as electron donors by sulfur-oxidizing acidophiles. In contrast, metal-sulfur bonds in acid-soluble sulfides (e.g. chalcocite; Cu2S) can be disrupted by protons as well as by ferric iron, liberating hydrogen sulfide, although more commonly, in the presence of ferric iron, the first free sulfur compound is thought to be an unstable sulfide cation (H2S+). This dimerizes to H2S2, which oxidizes, via various polysulfides, to S°, a process that Schippers and Sand (1999) referred to as the ‘polysulfide mechanism’. The various transformations of sulfur that occur during the oxidative dissolution of sulfide minerals are shown in Fig. 1.

While all current commercial biomining operations operate by bio-engineering mineral oxidation, such an approach is not applicable to ores that are already oxidized. These include the ‘nickel laterites’ (estimated to account for about 72% of the reserves of this base metal in the lithosphere), where much of the nickel is intimately associated with ferric iron minerals (chiefly goethite: FeO.OH). A novel approach for extracting nickel from limonitic ores, which operates via microbially catalysed reductive mineral dissolution, has been described (Hallberg et al., 2011). In this, oxidation of S° by an acidophilic bacterium (At. ferrooxidans) is coupled to the reduction of the ferric iron present in the mineral phase. This results in the dissolution of goethite and release of the associated nickel, which is retained in solution in the acidic leach liquor. Microbially catalysed reductive dissolution of goethite underpins the ‘Ferredox’ process, an integrated process described for extracting nickel from lateritic ores (du Plessis et al., 2011).

RISC oxidation in mining process waters

Wastewaters from sulfide mining treatment processes often contain high concentrations of RISCs (primarily S4O62− and S2O32−) that are created by several mechanisms that include reactions that occur during mineral grinding and flotation, and anaerobic reactions between hydroxide ions, pyrite and S° (Liljeqvist et al., 2011b). If left untreated, these RISCs may be microbially oxidized in recipient water bodies, causing acidification of the environment and oxygen depletion of water bodies (Silver and Dinardo, 1981; Li and Boucher, 1999).

One method to circumvent environmental problems associated with the release of RISCs from process waters is chemical treatment (Ikumapayi et al., 2009; Lagace, 2010). However, this invokes high costs as well as the need to use noxious chemical reagents. An alternative method is to remove the acid-generating potential before release of the wastewaters via biological oxidation of RISCs. A complicating factor in some geographical locations is the combination of large volumes of wastewater combined with low ambient temperatures (e.g. northern Sweden) that requires the RISC oxidation to be carried out at 4–10°C for much of the year. Continuous RISC removal has been achieved at 37°C using a mixed culture of acidophiles including At. caldus, At. thiooxidans, and a Sulfobacillus sp. (Sääf et al., 2009). Lowering the reactor temperature and inoculation with a cold temperature sulfide mine culture resulted in the dominating microorganism to be the psychrotolerant acidophile At. ferrivorans (Liljeqvist et al., 2011b). This confirmed an earlier study showing that At. ferrivorans was capable of low temperature RISC oxidation (Kupka et al., 2009).

Selective metal capture by acidophilic sulfidogens

One of the biotechnologies currently under development targets the selective removal of chalcophilic metals in waste streams (e.g. AMD) or mine process waters (e.g. pregnant leach liquors) using novel species of sulfate-reducing bacteria (SRB) that are either acid-tolerant or acidophilic. Many transition metals, and metalloids such as arsenic, which occur in mine waters form sulfide phases that have different solubility products. The concentration of the key reactant (S2−) can be controlled by varying solution pH [H2S ↔ HS- (pKa∼ 7); HS- ↔ S2− (pKa∼ 12)]. In extremely low pH liquors (e.g. pH 2), only metal sulfides that have very small solubility products (e.g. CuS) will form whereas other sulfides will only form at higher pH values (e.g. FeS, which requires a minimum pH of ∼ 6). The ‘BioSulphide’ process (Bratty et al., 2006) segregates metal sulfides in pH-controlled contact reactors, although here the sulfidogens are neutrophilic sulfur-reducers rather than SRB, and are cultivated in an off-line bioreactor to prevent contact with acidic liquors, which would otherwise inhibit or kill the bacteria. Recently, Ňancucheo and Johnson (2012) described a low pH bio-sulfidogenic system that involves consortia of acidophilic SRB that can be used to generate sulfide and to selectively precipitate metals from acidic liquors within a single bioreactor, thereby reducing both engineering complexity and operating costs. Tests with synthetic AMD waters (pH 2–2.5) that contained a variety of transition metals and aluminium demonstrated that copper and zinc could be selectively precipitated (as sulfides) by maintaining bioreactors at fixed pH values (ranging from pH 2.2 to 4.0), thereby facilitating recovery and recycling of target metals.

Acknowledgements

MD acknowledgements the Swedish Research Council (Vetenskapsradet Contract No. 621–2007-3537) for funding research.

Ancillary